New Sofacy Attacks Against US Government Agency

The Sofacy group, also known as APT28, is a well-known threat group that frequently conducts cyber espionage campaigns. Recently, Unit 42 identified a spear phishing e-mail from the Sofacy group that targeted the United States government. The e-mail was sent from a potentially compromised account belonging to the Ministry of Foreign Affairs of another government entity and carried the Carberp variant of the Sofacy Trojan. The developer implemented a clever persistence mechanism in the Trojan, one which had not been observed in previous attacks. The focus of this blog will be on the attacks and the infrastructure associated with Sofacy using the new persistence mechanism as a correlation point.

The Delivery

On May 28, 2016, attackers sent a spear-phishing e-mail to a U.S. government entity using an email address belonging to the Ministry of Foreign Affairs of another country. Analysis of the attack revealed a high likelihood that the sender’s email address was not spoofed and is instead a result of a compromised host or account belonging to that Ministry.

The targeted email had a subject of “FW: Exercise Noble Partner 2016”, which is a reference to a joint NATO training effort between the United States and Georgia. The email contained an RTF file as an attachment, with the filename “Exercise_Noble_Partner_16.rtf,” reflecting the same training exercise. We have also seen related delivery documents with filenames that have a Russian military theme (Putin_Is_Being_Pushed_to_Prepare_for_War.rtf and Russian anti-Nato troops.rtf), purportedly targeting organizations in Poland according to a blog published byPrevenity.

The RTF file is a weaponized document that attempts to exploit CVE-2015-1641 to drop two files to the system, specifically, “btecache.dll” and “svchost.dll”. The “btecache.dll” file is a Trojan that loads and executes “svchost.dll”, which is a Carberp variant the Sofacy Trojan. Surprisingly, unlike many other espionage actors who display decoy documents after successful exploitation, this RTF document does not drop or open a decoy document after exploiting the vulnerability.

In the installation process, we observed the delivery document creating a very interesting registry key that it uses for persistence to run the Trojan. The path to the “btecache.dll” file is added to the following registry key:

Software\Microsoft\Office test\Special\Perf\: “C:\Users\[username]\AppData\Roaming\btecache.dll”

This registry key is interesting, because unlike traditional methods of maintaining persistence, it does not automatically run the “btecache.dll” file at system start up. Instead, this registry key will cause the DLL to load only when the user opens any Microsoft Office application, such as Word or Excel. This is the first time Unit 42 has seen the Sofacy group, or any other threat group for that matter, use this tactic for persistence purposes. An added benefit for the threat actor to using this specific tactic for persistence is that it requires user interaction to load and execute the malicious payload, which can cause challenges for detection in automated sandboxes.

The Carberp variant of Sofacy

The “btecache.dll” file is the loader Trojan that is responsible for loading the “svchost.dll” DLL and executing it. Both the “btecache.dll” and “svchost.dll” files contain code from the leaked Carberp source code, specifically the API resolution functions, as well as the RC2 key. The Sofacy group has used the Carberp source code in the past, specifically discussed in a blog by F-Secure, which is the reason we call this Trojan the Carberp variant.

The “svchost.dll” file contains the bulk of the functionality of this Trojan, which at a high level is a downloader that allows the threat actors to gain an initial foothold on the system. The Trojan sends network beacons to its command and control (C2) serverallowing the threat actors to identify targets of interest. The threat actors can then respond to these network beacons to download and execute additional secondary payloads on the system.

The Trojan delivered in this attack contains two network locations that it will send network beacons to, specifically “google.com” and “191.101.31.6”. These beacons are sent to the legitimate website google.com as an attempt to hide the true C2 beacons sent to the actual C2 server hosted at 191.101.31.6.  The network beacons are sent using HTTP POST requests with URLs created largely with random characters. There are two exceptions where random characters are not used to construct the URL, specifically the file extension that is randomly chosen from .xml, .pdf, .htm or .zip and the base64 encoded value at the end of the URL. The base64 encoded data is a string (“J04aLsxVhHBkr19CYr0”) hardcoded within the Trojan that it will then encrypt using a custom algorithm. Figure 1 shows an example beacon sent from the Trojan to the C2 server during analysis.

