Security firms should consider exploiting coding mistakes in malware to take the security fight straight back to the hackers, a Websense security analyst has suggested in a discussion blog that outlines steps to take over the command and control (C&C) servers related to the Zeus banking Trojan.
A highly effective Trojan, Zeus was designed to steal banking and other types of personally identifiable information (PII). Despite its demonstrated efficacy, however, Websense researcher Abel Toro suggests in the blog that the Zeus code is riddled with weaknesses and that a publicly known vulnerability can be used to impersonate an infected system and upload carefully-crafted documents that to the C&C server.
The Zeus C&C servers expect stolen PII to be encrypted using the RC4 encryption technique and the server's encryption key. However, the analyst noted, RC4 is a 'symmetric cipher' so both the client Trojan and server use the same key – which means that the key must be contained somewhere within the malware itself.
Using the Volatility memory analysis tool, the RC4 key can be extracted and used to encrypt a web-shell payload that can be loaded onto the target C&C server using techniques that trick it into running the payload – which, in turn, can be exploited to gain full access to the Zeus C&C server's control panel “just as the original botnet owner would.”
“While Zeus is regarded as an 'advanced' banking Trojan it is also [subject] to bugs that may allow, in an ironic twist, an attacker with the technical skillset to take over a botnet's C&C server,” Toro says.
This approach could be used to take a more aggressive stance against malware, with suitably skilled security experts able to take on malware authors head on – for example, by proactively probing botnet attack networks or feeding their technical details to track down the suspects.
Such actions would represent an escalation of the cat-and-mouse game that currently characterises the relationship, with unknown consequences. Yet the knowledge that it can be done, the Websense blogger suggests, should at least prompt discussions about how and whether security analysts should consider a more active response to extant and new malware threats.
Hackers “are learning that it's not so easy to write secure code,” Toro says. “Most of us in the business of securing our applications and systems know that bulletproofing software is an extremely expensive and exhaustive undertaking. Malware creators who have to look to their own defences would have to slow down the production of new attacks.”
“Most attackers lack the necessary resources and community peer review to harden their malware, and that provides an opportunity for the security community to advance a conversation about what we should do about it.”