A Study on Advanced Persistent Threats

Ping Chen, Lieven Desmet, Christophe Huygens

To cite this version:

Ping Chen, Lieven Desmet, Christophe Huygens. A Study on Advanced Persistent Threats. 15th
IFIP International Conference on Communications and Multimedia Security (CMS), Sep 2014, Aveiro,
Portugal. pp.63-72, �10.1007/978-3-662-44885-4_5�. �hal-01404186�

HAL Id: hal-01404186

https://inria.hal.science/hal-01404186v1
Submitted on 28 Nov 2016

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.

Distributed under a Creative Commons Attribution 4.0 International License

 A study on Advanced Persistent Threats

Ping Chen, Lieven Desmet, and Christophe Huygens

iMinds-DistriNet, KU Leuven
3001 Leuven, Belgium
{firstname.lastname}@cs.kuleuven.be

Abstract A recent class of threats, known as Advanced Persistent Threats

(APTs), has drawn increasing attention from researchers, primarily from
the industrial security sector. APTs are cyber attacks executed by soph-
isticated and well-resourced adversaries targeting specific information in
high-profile companies and governments, usually in a long term campaign
involving different steps. To a significant extent, the academic community
has neglected the specificity of these threats and as such an objective ap-
proach to the APT issue is lacking. In this paper, we present the results
of a comprehensive study on APT, characterizing its distinguishing char-
acteristics and attack model, and analyzing techniques commonly seen
in APT attacks. We also enumerate some non-conventional countermeas-
ures that can help to mitigate APTs, hereby highlighting the directions
for future research.

Keywords: advanced threat, APT, sophisticated attacks, cyber security

1 Introduction
Cyber attacks have existed since the adoption of the Internet and have evolved
a lot in the past decades, from viruses and worms in the early days to malware
and botnets nowadays. In recent years, a new class of threat, the “Advanced
Persistent Threat” (APT) has emerged. Originally used to describe cyber in-
trusions against military organizations, the APT has evolved and is no longer
limited to the military domain. As highlighted in several large-scale security
breaches [12,15,1,29], APTs are now targeting a wide range of industries and
governments.
While APT has drawn increasing attention from the industrial security com-
munity, a comprehensive and clear understanding of the APT research problem
is lacking. This paper presents the result of a detailed study of the APT phe-
nomenon, and contributes a taxonomy of phases, mechanisms, and countermeas-
ures. In this paper, we first identify the characteristics of APT, and compare it
to traditional threats in Section 2. In Section 3, we dissect a typical APT attack
into six phases, analyzing the techniques that are commonly used in each stage.
We also enumerate various countermeasure that can be applied to defend against
APT attacks. In Section 3.2, we provide case studies of four APTs, illustrating
the adversaries’ tactics and techniques by applying our presented taxonomy and
technical analysis.
2 Definition: what is APT?