Figure 1 Network Beacon Sent from Carberp variant of Sofacy

The POST data seen in the beacon in Figure 1 is base64 encoded and encrypted using the same custom algorithm used to encrypt the data in the beacon URL. We decrypted the data to determine its purpose and found the cleartext seen in Figure 2.

,^Bid=I,;<&w@[System Process]
System
smss.exe
csrss.exe
wininit.exe
csrss.exe
winlogon.exe
services.exe
lsass.exe
lsm.exe
svchost.exe
svchost.exe
svchost.exe
svchost.exe
svchost.exe
svchost.exe
svchost.exe
spoolsv.exe
svchost.exe
taskhost.exe
userinit.exe
dwm.exe
explorer.exe
svchost.exe
cmd.exe
conhost.exe
reader_sl.exe
svchost.exe
cmd.exe
conhost.exe
SearchIndexer.exe
SearchProtocolHost.exe
SearchFilterHost.exe
SearchProtocolHost.exe
explorer.exe
svchost.exe
svchost.exe
disk=IDE\DiskMAXTOR_HARDDISK_________________________2.2.1___\5&2770a7af&0&0.0.0
build=0x7caa0e19

Figure 2 Decrypted HTTP POST Data Shows System Information

The clear text of the data sent in the network beacons contains information regarding the compromised system, as well as malware-specific information. The data is comprised of the following fields of data:

id = The serial number of the storage device

w = This parameter (whose name ‘w’ could change to any character between samples) begins with a one byte value denoting the OS version followed by a one byte value for the CPU architecture. These values are immediately followed by a new line delimited list of running processes on the system.

disk = The name of the system’s hard drive, obtained from the registry key “SYSTEM\CurrentControlSet\Services\Disk\Enum\0”

build = The hardcoded build identifier for the Trojan version

inject = (Optional, not displayed in Figure 2) If the Trojan injected its code into other processes to interact with the C2 server

This callback data allows the threat actors to determine if the infected machine is a target of interest, as the beacon contains a list of running processes and the name of the storage device that could be used to filter out analysis systems or researchers. If the actors believe the system is of interest, they will respond to these network beacons to download and execute additional secondary payloads on the system. The Trojan parses the response to the beacons for two actions “Execute” and “Delete” between the tags “[file]” and “[/file]”, as well as settings labeled “FileName”, “PathToSave”, “Rundll” and “IP” between the tags “[settings]” and “[/settings]”. This allows the threat actors to download additional files to the system, execute both executables and DLLs and delete files.

The Infrastructure

The initial analyzed sample in this attack only contained a single malicious command and control location, 191.101.31.6. We have not observed this IP address used by the Sofacy group in any previous attack campaigns, and examining passive DNS data showed no other correlations to potentially related attacks. The sample also seen by Prevenity appeared to only have a single primary C2 domain, servicecdp[.]com. This domain also appears to be newly created for this specific attack campaign, with no strong links to any previous attacks.

Pivoting off the unique registry key used for persistence revealed links to a previously observed Sofacy campaign, from mid-2015. Two additional payloads with recent compile dates of March 7, 2016, were discovered using the same persistence mechanism, and analysis of those payloads revealed one primary C2 domain, munimonoce[.]com, and three secondary C2 domains, http://www.wscapi[.]com, http://www.tabsync[.]net, and storsvc[.]org. The secondary C2 domains may appear familiar, as they were widely publicized in a report from iSight Partners in July 2015 as C2 domains related to the Sofacy group aka Tsar Team.

