text
stringlengths 4
429
|
---|
vestigate: attacks against activists, dissidents, trade unionists, |
likewise elicited significant research interest. Where these two |
human rights campaigners, journalists, and members of NGOs |
problem domains intersect, however |
targeted cyber attacks by |
(henceforth |
targets |
) in the Middle East. The attacks we have |
nation-states against individuals |
has received virtually no sig- |
documented usually involve the use of malicious links or e-mail |
nificant, methodical research attention to date. This new prob- |
attachments, designed to obtain information from a device. On |
lem space poses challenges that are both technically complex |
the one hand, we have observed attacks using a wide range of |
and of significant real-world importance. |
off-the-shelf spyware, as well as publicly available third-party |
In this work we undertake to characterize the emergent prob- services, like iplogger.org. On the other hand, some at- |
lem space of nation-state Internet attacks against individuals tacks use so-called |
lawful intercept |
trojans and related equip- |
engaged in pro-democracy or opposition movements. While |
we lack the data to do so in a fully comprehensive fashion, 1 Dates in the paper are given MM/DD/YY. |
USENIX Association |
23rd USENIX Security Symposium |
511 |
ment, purportedly sold exclusively to governments by compa- do so in a sufficiently well-grounded, meaningful manner first |
nies like Gamma International and Hacking Team. The lat- requires developing an understanding of the targets |
knowledge |
ter advertises that governments need its technology to |
look of security issues, their posture regarding how they currently |
through their target |
s eyes |
rather than rely solely on |
passive protect themselves, and the resources (including potentially ed- |
monitoring |
[1]. Overall, the attacks we document are rarely ucation) that they can draw upon. To this end, we are now con- |
technically novel. In fact, we suspect that the majority of at- ducting (with IRB approval) in-depth interviews with potential |
tacks could be substantially limited via well-known security targets along with systematic examination of their Internet de- |
practices, settings, and software updates. Yet, the attacks are vices in order to develop such an understanding. |
noteworthy for their careful social engineering, their links to |
governments, and their real-world impact. |
We obtained the majority of our artifacts by encouraging in- 2 Related Work |
dividuals who might be targeted by governments to provide us |
with suspicious files and unsolicited links, especially from un- In the past decades, a rich body of academic work has grown to |
familiar senders. While this process has provided a rich set of document and understand government Internet censorship, in- |
artifacts to analyze, it does not permit us to claim our dataset is cluding nationwide censorship campaigns like the Great Fire- |
representative. wall of China [9, 10, 11]. Research on governmental Internet |
Our analysis links these attacks with a common class of ac- surveillance and activities like law-enforcement interception is |
tor: an attacker whose behavior, choice of target, or use of in- a comparatively smaller area [12]. Some academic work looks |
formation obtained in the attack, aligns with the interests of a at government use of devices to enable censorship, such as key- |
government. In some cases, such as Ali |
s, the attackers appear word blacklists for Chinese chat clients [13], or the Green Dam |
to be governments themselves; in other cases, they appear in- censorware that was to be deployed on all new computers sold |
stead to be pro-government actors, ranging from patriotic, not in China [14]. We are aware of only limited previous work |
necessarily skilled volunteers to cyber mercenaries. The phe- looking at advanced threat actors targeting activists with hack- |
nomenon has been identified before, such as in Libya, when ing, though this work has not always been able to establish ev- |
the fall of Gaddafi |
s regime revealed direct government ties to idence of government connections [15]. |
hacking during the 2011 Civil War [2]. Platforms used by potential targets, such as GMail [16], |
We make the following contributions: Twitter [17], and Facebook [18] increasingly make transport- |
We analyze the technology associated with targeted at- layer encryption the default, obscuring communications from |
tacks (e.g., malicious links, spyware), and trace it back most network surveillance. This use of encryption, along with |
to its programmers and manufacturers. While the attacks the global nature of many social movements, and the role of |
are not novel |
and indeed often involve technology used diaspora groups, likely makes hacking increasingly attractive, |
by the cybercrime underground |
they are significant be- especially to states who are unable to request or compel content |
cause they have a real-world impact and visibility, and from these platforms. Indeed, the increasing use of encryption |
are connected to governments. In addition, we often find and the global nature of targets have both been cited by pur- |
amateurish mistakes in either the attacker |
s technology or veyors of |
lawful intercept |
trojans in their marketing materi- |
operations, indicating that energy spent countering these als [1, 19]. In one notable case in 2009, UAE telecom firm Eti- |
threats can realize significant benefits. We do not, how- salat distributed a system update to its then 145,000 BlackBerry |
ever, conclude that all nation-state attacks or attackers subscribers that contained spyware to read encrypted Black- |
are incompetent, and we suspect that some attacks have Berry e-mail from the device. The spyware was discovered |
evaded our detection. when the update drastically slowed users |
phones [20]. In con- |
trast to country-scale distribution, our work looks at this kind of |
When possible, we empirically characterize the attacks |
pro-government and government-linked surveillance through |
and technology we have observed. We map out global |
highly targeted attacks. |
use of two commercial hacking tools by governments by |
searching through Internet scan data using fingerprints for The term APT (Advanced Persistent Threat) refers to a |
command-and-control (C&C) servers derived from our sophisticated cyber-attacker who persistently attempts to tar- |
spyware analysis. get an individual or group [21]. Work outside the academic |
community tracking government cyberattacks typically falls |
We develop strong evidence tying attacks to govern- under this umbrella. There has been significant work on |
ment sponsors and corporate suppliers, countering de- APT outside the academic community, especially among se- |
nials, sometimes energetic and sometimes indirect, of curity professionals, threat intelligence companies, and human |
such involvement [3, 4, 5, 6], in contrast to denials [7] rights groups. Much of this work has focused on suspected |
or claims of a corporate |
oversight |
board [8]. Our scan- government-on-government or government-on-corporation cy- |
ning suggests use of |
lawful intercept |
trojans by 11 ad- ber attacks [22, 23]. Meanwhile, a small but growing body |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.