APTs frequently made global headlines in recent years, and many feel that this
term is overloaded, since different people refer to it as different things. Because
so many different opinions of what constitutes an APT exist in the commercial
market [2,14,23], a clear definition is needed. In this paper, we adopt the defini-
tion given by US National Institute of Standards and Technology (NIST), which
states that an APT is [17]:
“An adversary that possesses sophisticated levels of expertise and significant
resources which allow it to create opportunities to achieve its objectives by using
multiple attack vectors (e.g., cyber, physical, and deception). These objectives
typically include establishing and extending footholds within the information
technology infrastructure of the targeted organizations for purposes of exfiltrat-
ing information, undermining or impeding critical aspects of a mission, program,
or organization; or positioning itself to carry out these objectives in the future.
The advanced persistent threat: (i) pursues its objectives repeatedly over an ex-
tended period of time; (ii) adapts to defenders’ efforts to resist it; and (iii) is
determined to maintain the level of interaction needed to execute its objectives”.
This definition provides a good base for distinction between traditional threats
and APTs. The distinguishing characteristics of APTs are: (1) specific targets
and clear objectives; (2) highly organized and well-resourced attackers; (3) a
long-term campaign with repeated attempts; (4) stealthy and evasive attack
techniques. We elaborate on each of these characteristics below.
Specific targets and clear objectives APT attacks are highly targeted at-
tacks, always having a clear goal. The targets are typically governments or organ-
izations possessing substantial intellectual property value. Based on the number
of APT attacks discovered by FireEye in 2013 [11], the top ten industry vertical
targets are education, finance, high-tech, government, consulting, energy, chem-
ical, telecom, healthcare, and aerospace. While traditional attacks propagate as
broadly as possible to improve the chances of success and maximize the harvest,
an APT attack only focuses on its pre-defined targets, limiting its attack range.
As for the attack objectives, APTs typically look for digital assets that bring
competitive advantage or strategic benefits, such as national security data, intel-
lectual property, trade secrets, etc., while traditional threats mostly search for
personal information like credit card data, or generically valuable information
that facilitates financial gain.
Highly organized and well-resourced attackers The actors behind APTs
are typically a group of skilled hackers, working in a coordinated way. They may
work in a government/military cyber unit [15], or be hired as cyber mercenaries
by governments and private companies [9]. They are well-resourced from both
financial and technical perspectives. This provides them with the ability to work
for a long period, and have access (by development or procurement) to zero-day
vulnerabilities and attack tools. When they are state-sponsored, they may even
operate with the support of military or state intelligence.
A long-term campaign with repeated attempts An APT attack is typically
a long-term campaign, which can stay undetected in the target’s network for
several months or years. APT actors persistently attack their targets and they
repeatedly adapt their efforts to complete the job when a previous attempt fails.
This is different from traditional threats, since traditional attackers often target
a wide range of victims, and they will move right on to something less secure if
they cannot penetrate the initial target.
Stealthy and evasive techniques APT attacks are stealthy, possessing the
ability to stay undetected, concealing themselves within enterprise network traffic,
and interacting just enough to achieve the defined objectives. For example, APT
actors may use zero-day exploits to avoid signature-based detection, and encryp-
tion to obfuscate network traffic. This is different from traditional attacks, where
the attackers typically employ “smash and grab” tactics that alert the defenders.
In Table 1, we summarize the differences between traditional threats and
APTs for several attack attributes.

Traditional Attacks APT Attacks

Attacker Mostly single person Highly organized, sophisticated,
determined and well-resourced group
Target Unspecified, mostly individual Specific organizations, governmental
systems institutions, commercial enterprises
Purpose Financial benefits, Competitive advantages,
demonstrating abilities strategic benefits
Approach Single-run, “smash and grab”, Repeated attempts, stays low and slow,
short period adapts to resist defenses, long term

Table 1: Comparison of traditional and APT attacks

3 Attack Model: how does APT work?

APT attacks are meticulously planned, and typically have multiple steps in-
volved. While a specific APT attack may have its unique features, the stages of
APT attacks are similar and they differ mostly in the techniques used in each
stage. To describe the phases of an APT attack, we adopt a six-stage model
based on the concept of an “intrusion kill chain” introduced in [7]. Using such
a kill chain model helps to understand threat actors’ techniques in each stage,
and provides guidance for defense against APT attacks as well.

3.1 Phases of an APT attack

A typical ATP attack will have the following six phases: (1) reconnaissance and
weaponization; (2) delivery; (3) initial intrusion; (4) command and control; (5)
lateral movement; (6) data exfiltration.
(1) Reconnaissance and Weaponization Reconnaissance is also known as
information gathering, which is an important preparation step before launching
attacks. In this stage, attackers identify and study the targeted organization,
collecting as much as information possible about the technical environment and
key personnel in that organization. This information is often gathered via open-
source intelligence (OSINT) tools and social engineering techniques.

– Social Engineering. Social engineering refers to psychological manipula-

tion of people into accomplishing goals that may or may not be in the target’s
best interest. In cyber attacks, it is often used for obtaining sensitive inform-
ation, or getting the target to take certain action (e.g. executing malware).
– OSINT. OSINT is a form of intelligence collection from publicly available
sources, and nowadays it typically refers to aggregating information about a
subject via either paid or free sources on the internet. Various information
can be collected via OSINT, ranging from the personal profile of an employee
to the hardware and software configurations in an organization.