In addition, the primary C2 domain munimonoce[.]com previously had resolved to the IP 66.172.11.207, which was previously identified as a primary C2 IP for a Sofacy payload with a compile timestamp of June 11, 2015. This particular sample also happened to use the exact same secondary C2 domains of www.wscapi[.]com, www.tabsync[.]net, and storsvc[.]org, but lacked the newly discovered persistence mechanism.

The Sofacy group often re-uses infrastructure components across multiple attack campaigns, whether to speed the flow of attacks, for a lack of available resources committed, or out of sheer laziness. In this case, the newer attack campaign appears to use newly created infrastructure, but still maintains some overlap with previous Sofacy-related C2s. We believe this overlap could possibly be due to an oversight when adapting a previous code base with the new persistence method discussed in this blog for the new attack campaign.

The threat appears to be moving toward deployment of one-off infrastructure that can make analysis of attack campaigns and correlation more challenging. This shift stresses the importance of analysts and researchers being able to pivot on all artifacts of a given attack, not simply relying on network indicators. In this case, we were able use AutoFocus to pivot on a common registry key unique to this attack campaign to quickly identify where it correlates with characteristics of previous attacks.

Conclusion

The Sofacy group continues its attack campaigns on government organizations, specifically the U.S. government in this latest spear-phishing example. The threat group added a new persistence mechanism that requires user interaction by loading its payload into Microsoft Office applications when opened, which may help the actors to evade detection. The use of this new persistence method shows the continued development of tactics and techniques employed by this threat group, often times in clever ways as we observed in this instance.

Palo Alto Networks customers are protected from the new Sofacy Carberp variant and can gather additional information using the following tools:

  • WildFire detection of all known samples as malicious
  • All known C2s are classified as malicious in PAN-DB
  • AutoFocus tags have been created SofacyCarberp

Indicators

Delivery Documents

03cb76bdc619fac422d2b954adfa511e7ecabc106adce804b1834581b5913bca (Exercise_Noble_Partner_16.rtf)
12572c2fc2b0298ffd4305ca532317dc8b97ddfd0a05671066fe594997ec38f5 (Putin_Is_Being_Pushed_to_Prepare_for_War.rtf and Russian anti-Nato troops.rtf)

Loader Trojans

c2551c4e6521ac72982cb952503a2e6f016356e02ee31dea36c713141d4f3785 (btecache.dll)
be1cfa10fcf2668ae01b98579b345ebe87dab77b6b1581c368d1aba9fd2f10a0 (bitsprex3.dll)
fbd5c2cf1c1f17402cc313fe3266b097a46e08f48b971570ef4667fbfd6b7301 (amdcache.dll)

Payloads

69940a20ab9abb31a03fcefe6de92a16ed474bbdff3288498851afc12a834261 (svchost.dll)
aeeab3272a2ed2157ebf67f74c00fafc787a2b9bbaa17a03be1e23d4cb273632 (clconfg.dll)
dfa8a85e26c07a348a854130c652dcc6d29b203ee230ce0603c83d9f11bbcacc (iprpp.dll)
57d230ddaf92e2d0504e5bb12abf52062114fb8980c5ecc413116b1d6ffedf1b (clconfg.dll)

Command and Control

191.101.31.6
munimonoce[.]com
wscapi[.]com
tabsync[.]net
storsvc[.]org
servicecdp[.]com

and

[Palo Alto Networks Research Center]

Creating Her Own Cybersecurity Path: From JD to FBI to CISSP

My career in cybersecurity was not mapped out. I earned my bachelor of science degree in business administration, with a concentration in information systems from Xavier University, but one thing was for sure: I did not want to be a coder. I found that as I pursued my education, different opportunities in a variety of fields opened themselves up to me. Next, I decided to pursue my law degree – I earned my J.D. from Texas Southern University – and while I was in law school, the FBI came in to recruit. I was intrigued; so after practicing law for a short period of time in Washington, D.C., I began working as a special agent for the FBI, and in addition, served as the information systems security officer for the Los Angeles Field Division.

