Trojan 2.0 Crafted Using Web 2.0 Technology

Web 2.0, meet Trojan 2.0.

In its Q4 Web Security Trends Report, the Finjan Malicious Code Research Center warns that traditional methods of defense against crimeware Trojans -- signature-based detection and blocking command-and-control channels -- will be increasingly ineffective as crimeware creators take advantage of Web 2.0 technology.

Trojans, in the context of computers, are simply programs that appear to be benign but are in fact malicious. When run, they install programs or execute scripts that can steal or damage data on the host system. Trojan keyloggers, which record the keystrokes of users on a compromised system and send those keystrokes to a data thief, are one popular manifestation of this kind of software. Typically, Trojan programs can be controlled by the remote attacker.

While Trojan 2.0 may sound like the latest in a long line of terms that have attempted to piggyback on the market recognition of Web 2.0, Finjan's decision to refer to the emerging generation of Trojans thus has some merit because Trojan 2.0 code exploits Web 2.0 systems and software.

As Finjan explains in its report, blocking the Trojan command-and-control structure becomes more difficult when that communication happens over open channels. The commands issued to Trojans can easily be converted into an RSS feed and passed to a free Web-based RSS reader, such as Google Reader or My Yahoo. "This is the first step in 'legitimizing' the control communications that the Trojans would use," the report states. "By generating the traffic through a trusted third-party Web service, the Trojan has elegantly avoided shutdown by Web security enforcement entities."

The command center could thus be almost any RSS-enabled blog, the report states. And taking that one blog down would do nothing because the Trojan could be directed to pay attention to a different command-and-control feed.

Web 2.0 sites such as Blogger, MySpace, and Facebook also provide an easily accessible place to store stolen data until it can be collected and deleted.

For businesses scrambling to prevent data breaches, this presents a problem. "Since this model uses legitimate Web sites and domains for distributing instructions to botnets, these communications appear as regular Web traffic and in most cases cannot be detected by enterprises' existing security solutions," Finjan's report says.

Finjan concludes that real-time data inspection is necessary to defend against the Trojan 2.0 threat. And while that may be a self-serving argument for the company to make, given that it traffics in that very type of technology, it's also something that many others in the security industry have been saying for some time.

Signature-based blocking alone won't keep networks Trojan-free in an era when every Trojan may have its own signature. And port blocking won't help when your data is leaving through ports you have to leave open.