Besides simply grabbing information from the web, attackers may also em-
ploy data mining techniques and big data analytics to automatically process the
gathered data, in order to produce actionable intelligence. Based on the gathered
intelligence, APT actors construct an attacking plan and prepare the necessary
tools. In order to be successful, attackers typically prepare various tools for dif-
ferent attack vectors, so that they can adapt tactics in case of failure.
(2) Delivery In this stage, attackers deliver their exploits to the targets. There
are two types of delivery mechanisms: direct and indirect delivery. For direct
delivery, the attackers send exploits to their targets via various social engineering
techniques, such as spear phishing.
Indirect delivery is stealthy. In this approach the attackers will compromise
a 3rd party that is trusted by the target, and then use the compromised 3rd
party to indirectly serve exploits. A trusted 3rd party can be a supplier of soft-
ware/hardware used in the targeted organization, or a legitimate website that is
frequently visited by the targeted persons (watering hole attack).

– Spear Phishing. Spear phishing is a targeted form of phishing in which

fraudulent emails only target a small group of selected recipients. It typ-
ically use information gathered during reconnaissance to make the attack
more specific and “personal” to the target, in order to increase the probabil-
ity of success. The recipient is lured to either download a seemingly harmless
attachment that contains a vulnerability exploit, or to click a link to a mali-
cious site serving drive-by-download exploits [27]. In APT attacks, malicious
attachments are used more often than malicious links, as people normally
share files (e.g., reports, business documents, and resumes) via email in the
corporate or government environment.
– Watering Hole Attack The concept of a watering hole attack is similar
to a predator waiting at a watering hole in a desert, as the predator knows
that the victims will have to come to the watering hole. Similarly, rather
 than actively sending malicious emails, the attackers can identify 3rd party
websites that are frequently visited by the targeted persons, and then try to
infect one or more of these websites with malware. Eventually, the delivery
accomplishes when the infected webpages are viewed by victims [18]. The use
of watering hole attacks have been seen in several APT campaigns [5,6,10].

(3) Initial Intrusion Initial intrusion happens when the attacker get a first
unauthorized access to the target’s computer/network. While the attackers may
obtain access credentials through social engineering, and simply use them for
“legitimate” access, the typical way for intrusion is executing malicious code
that exploits a vulnerability in the target’s computer. The attackers first deliver
malicious code in the delivery stage, and then in the intrusion stage gain access
to target’s computer when the exploit is successfully executed.
In APT attacks, the attackers often focus on vulnerabilities in Adobe PDF,
Adobe Flash and Microsoft Office as well as Internet Explorer. While several
APT attacks [12,20] have leveraged zero-day exploits for initial intrusion, many
APT attacks also employ older exploits that target unpatched applications.
The initial intrusion is a pivotal phase in an APT attack, since the APT
actors establish a foothold in the target’s network in this stage. A successful
intrusion typically results in the installation of a backdoor malware. From this
point, the threat actors connects to the targets’ network. As a result, network
traffic is generated, and file evidences are left on the victims’ computers, which
gives defenders the chance to detect an APT in an early phase.
(4) Command and Control Upon successfully establishing a backdoor, APT
actors use Command and Control (C2) mechanisms to take control of the com-
promised computers, enabling further exploitation of the network. In order to
evade detection, the attackers increasingly make use of various legitimate services
and publicly available tools.
– Social networking sites. The attackers register accounts on various so-
cial networking sites, and put control information into blog posts or status
messages [16].
– Tor anonymity network. Servers configured to receive inbound connec-
tions only through Tor are called hidden services. Hosting C2 servers in Tor
as hidden services makes them harder to identify, blacklist or eliminate.
– Remote access tools (RATs) Although often used for legitimate remote
administration, RATs are often associated with cyber attacks [3,28]. A RAT
contains two components: a “server” residing on a victim’s endpoint, and a
“client” that is installed on the attackers machine. In order to make it work,
the “server” component needs to be delivered to the target’s machine first,
which is often accomplished via spear-phishing emails.