Upon leaving the FBI, I looked around at the skills I had obtained – from IT to legal to investigating cybercrime – and believed that a career in information security was the right fit for me. I thought that obtaining the Certified Information Systems Security Professional (CISSP®) certification would help me validate what I knew based on my experience, as well as add credibility to me as a professional when I transitioned from government service into the private sector.

Having my CISSP has helped open doors for me as it is looked upon as one of the highest industry certifications you can obtain. It validates that an experienced cybersecurity professional is working from a standard of common knowledge. The CISSP also helps to set the foundation for consistency in the industry.

Along with having my certification, one of the best things about being a member of (ISC)² is the special programs they provide. Early on after obtaining my CISSP, I signed up and delivered several Safe and Secure Online® presentations in my community. I had the opportunity to speak with school-age children, Girl Scout organizations and parents about the challenges faced by children online. Being able to speak to children about cyber dangers, as well as their parents, regarding what to look out for, was one of the most rewarding experiences in my career.

Luckily, I followed my mother’s advice and got an education. Because of that, I’ve had the pleasure of experiencing so many wonderful career highlights: Being a licensed attorney, an FBI special agent, a business owner, a college instructor and now an executive – these achievements make me take a deep exhale when I think about it. I am one of those people who always has some type of goal for myself – both professionally and personally. I believe that if you are not striving for success in life, you miss out on the opportunity to be the best version of yourself. Of course, I have specific goals to become a chief information security officer, hold board membership positions and more; but essentially, my goals all come down to using my knowledge and experience to help move the needle forward in cybersecurity.

I was once told that when you feel passionate about something, you have to get involved to effectuate change. As an African American woman, I believe that it is part of my responsibility, as I obtain success, to be visible and to inspire others. This is what led to my involvement with the International Consortium of Minority Cybersecurity Professionals (ICMCP). As a single mother and an executive, I have limited time. However, I believe in the mission of the ICMCP – to find ways to bring more women and minorities into the field of cybersecurity. The organization also provides support to those already in the field, because historically, women and minorities tend to leave cybersecurity careers early. We have a severe shortage of talent in cybersecurity for various reasons. Part of the ICMCP’s mission is to attract talent from a multitude of areas – both from traditional and nontraditional disciplines – so that we can bring more talent into the pipeline.

Personally, I believe that we need diverse talent. This isn’t limited to sex, race or cultural diversity, but diversity in thoughts, actions, problem-solving skills and leadership abilities as well. This will take us much further in our “secure the world” directive.

Mary Chaney, Esq., CISSP is the director of worldwide information security at Johnson & Johnson and co-chair of the fundraising committee for the International Consortium of Minority Cybersecurity Professionals (ICMCP).

(ISC)² Management

[(ISC)² Blog]

Conventional AV Systems Can Actually Harm You

There’s barely a day goes by when I’m not reading another batch of stories about how unsuitable conventional endpoint AV security is for dealing with modern malware, APTs, zero-day threats and so forth.

So replete are these tales of woe that it’s almost possible to switch off from the basic fact that in this uber-connected, cloud-enabled, everything-as-a-service, internet-of-thingamajigs world, most conventional endpoint AV systems are impotent and probably do more harm than good. I write almost, but not quite, because every now and again the occasional story jumps off the screen and gives you that all important wake-up call.

One such story that came to light a few weeks ago, centered around an Adverse Event Reportpublished by the U.S. Food and Drug Administration (FDA). Wherein, a patient (not named), undergoing a cardiac catheterisation procedure at a US Hospital (also not named) had to be sedated, mid-operation, for five minutes, while the procedure was suspended following the system crash of a vital piece of monitoring equipment.

The system in question monitors, measures and records patient data during cardiac catheterization procedures. The system is made up of a patient data module, used to capture the patient’s vitals, and a hemo monitor PC to display them. The two elements are connected via a serial interface.

