Is APT27 Abusing COVID-19 To Attack People ?!

Security researcher Marco Ramilli analyzed a new Coronavirus (COVID-19)-themed attack gathering evidence of the alleged involvement of an APT group.

Scenario

We are living hard time, many countries all around the world are hit by COVID-19 which happened to be a very dangerous disease. Unfortunately many deaths, thousands of infected people, few breathing equipment, stock burned Billion of dollars and a lot of companies are entering into a economic and financial crisis. Governments are doing their best to mitigate such a virus while people are stuck home working remotely using their own equipment.

In that scenario, jackals are luring people using every dirty way to attack their private devices. At home it’s hard to have advanced protection systems as we have in companies. For example it’s hard to have Intrusion Prevention Systems, proxies, advanced threat protection, automated sandbox and again advanced end-point protections letting personal devices more vulnerable to be attacked. In this reality ruthless attackers abuse of this situation to attack digitally unprotected people.

Today, many reports are describing how infamous attackers are abusing such an emergency time to lure people by sending thematic email campaigns or by using thematic IM within Malware or Phishing links. Following few of them that I believe would be a nice reading:

• New Cyber Attack Campaign Leverages the COVID-19 Infodemic.
• Hackers taking advantage of COVID-19 to spread malware
• New COVID-19-themed malspam campaign delivers FormBook Malware

Today I want to contribute to such a blog-roll analyzing a new spreading variant that hit my observatory. I want to “spoil” the conclusions now, but it’s getting pretty sad if an APT group makes use of its knowledge to take advance from COVID-19 outbreak.

Stage 1

The first stage is a fake PDF file. It looks like a real PDF, it has a hidden extension and a nice PDF icon, but it really isn’t a PDF, it’s actually a .lnk file, or in other words a “Microsoft Linking File”.

Opening up the .lnk file we might appreciate a weird linking pattern. Two main sections: one is a kind of header where it is possible to observe commands, and the other section is a big encoded payload.

Once beautified the first section it looks easier to understand what it does. It basically copies itself into a temporary folder (through cmd.exe), it extracts bytes from its body (from section two), it decodes such a bytes from Byte64 (through msoia.exe ) and it places the extracted content into the temporary user folder. It deflates the content (through expand) and it finally it executes a javascript file (through wscript) which was included into the compressed content. The following image shows the beautified code section of the analyzed file.

It is quite nice to see how the attacker copied certutils from local system, by using (ertu.exe) in order to avoid command line detection from public sandboxes. Indeed many sandboxes have signatures on certutils, since it’s quite a notorious tool used by some attackers, so that avoiding the behavior signature match it would take a lower score from public sandboxes.

Stage 2

Stage 1 carved Stage 2 from its body by extracting bytes and decoding them using base64 encoding. The new stage is a Microsoft compressed CAB file described in the following table.

Extracting files from Microsoft CAB we observe 6 more files entering in the battlefield:

• 20200308-sitrep-48-covid-19.pdf.
• The original PDF from WHO explaining the COVID-19 status and how to fight it.
• 3UDBUTNY7YstRc.tmp.
• PE32 Executable file (DLL) 486AULMsOPmf6W.tmp.
• PE32 Executable (GUI) 9sOXN6Ltf0afe7.js.
• Javascript file (called by .lnk) cSi1r0uywDNvDu.tmp.
• XSL StyleSheet Document MiZl5xsDRylf0W.tmp.
• Text file including PE32 file

Stage 1 executes the Javascript included in the CAB file. 9sOXN6Ltf0afe7.js performs an ActiveXObject call to WScript.Shell in order to execute Windows command lists. Once” deobfuscated” and beautified the command line looks like the following (9sOXN6Ltf0afe7.js payload beautified) . The attacker creates a folder that looks like a “file” by calling it cscript.exe trying to cheat the analyst. Then the attacker populates that folder with the needed files to follow the infection chain.

A special thought goes to WINRM.VBS which helped the attacker to manage remote systems (lateral movements). According to Microsoft: “WINRM is the CLI interface to our WS-MGMT protocol. The neat thing about this is that you can call it from PowerShell to manage remote systems that don’t have PowerShell installed on them (including Server Core systems and Raw hardware).” The attacker also places a file called Wordcnvpxy.exe on the OFFICE12 folder. We will analyze it in a few steps but at that stage we might observe that is the “last call” before luring the victim by showing the good PDF file (also included in the CAB). But according with 9sOXN6Ltf0afe7.js the first run is on WsmPty.xsl which is the renamed version of cSi1r0uywDNvDu.tmp.

Stage 3

Stage 3 is run by stage 2 and it is a XSL (StleSheet Office file) wrapping a VBScript object.

The following VBScript is run through cscript.exe, It’s an obfuscated and xor-encrypted payload. The encryption is performed by a simple xor having as key the single byte 0 while the encoding procedure is a multi conversion routine which could be summarized as follows:

chr(asc(chr(“&h”&mid(x,y,2))))

The attacker tried to confuse the analyst by reusing variable names in private or local contexts, but after a couple of minutes, you might eventually come out with the following decryption loop.