(5) Lateral Movement Once the communication between the compromised

systems and C2 servers is established, threat actors move inside the network,
in order to expand their control over the targeted organization, which in turn
enables them to discover and collect valuable data. Lateral movement usually
involves the following activities: (1) performing internal reconnaissance to map
the network and acquire intelligence; (2) compromising additional systems in
order to harvest credentials and gain escalated privileges; (3) identifying and
collecting valuable digital assets, such as development plans, trade secrets, etc..
This stage typically lasts a long period, because (1) the attackers want to
harvest a maximum of information over a long term; (2) the activities are de-
signed to run low and slow in order to avoid detection. As APT actors move
deeper into the network, their movements become difficult to detect. APT act-
ors often utilize legitimate OS features and tools that are typically used by IT
administrators, and they may also crack or steal credentials to gain legitimate
access, which both make their activities undetectable or even untraceable.
(6) Data Exfiltration The primary goal for an APT attack is to steal sensitive
data in order to gain strategic benefits, thus data exfiltration is a critical step for
the attackers. Typically the data is funneled to an internal staging server where it
is compressed and often encrypted for transmission to external locations under
the attackers’ control. In order to hide the transmission process, APT actors
often use secure protocols like SSL/TLS, or leverage the anonymity feature of
Tor network [16].

3.2 Case study of APT attacks

In order to better understand the APT attack model, we studied four APT
attacks reported in various sources [12,20,29,10], mapping the attackers’ action
into our six-stage model. The results are shown in Table 2.

3.3 Countermeasures

Due to the complexity and stealthiness of APTs, there is no single solution that
offers effective protection. The current best practice is a wide range of security
countermeasures resulting a multi-layered defense. However, due to the specific
nature of APTs, some of the existing defense systems need to be reengineered
to work in the APT context, hereby requiring additional research. For example,
while genetic algorithms have been proved useful for malware detection, their
applicability in a large dataset is subject of further study. We elaborate on some
defense techniques below.
Security Awareness Training Considering the wide use of social engineering
techniques (e.g., spear-phishing emails) in APT campaigns, security awareness
training plays an important role in defense. Besides the general best security
practices, the training should also provide education about APT attacks. Ac-
cording to an APT awareness study [8], more than half of the industries are not
awareness of the differences between APTs and traditional threats, and 67% of
respondents report the lack of awareness training relative to APTs.
Traditional Defense Mechanisms Traditional defense mechanisms are ne-
cessary since they block known attack vectors, and hence increase the difficulty
Name Operation RAS Breach [20] Operation Operation
Aurora [12] Ke3chang [29] SnowMan [10]
Active June 2009 - Unknown - May 2010 - Unknown -
Time December 2009 March 2011 December 2013 February 2014
Recon. and employees’ emails, employees’ emails, officials’ emails, identify weakness
Weaponi- zero-day exploits, zero-day exploits, trojanized docs, in vfw.org, RAT,
zation backdoor, and C2 trojanized docs, backdoor, and C2 backdoor
tools backdoor, RAT tools
Delivery spear phishing spear phishing spear phishing watering hole at-
(malicious links) (malicious xls file) (malicious zip file)tack (compromise
& infect vfw.org)
Initial drive-by download xls vulnerability victims open the drive-by download
Intrusion (CVE-2010-0249) (CVE-2011-0609) executable file (CVE-2014-0322)
Command custom C2 pro- Poison Ivy RAT custom C2 pro- ZxShell,
and Con- tocol, operating tocol, based on Gh0st RAT
trol on TCP port 443 HTTP protocol
Lateral compromise SCM, Perform privilege compromise in- unknown
Movement and obtain source escalation, gather ternal systems,
code SecureID data collect data
Data upload data to C2 compress, encrypt compress, encrypt unknown, could
Exfiltration servers data as RAR files, data as RAR files be US military
use FTP for trans- intelligence
mission

Table 2: Comparison of different APTs

for APT actors. Common countermeasures that must be used are: patch man-
agement, anti-virus software, firewalls, host-based intrusion detection systems
(HIDS), network-based intrusion detection systems (NIDS), intrusion prevention
system (IPS), Security Information and Event Management (SIEM), content fil-
tering software, etc..
Security awareness training and traditional defense mechanisms do not ad-
equately address APTs. Defenders should combine them with the following state-
of-the-art countermeasures that are proposed to mitigate APTs.

Advanced Malware Detection Malware is critical for the initial intrusion.

