Trojan Tied to RSA SecureID Breach Linked to Multiple Cyber-Campaigns

Dell SecureWorks has connected malware used in the RSA SecureID breach to attack campaigns going back several years that have targeted companies in countries around the globe.

Using a Remote Access Trojan (RAT) known as “Comfoo,” the attackers have been traced to at least 64 campaigns, according to Joe Stewart, director of malware research at Dell SecureWorks, who noted that development of the Trojan goes back as far as 2006.

Some of the victims were government organizations, media companies and corporations specializing in audio and video conferencing technology. Other victims included the networks of mineral exploration, electronic and telecommunication companies.

The largest numbers of victims were found in Japan, India and South Korea, though other targets were found in the United States, Taiwan and Europe. Once Comfoo was on the victims’ systems, it was able to gather system and network information as well as log keystrokes, upload files and take screenshots.

To stay persistent on the system’s it infects, Comfoo replaces the path to the DLL of an existing unused service as opposed to installing a new service that might be noticed during system audits. In addition, the malware starts the existing “ipnat” system service, which causes remote inbound connections to the infected system to fail, thereby blocking remote maintenance by network administrators. In some instances, Comfoo is also delivered with a rootkit that hides its files on disk.

“The malware slipped under the radar because the attackers choose to tweak it to ensure the target’s antivirus won’t detect it before they deploy it,” Stewart said. “The rest of the time when it is detected, targets may not even realize it’s anything other than ordinary commodity malware, because AV doesn’t really tell you that information. Finally, those that do recognize the threat for what it is don’t often share the indicators because they don’t want to publicize a breach.”

In a blog post, Dell SecureWorks’ researchers explain that they also found a “rendezvous-type traffic relay program” that passes traffic between Comfoo victims and the Comfoo master console operated by the attackers.

“Unlike ‘dumb’ traffic relay servers such as HTran, the Comfoo relay server does not know the location of the master console,” according to the blog post by Stewart and senior security researcher Don Jackson. “Instead, the master console program connects to the relay server on-demand, and any incoming victim data is passed to the master console connection.”

“The unauthenticated nature of the Comfoo relay server’s administrative connections makes it possible to take control of the C2 [command and control] server and all victims’ systems, armed only with knowledge of the protocol, the encryption method and the static encryption key hard-coded into every Comfoo binary,” according to the blog.

“Researchers can passively monitor victims’ logins to the relay servers (sending no commands) by connecting to the correct port on the correct IP address at the right time. This technique is analogous to viewing Web server log data stored in a publicly accessible directory on a C2 server.”

To help identify and notify victims of Comfoo-related espionage, the researchers set up a passive monitoring system for dozens of active Comfoo C2 relays that has been running since January 2012. To date, Dell SecureWorks Counter Threat Unit has identified more than 200 variants of the malware developed by the Comfoo gang.

“We are going public now because the activity of this one RAT has fallen off significantly, so tipping the attackers off about how we were able to monitor it for the last 18 months is now a non-issue,” Stewart said.