Original URL: http://www.theregister.co.uk/2009/10/01/facebook_automated_attacks/

Automated attacks push malware on Facebook

More captcha busting suspected

By Dan Goodin

Posted in Security, 1st October 2009 19:18 GMT

Hackers have figured out how to create computer-generated Facebook profiles and are using them to trick unsuspecting users into installing malware, a security researcher warned Thursday.

The fraudulent profiles display the same picture of a blond-haired, blue-eyed woman, but with slightly different names and birthdates, said Roger Thompson, chief of research at security firm AVG Technologies. Each invites visitors to click on what purports to be a video link that ultimately tries to trick viewers into installing rogue anti-virus software.

AVG's LinkScanner product, which monitors webpages in real time to make sure they're not malicious, has encountered "hundreds" of separate pages. But because AVG only sees a page when one of its subscribers tries to click on one, Thompson suspects the total number of fake profiles is in the thousands.

"There are enough of them that it's probably an indication of an automated attack," Thompson told The Reg. "I just can't see someone creating the same profile time after time after time."

That means the attackers have figured out how to crack the captcha Facebook uses to ensure profiles are created by humans, rather than computer scripts that automate the process so it can be carried out thousands of times.

If Thompson is correct, it's by no means the first time hackers have figured out how to bypass the measure on a high-profile website. Captchas for Google Mail and Microsoft's Windows Live email services have been successfully cracked before. In some cases, scripts that use optical recognition technology are suspected to be at work. In other cases, sweat shops that rely on people to solve the captcha puzzles are likely at play.

In any case, the availability of an unlimited number of fraudulent accounts is extremely valuable to scammers. Web-based email accounts typically get the green light from anti-spam products, and end users have an inherent, if misplaced, trust in social networking profiles.

Thompson's report came the same day that the FBI issued this advisory warning people to be wary of fraud on social networking sites.

Facebook engineers are doing a good job killing the fake profiles, Thompson said. But at time of writing, many were still available, as pages like this one attests. ®