Since APT actors often leverage zero-day exploits or custom-developed evasive
tools that bypass traditional defenses, the ability to detect advanced malware is
important for defense against APTs. Sandboxing execution is a proven technique
for analyzing malware’s behavior, which allows defenders to identify unknown
advanced malware [19]. As advanced malware may leverage various sandbox-
evasion techniques [22] to detect the VM environment, it is important to take
these sandbox-evasion techniques into consideration when using sandboxing ex-
ecution. Also, the research challenge in this area is to perform the malware
analysis on-line, and in a non-intrusive fashion.
Event Anomaly Detection Since APT actors use various stealthy and evasive
techniques, there is no “known bad” pattern that traditional signature-based de-
fense mechanisms could use. Instead of looking for “known bad” item, an effective
APT detection approach is to study normal behavior and search for anomalous
activities. Anomaly detection includes the detection of suspicious network traffic
and suspicious system activities, or “irregular” clusters of activities (potentially
obtained through machine learning). Due to the massive amount of data the
need to be analyzed in a reasonable time, anomaly detection typically relates
to the research problem of big data analytics. There are several researchers pro-
posing the use of big data analytic for APT detection. In [4], Giura & Wang
implemented a large-scale distributed computing framework based on MapRe-
duce to process all possible events, which can be used to detect APT attacks.
Liu et. al. [13] proved that analyzing a huge volume of HTTP requests with
Hadoop and Lucene can help to quickly uncover potential victims based on a
known APT victim.
Data Loss Prevention Since the ultimate goal of an APT attacks is the trans-
mission of valuable data from the target’s network to outside, a fully contextually
aware data loss prevention (DLP) solution can be deployed as the last line of
defense to protect sensitive data against exfiltration. A DLP solution is a sys-
tem that is designed to detect and prevent potential data breach by monitoring
and blocking sensitive data while in-use, in-motion, and at-rest. It requires the
defender to identify its sensitive and critical data first, and define policies and
rules in a DLP application for protection. An example research solution is [21].
Intelligence-driven Defense Intelligence-driven defense is not a specific de-
fense solution, it is a defense strategy that leverage the knowledge about the
adversaries, and adapt defense based on the gathered intelligence [7]. Since APT
actors are determined, and typically launch repeated attacks against the target,
defenders can create an intelligence feedback loop, which allow them to identify
patterns of previous intrusion attempts, understand the adversaries’ techniques,
and then implement countermeasures to reduce the risk of subsequent intrusions.
In Table 3, we summarize the attack techniques and tools that commonly
seen in each stage of an APT attack. Additionally, we also identify the counter-
measures that can be applied in each stage.

4 Related Work

Existing research on APTs are mostly from industrial security community. Tra-
ditional security service providers (e.g., McAfee, Symantec) and emerging APT-
focused companies (e.g., FireEye, Mandiant) regularly publish technical reports
that document cases of APT attacks [18,1,15,11]. In [26], Thonnard et al. con-
ducted an in-depth analysis of of 18,580 email attacks that were identified as
targeted attacks by Symantec, and through the analysis, they showed that a
targeted attack is typically a long-running campaign highly focusing on a lim-
ited number of organizations.
Stages Attack techniques/tools Countermeasures
Reconnaissance and OSINT, Social engineering Security awareness training,
Weaponization Preparing malware Patch management, Firewall
Delivery Spear phishing, Content filtering software,
Watering hole attack NIDS, Anti-virus software
Initial Intrusion Zero-day exploits, Patch management, HIDS,
Remote code execution Advanced malware detection
Command and Exploiting legitimate services, NIDS, SIEM,
Control RAT, Encryption Event Anomaly detection
Lateral Movement Privilege Escalation, Access control, HIDS, NIDS,
Collecting data Event Anomaly detection
Data Exfiltration Compression, Encryption, Data Loss Prevention
Intermediary Staging

Table 3: Attack techniques and countermeasures in each stage of an APT attack

There are several articles [24,25] that briefly explained APT attacks and
discussed the detection techniques. However, they are not as comprehensive as
our presented analysis. As for the countermeasures, several academic researchers
proposed the use of big data analytics for APT detection [4,13].

5 Conclusion

APTs are sophisticated, specific and evolving threats, yet certain patterns can be
identified in the their process. In this paper, we focused on the identification of
these commonalities. Traditional countermeasures are needed but not sufficient
for the protection against APTs. In order to mitigate the risks posed by APTs,
defenders have to gain a baseline understanding of the steps and techniques
involved in the attacks, and develop new capabilities that address the specifics
of APT attacks. By studying public APT cases and the offerings of the security
industry, we presented this broad perspective on APT, which should establish
common ground within the security community and provide guidance for further
defensive research.