During this particular procedure the monitor PC lost communication with the patient data module resulting in a black screen on the monitor and the patient having to be sedated while the system was rebooted. As the FDA report describes, the cause of this blackout was attributed to the installed conventional AV software, which at a critical point in the procedure initiated a scan of the system.

Although the system could be rebooted and the patient fortunately survived it got me thinking about the real-life harm a conventional AV could do to me. Quoting from the Manufacturers Narrative in the FDA Report, “Our experience has shown that improper configuration of anti-virus software can have adverse effects including downtime and clinically unusable performance.” So, although I may be sensationalizing the FDA’s paragraph a little, I’m not feeling that confident after reading the manufacturer’s narrative. Let’s face it: the team performing a standard cardiac catheterisation procedure is not likely to include an IT security engineer who can be called upon at a moment’s notice.

Could this scenario have been avoided with an Advanced Endpoint Protection system? The answer is probably yes. Traps, our advanced endpoint protection product, is not a conventional AV system — indeed, it’s a paradigm shift from “the way things used to be done.” Traps secures endpoints by preventing known and unknown malware and exploits from executing by focusing on blocking the few, core techniques used by attackers rather than application-specific characteristics. Furthermore, It does this in a lightweight, nonintrusive agent that definitely does not rely on system scanning.

Learn more:

[Palo Alto Networks Research Center]

Leaky End Users Star in DBIR 2016

Insider threat once again tops the list of enterprise cyber security threats in the 2016 Verizon Data Breach Investigations Report (DBIR). For the second straight year, Verizon research showed that the average enterprise is less likely to have its data stolen than to have an end user give away sensitive credentials and data—whether unintentionally or maliciously.

From insecure storage, transfer or disposal of sensitive information, to lost or stolen endpoint devices, to intentional data theft and privilege abuse, to simply entering the wrong recipient name in the email address field, the vast majority of breaches can be traced back to end users. “Our findings boil down to one common theme,” said Verizon Enterprise Solutions Executive Director of Global Services Bryan Sartin, “the human element.”

Overall, 2015 trends persist in 2016
The 2016 DBIR pulls trends and insights from more than 100,000 incidents—and 3,141 confirmed data breaches—across 82 countries. Is there anything groundbreaking in this year’s DBIR? Nope. Verizon reports “no drastic shifts” and no “show-stopping talking point.” For the most part, last year’s trends and patterns continued. But to “strike a deceased equine” (as Verizon put it), these persistent trends bear reviewing.

Phishing still works—end users are more likely than ever to click the link
The 2016 DBIR found hackers increasingly targeting devices and people instead of servers and networks, with phishing attacks growing from less than 10 percent of all attacks in 2009 to more than 20 percent in 2015. Why? Because people are more likely than ever to “click the link.” Verizon says 12 percent of people tested will click on a phishing attachment—up from 11 percent in 2014. Also of note: the same study found only three percent of users that receive a phishing email report the attack attempt. The IT department is stuck between a rock and a hard place. More people fall for the scam, and no one gives IT a heads-up.

Privilege abuse is still a top insider threat—with an emerging twist
Traditional privilege abuse involves an internal user stealing or corrupting sensitive data—whether for personal gain or in collusion with an external actor. Verizon noted an emerging twist: external parties with legitimate access credentials (a customer or vendor, for example) colluding with another external actor. Verizon also showed that insider threat detection is extremely difficult in cases of privilege abuse, with most incidents taking months for the enterprise to discover. This year, privilege abuse was the top defined category of cyber security threats, second only to the catchall category of “Miscellaneous Errors.”

Something new: the three-pronged attack
Cybercriminals aren’t just getting smarter—they’re growing more patient. Verizon highlighted what it called the “new three-pronged attack”:

  1. Phishing email lures user to malicious link or attachment.
  2. Clicking the link installs malware that targets a user’s various digital access credentials. Sophisticated malware can even compromise other users’ credentials through this one entry point.
  3. Those credentials are later used in other attacks.

