Posted by Virus Bulletin on Dec 19, 2007
Virus Bulletin finds 40% of users think governments should write viruses to bug terrorists.
In a survey of visitors to the Virus Bulletin website, 40% of computer users said they thought it was a good idea for governments to write viruses to bug terrorists - however, security professionals disagree.
It is an idea that surfaces every now and again: since modern malware allows criminals to get a very detailed picture of what's happening on a user's computer, why not switch roles and write 'malware' that gives good people, say a government, insight in what bad people, say terrorists, are doing on their computers?
The latest to come up with such a plan was the German government who, as reported by The Register last month, have started hiring coders to write software that can be used to hack into terrorists' computers.
Righard Zwienenberg, Chief Research Officer at Norman, commented on the subject in Virus Bulletin earlier in the year, describing why he thought such 'magic lanterns' were a bad idea. He still feels strongly: "It's a rather stupid and ignorant idea. The security industry is progressing into advanced proactive detection techniques and trojans like this will most likely be found by security software that is acting on the behaviour."
Of course, a government could force vendors to let their trojans bypass security, but the story does not end there, Zwienenberg continues: "The captured information has to go somewhere or has to be picked up. These communications may be blocked by a firewall or logged. If the trojan installs a backdoor so the police can enter the system, this may open up possibilities for others to do the same. And what if the police system is infected and a piece of malicious code piggy-backs on the communication through the backdoor into the suspect's computer?"
Alex Eckelberry, president of security software company Sunbelt, concurs: "In order for law enforcement to truly make such a trojan undetectable, they would need to get the anti-malware vendors on board, which would create a serious problem of ethics for vendors." Eckelberry feels it would be extremely difficult to get vendors to agree to non-detection, pointing out that it would actually be to an anti-malware vendor's advantage to advertise that they detect magic lanterns, as a PR/marketing push. (Indeed in 2002, VB reported that approximately 100% of the audience in Graham Cluley's VB2002 presentation 'e-bugs: should anti-virus products detect them?' felt they would like AV vendors to include detection of such programs.)
And what, Eckelberry wonders, will happen when magic lanterns start to spread in the wild? "This will certainly occur. For example, a suspected criminal could sell his laptop on eBay, and the magic lantern could then spread."