Malware & Threats

Clever Techniques Help Malware Evade AV Engines

FireEye researchers have identified several pieces of malware that managed to go undetected by antivirus engines for extended periods of time by using some interesting techniques.

<p><strong><span><span>FireEye researchers have identified several pieces of malware that managed to go undetected by antivirus engines for extended periods of time by using some interesting techniques.</span></span></strong></p>

FireEye researchers have identified several pieces of malware that managed to go undetected by antivirus engines for extended periods of time by using some interesting techniques.

In the first part of a new series called “Ghosts in the Endpoint,” the security firm detailed the techniques used by several pieces of malware — including families possibly used by APTs and ones that could not be attributed — to remain largely undetected by AV engines.

Researchers analyzed malicious Win32 binaries and documents (Microsoft Office, RTF and Hangul Word Processor) that had been uploaded to the VirusTotal scanner in 2015 and remained undetected until at least January 2016. The study also includes samples detected by a small number of AV engines but which leveraged interesting bypass techniques.

One of the threats, which FireEye believes has been used by an APT group targeting Taiwan, had a 0/53 detection rate on VirusTotal for at least 6 months. The malware, a backdoor dubbed by experts “GOODTIMES,” was disguised as an Excel file (XLSX) and leveraged a Flash Player exploit leaked after the Hacking Team breach last year.

Researchers believe the threat might have remained undetected by AV engines because instead of being hosted on a web page, the Flash exploit was embedded as an ActiveX object directly in the Excel file.

Another threat, which FireEye believes may have been used by the Chinese group known as APT3, is a variant of a backdoor named UPS. This piece of malware also managed to remain undetected for at least 6 months, possibly because the sample includes a significant amount of junk code that could mask its malicious nature and make analysis efforts more difficult.

A piece of malware that includes a VBA macro and a Metasploit shellcode loader backdoor had only been detected by one AV engine when it was analyzed by FireEye in January. The threat, uploaded to VirusTotal in September 2015, has been possibly used by an APT actor based in the Middle East. Evidence points to the Iran-linked cyber espionage group dubbed Rocket Kitten.

The threat might have bypassed signature-based detection thanks to a byte concatenation technique used inside the VBA macro.

Advertisement. Scroll to continue reading.

The last APT-related malware that went largely undetected for at least six months was delivered via a Hangul Word Processor document and it might have been used by malicious actors to target South Korea. Researchers believe the malware could have evaded detection through a change in heap spraying technique, and by triggering the vulnerability it exploited in a different format.

FireEye also discovered several pieces of malware that it could not attribute to any particular threat actor, including a backdoor dubbed “OccultAgent,” a RAT used in attacks in Brazil, and a downloader that went undetected for more than a year.

The list of techniques used by these threats to evade detection includes the use of multiple scripting languages, multilayer packing, multistage infections, and various techniques for loading malicious content from Office documents.

“For proper detection, it is essential to monitor an attack through its entire life cycle – not simply when a suspicious document or file first enters a network. This approach is necessary to detect and block multi-stage infection strategies. While initial events (such as the delivery of a macro-enabled spreadsheet) may appear innocuous, eventually a later stage of the attack will trigger detection,” FireEye said in a blog post.

“It is much easier to stop an attack – including a multi-stage attack – when it first occurs, to include detecting known and unknown exploits (zero days), or even threats that require user interaction such as macros inside documents,” the company added.

Related Content

Copyright © 2024 SecurityWeek ®, a Wired Business Media Publication. All Rights Reserved.

Exit mobile version