Security researchers observed an Astaroth attack that used only living-off-the-land techniques to run the backdoor directly in memory on Windows machines.
The Microsoft Defender ATP Research Team detected the Astaroth attack after noticing an anomaly from a Windows Defender Antivirus algorithm used for catching fileless campaigns. This led the researchers to an infection chain that relied strictly on living-off-the-land techniques — the use of regular tools already present on the system — to avoid raising any red flags.
The infection chain began with a series of spear phishing emails that contained malicious links to redirect users to a LNK file. When double-clicked, the LNK file executed the Windows Management Instrumentation Command (WMIC) tool with the /Format parameter. The campaign then executed JavaScript code to download payloads using the Bitsadmin tool.
At this point, the attacks used Certutil to decode the payloads and then the Regsvr32 tool to run a series of dynamic link libraries (DLLs). This chain of events eventually led one DLL to load into Userinit and then load Astaroth as its final payload. With Astaroth activated, threat actors could use the backdoor to steal sensitive information and move laterally across the network.
Astaroth Abuses Windows Services and Legitimate Tools
Astaroth attacks often abuse legitimate Windows services. In September 2018, Cofense spotted a campaign that compromised 8,000 machines in one week by exploiting WMIC and Certutil. News of this attack came several months before Cybereason detected a campaign in which threat actors abused legitimate operating system functionality and security-related products to distribute the backdoor to users in Brazil.
Awareness Is the Best Defense Against an Astaroth Attack
Security awareness training to educate employees about phishing campaigns and other digital threats can help defend against an Astaroth attack. Analysts should use this education framework in tandem with a patch management strategy and a deep familiarity with built-in Windows code to spot anomalies that could be indicative of fileless attacks.