Acknowledgements We want to thank the anonymous reviewers for the valu-

able comments. This research is partially funded by the Research Fund KU
Leuven, iMinds, IWT, and by the EU FP7 projects WebSand, NESSoS and
STREWS. With the financial support from the Prevention of and Fight against
Crime Programme of the European Union (B-CCENTRE).

References

1. Dmitri Alperovitch. Revealed: Operation Shady RAT. 2011.

 2. Richard Bejtlich. What Is APT and What Does It Want? http://taosecurity.
blogspot.be/2010/01/what-is-apt-and-what-does-it-want.html, 2010.
3. James T. Bennett et al. Poison Ivy: Assessing Damage and Extracting Intelligence.
2013.
4. Paul Giura and Wei Wang. Using large scale distributed computing to unveil
advanced persistent threats. SCIENCE, 1(3), 2013.
5. Will Gragido. Lions at the Watering Hole – The “VOHO” Affair. http://blogs.
rsa.com/lions-at-the-watering-hole-the-voho-affair/, 2012.
6. Thoufique Haq and Yasir Khalid. Internet Explorer 8 Exploit Found in Watering
Hole Campaign Targeting Chinese Dissidents. 2013.
7. E. M. Hutchins et al. Intelligence-Driven Computer Network Defense Informed by
Analysis of Adversary Campaigns and Intrusion Kill Chains. In Proceedings of the
6th International Conference on Information Warfare and Security, 2013.
8. ISACA. Advanced Persistent Threat Awareness. 2013.
9. Kaspersky. The Icefog APT: A Tale of Cloak and Three Daggers. 2013.
10. Darien Kindlund et al. Operation SnowMan: DeputyDog Actor Compromises US
Veterans of Foreign Wars Website. 2014.
11. FireEye Labs. Fireeye advanced threat report 2013. 2014.
12. McAfee Labs. Protecting Your Critical Assets: Lessons Learned from “Operation
Aurora”. 2010.
13. Shun-Te Liu et al. A novel search engine to uncover potential victims for apt in-
vestigations. In Proceedings of the 10th IFIP International Conference on Network
and Parallel Computing, 2013.
14. Mandiant. The Advanced Persistent Threat. 2010.
15. Mandiant. APT1: Exposing One of China’s Cyber Espionage Unit. 2013.
16. Information Warfare Monitor and Shadowserver Foundation. Shadows in the
Cloud: Investigating Cyber Espionage 2.0. 2010.
17. NIST. Managing Information Security Risk: Organization, Mission, and Informa-
tion System View. SP 800-39, 2011.
18. Gavin O’Gorman and Geoff McDonald. The Elderwood Project. 2012.
19. M. Zubair Rafique et al. Evolutionary algorithms for classification of malware
families through different network behaviors. In Proceedings of the Genetic and
Evolutionary Computation Conference, 2014.
20. Uri Rivner. Anatomy of an Attack. https://blogs.rsa.com/
anatomy-of-an-attack/, 2011.
21. Matthew Schmid et al. Protecting data from malicious software. In Proceedings of
the 18th Annual Computer Security Applications Conference. IEEE, 2002.
22. Abhishek Singh and Zheng Bu. Hot Knives Through Butter: Evading File-based
Sandboxes. 2014.
23. Symantec. Advanced Persistent Threats: A Symantec Perspective. 2011.
24. Colin Tankard. Advanced Persistent Threats and how to monitor and deter them.
Network security, 2011(8):16–19, 2011.
25. Gordon Thomson. APTs: a poorly understood challenge. Network Security,
2011(11):9–11, 2011.
26. Olivier Thonnard et al. Industrial espionage and targeted attacks: Understanding
the characteristics of an escalating threat. In Proceedings of the 15th Symposium
on Research in Attacks, Intrusions, and Defenses, pages 64–85. Springer, 2012.
27. TrendLabs. Spear-Phishing Email: Most Favored APT Attack Bait. 2012.
28. Nart Villeneuve and James T. Bennett. XtremeRAT: Nuisance or Threat? 2014.
29. Nart Villeneuve et al. Operation Ke3chang: Targeted Attacks Against Ministries
of Foreign Affairs. 2013.