The first challenge here is tracing the subsequent attack back to the initially-targeted user and the original phishing email. The second is figuring out just how deep the attack went—which credentials were compromised and which data may have been exposed or stolen. Playing the “long con” gives cybercriminals a chance to slowly, silently extend the reach of the breach, with users and IT unaware.

Biggest cost: tracking down data during breach recovery
With sophisticated attacks leveraging insider credentials to go deeper and broader, it’s no surprise that the biggest cost of an enterprise data breach comes from the daunting task of forensic analysis. Figuring out what data was compromised, and tracking down copies of the files, puts an enormous strain on IT resources, and accounts for nearly 50 percent of the average total cost of an enterprise data breach.

TL;DR—Breaches are inevitable; data visibility is key
The DBIR is great reading (really—you’re guaranteed a laugh or two), but it’s 85 pages long. Here’s the quick-and-dirty:

  • “No locale, industry or organization is bulletproof.” In other words, breaches are inevitable.
  • Know your biggest threats. Take five minutes to check out the tables on pages 24 and 25, showing incident patterns by industry.
  • “You cannot effectively protect your data if you do not know where it resides.” Breach remediation is crucial. Data visibility is key.

Next, we’ll tackle this last point—why data visibility is essential to effective breach remediation, and how an enterprise can enhance data visibility.

Download The Guide to Modern Endpoint Backup and Data Visibility to learn more about selecting a modern endpoint backup solution in a dangerous world.

Susan Richardson, Manager/Content Strategy, Code42

[Cloud Security Alliance Blog]

Traveling the Road to GRC Maturity

Today’s business environment is fraught with risk. However, to successfully seize growth opportunities in the market, organizations need to effectively position themselves to embrace risk with confidence instead of simply avoiding it.

Technology, economic and market conditions affect organizations on a daily basis. The constantly changing landscape of risk is a leading topic in headlines, industry forums, media outlets and board rooms. We are moving to a world where your risk management approach is not only your defense against known and unknown risks, but a critical component of your competitive advantage.

The next five years will bring even more pressure and greater shifts in the way governance, risk and compliance (GRC) programs need to operate to succeed. Not only are regulations and risk management needs changing, but the technology used to power businesses is radically shifting. The delineation between digital strategies and business strategies has disappeared, and any differentiation between cyber risk and business risk has vanished, as well.

What’s Your Cyber Risk Appetite?
Given that many new business growth strategies rely on technology, organizations must be able to manage cyber risk as part of their risk management strategy. This convergence is leading many organizations to think in terms of cyber risk appetite – what risks can the business take and which are beyond the tolerance of the organization.

GRC programs must create a unified risk culture and a common language across the enterprise in order to understand risk in the context of the organization’s overall objectives. Therefore, effective risk management practices must address cyber risk and business risk in equal measure and provide a consolidated view of risk to executives and practitioners.

These capabilities do not magically appear within an organization overnight. The transition from a compliance-driven approach to a risk-based strategy takes commitment and collaboration across all “lines of defense” within an organization. The road to a mature risk management strategy has many twists and turns. Organizations will face multiple forks in that road and a fair share of distractions and derailments. However, the benefits of staying the course are clearly evident.

Risk management is becoming a core capability that separates the winners from the losers. Organizations that understand and effectively manage risk will prosper, while those that cannot will fail. Success starts with the ability to manage risk in a manner that frees up resources to focus on the company’s long term, strategic objectives. Executives need relevant, up-to-date risk information in order to make the right decisions and pursue the right opportunities.

I am pleased to be participating in a 16 June (11 a.m. CDT) webinar titled Next-Gen GRC: Building a Road to GRC Maturity with fellow risk and compliance professionals to discuss this maturity journey. Every organization has its own challenges but there are some fundamental elements that can make a significant impact in your program’s success. We will be discussing strategies to push risk management across your organization and take your program to the next level.

Patrick Potter, GRC Strategist, RSA

[ISACA Now Blog]

English
Exit mobile version