Spammers employ stripper to crack CAPTCHAs
LATEST NEWS
SUBSCRIBE
Computerworld is New Zealand's only specialised information systems fortnightly. Subscribe now for $100 (23 issues) and save more than 37% off the cover price!
SIGN UP
Hackers are using human beings in semi-real time to translate CAPTCHAs by proxy, says Trend Micro
By Gregg Keizer | Framingham | Friday, 2 November, 2007
Spammers are using a virtual stripper as bait to dupe people into helping criminals crack codes they need to send more spam or boost the rankings of parasitic websites, say security researchers.
A series of photographs shows "Melissa," no relation to the 1999 worm by the same name, with progressively fewer clothes and more skin each time the user correctly enters the characters in an accompanying CAPTCHA (Completely Automatic Public Turing Test to Tell Computers and Humans Apart), the distorted, scrambled codes that most web-mail services use to block bots from registering hundreds or thousands of accounts. Spammers rely on web email accounts because they're disposable; by the time filters have blocked the address, the spammers throw it away and move on to another.
The CAPTCHAs that Melissa feeds to users are, in fact, legitimate codes snatched from Yahoo Mail's signup screens, say analysts at Trend Micro. The hackers, frustrated at their inability to come up with a way to automate account registration, are getting users to do their dirty work.
"They're using human beings in semi-real time to translate CAPTCHAs by proxy," says Paul Ferguson, a network architect at Trend Micro. "You have to give them this, it's clever."
Each time the user correctly decodes the CAPTCHA, a new Melissa photo is revealed, pulled from a hacker-controlled server in Israel, according to Symantec. The plain-text decodes are sent to that same server, where they are presumably banked for future use in generating large numbers of Yahoo Mail accounts.
Fumble-fingered typists are even encouraged by Melissa to try their luck again: "Hmmm, nope, the word you entered is incorrect honey! Lets [sic] try again?" the virtual stripper replies.
Trend Micro says the striptease was part of a Trojan horse called CAPTCHA.a; rival Symantec dubbed it Captchar.a instead. The Trojan horse may be part of a multistage attack, downloaded to a PC that's been compromised by other, more malicious code, or can be encountered as a drive-by web-based exploit.
"This isn't the first time that they've tried to bust CAPTCHAs," says Ferguson, noting past attempts by bot-driven malware to apply optical character-recognition technology to deciphering the squiggles and obscured letters. Nor is it the first time human beings have been put to work decoding CAPTCHAs. "Work-at-home money mule schemes run by criminals have hired people to do this same thing," Ferguson says. "They're told to log on to this web page and type the CAPTCHA. They have a quota."
In some cases, those CAPTCHAs have been used to sidestep bot protection for blog commenting rights; hackers will flood a blog they've created with fraudulent comments to drive up its search-engine ranking, expecting that the higher placement will translate into more traffic and thus more clicks on the ads displayed on the blog page. "Sometimes they use [CAPTCHAs] just to bump up their page [ranking]," Ferguson says.
The Trojan horse can strike PCs running Windows 98, Me, NT, 2000, XP and Server 2003.
MOST POPULAR
Social Media @Computerworld NZ

Computerworld NZ has now reached LinkedIn! Join to expand your networks and meet others interested in information systems.







