HP thinks two can play at that game. The company has released its Active Countermeasures technology to a limited beta audience. Its an innovative network scanning tool that looks for systems on the network that "are unmapped or do not comply with security policy, and therefore represent vulnerable points in the network."
When it finds these systems, it "automatically deploys policy-driven mitigation techniques." It appears that the scanner actually exploits the vulnerabilities in order to gain control and deploy the mitigation techniques.
Taking the biological metaphor for all its worth, HP says this is part of a "corporate immune system" that includes other innovative techniques such as a mail server that implements the companys "Virus Throttler," which sets rate-limiting on mail connections to limit the damage that mail worms can do.
Joe Pato, a distinguished technologist at HP Labs, spoke about this technology at the RSA conference in San Francisco earlier this year, where he likened the technique to vaccination, in which the patient receives a less virulent form of the infection.
So, its a network vulnerability scanner with a difference. One might expect Active Countermeasures to be more effective against rogue systems on the network than a conventional scanner, but to what degree? If a system is not supposed to be there, do you really want to patch it and install your anti-virus client, or do you want to block it off the network somehow and alert the administrator?
Its not hard to imagine many problems resulting from aggressive use of this technology, although not everyone would call all of them problems. For instance, the guest or consultant who connects to the network without going through all of the proper channels first—and ends up getting his system "infected."
Im also curious as to how hard it would be to write a custom honeypot to bait Active Countermeasures. Nothings perfect, I guess. Im sure Active Countermeasures is designed to address the common cases, not every theoretical crack of itself.
Because of some poorly worded articles, there was some talk about Active Countermeasures being an actual "good worm," but this was misplaced. Worms spread, as I understand them, and Active Countermeasures simply uses wormlike attacks against unsecured systems, but it doesnt spread any further than that.
Still, one could wonder whether its a good idea to exploit vulnerabilities for good purposes. Does the end justify the means? There are several recent examples of worms written to do "good" things, most famously Welchia, which used the RPC/DCOM vulnerability and then tried to patch it.
Go ask the corporate administrators that HP wants to sell to if theyd like you to spread Welchia on their networks, but make sure youre wearing a cup first.
Of course, Welchia is a worm and Active Countermeasures isnt, but its going to take a whole lot of testing with detailed record-keeping on other peoples networks before I trust this technology on mine. Is it really better than other scanning and remediation techniques? Lets exhaust the honest means before we start enshrining attack code as an active security tool.
Security Center Editor Larry Seltzer has worked in and written about the computer industry since 1983.
More from Larry Seltzer