If you run it against the embedded payload you will eventually see a new stage: Stage 4. A brand new script targeting old version of MSOffice.

Stage 4

Stage 4 is decoded and run by Stage 3. That stage runs an attacker version of MSOSTYLE.exe copied from Stage 2. It hijacks method on an old office 2007 component (Office Data Provider for – MSOSTYLE.exe).

Stage 5

Stage 5 is activated by Stage 2 but only after the execution of Stage 3 and Stage 4. Stage 5 is a multi-step session composed by the following additional artifacts: (i) 3UDBUTNY7YstRc.tmp renamed by Stage 2 into OINFO12.OCX and (ii) MiZl5xsDRylf0W.tmp renamed by Stage 2 into Wordcnvpxy.exe . Every single artifact is available after the execution of Stage 2 into the crafted folder called: OFFICE12 from the user home.

MSOSTYLE.EXE is an old Microsoft Office Data Provider for WBEM. Web-Based Enterprise Management (WBEM) comprises a set of systems-management technologies developed to unify the management of distributed computing environments. So it could not be considered malicious, but it could be considered accountable of the entire infection chain.

At the time of writing only three AVs detect OINFO12.OCX as a malicious file. Rising AV is actually the only company which attributes it to a well-known PlugX sample. According with Trend Micro, the PlugX malware family is well known to researchers having samples dating back to as early as 2008. PlugX is a fully featured Remote Access Tool/Trojan (RAT) with capabilities such as file upload, download, and modification, keystroke logging, webcam control, and access to a remote cmd.exe shell.

Taking it on static analysis it will expose three callable functions: DeleteOfficeData (0x10001020), GetOfficeData (0x10001000) and EntryPoint 0x100015ac).

Both of the methods DeleteOfficeData and GetOfficeData looks like recalling a classic method to hijacking old Office Parser (take a look to here and figure 3 in here ) to execute commands.

Indeed if run from its Entry Point, the DLL executes Wordcnvpxy.exe (as it is the default plugin component). The executable DLL must be in the same path of Wordcnvpxy.exe and it needs to have such a filename (imposed by Stage 2 and hardcoded into the library). On the other side of the coin if commands are passed through stdin, it executes the given parameters as commands.

The following image shows when parameters are given and Commands are executed.

Finally we have Wordcnvpxy.exe which is run in the same stage (Stage 5) by OINFO12.OCX . At the time of writing, it is well-known from static engines, it looks like a standard backdoor beacon-ing to own command and control installed as PlugX module.

The sample uses dynamic function loading avoiding static enumeration and guessing. It grabs information on the victim, PC-name, username, IP-location and send them to C2 as a first beacon.

The used Command and Control resolves to the following URL hxxp://motivation[.]neighboring[.]site/01/index.php

Unfortunately the attacker has shut down everything few hours after I started my analysis, so that I do not have more information about network, commands and additional Plugins. However the overall structure reminds me PlugX RAT as nicely described here.

Attribution

According to MITRE (BTW thank you @Arkbird_SOLG for the great suggestions on attribution) PlugX is a well known RAT attributed to China’s APT. APT27 (aka Emissary Panda) are the mostly notable APT group that used it. Moreover (thanks to @Arkbird_SOLG) “[…] on China culture, hijacking method are a mandatory knowledge for a job like pentesting […]” which could confirm the theory of APT27

We all are passing a bad time. COVID-19 caused many death and is threatening entire economies. Please, even if you are an attacker and you gain profit from you infamous job, stop cyber attacks against peoples that are suffering this pandemic and rest. Ethics and compassion should be alive – even behind you monitors.

Additional details, including Indicators of Compromise (IoCs) and Yara Rules, are included in the analysis published by Marco Ramilli on its blog:

About the author: Marco Ramilli, Founder of Yoroi

I am a computer security scientist with an intensive hacking background. I do have a MD in computer engineering and a PhD on computer security from University of Bologna. During my PhD program I worked for US Government (@ National Institute of Standards and Technology, Security Division) where I did intensive researches in Malware evasion techniques and penetration testing of electronic voting systems.

I do have experience on security testing since I have been performing penetration testing on several US electronic voting systems. I’ve also been encharged of testing uVote voting system from the Italian Minister of homeland security. I met Palantir Technologies where I was introduced to the Intelligence Ecosystem. I decided to amplify my cyber security experiences by diving into SCADA security issues with some of the most biggest industrial aglomerates in Italy. I finally decided to found Yoroi: an innovative Managed Cyber Security Service Provider developing some of the most amazing cyber security defence center I’ve ever experienced ! Now I technically lead Yoroi defending our customers strongly believing in: Defence Belongs To Humans

Pierluigi Paganini

(SecurityAffairs – COVID19, COVID19)

[adrotate banner=”5″]

[adrotate banner=”13″]

Share On