Stanford Researchers Crack “Captcha” Test

Oh, Captcha.  The bane of my online-buying existance.  Is that an “E” or a “C”?  Lower case “B” or lower case “H”?  Hopefully this recent development will force developers to think of a new, slightly less annoying way to protect our identities online.

Their bot is 25% effective in breaking the security used on Wikipedia, 43% on eBay and up to 66% on the Visa site belongs to authorize.net. The program worked pretty good on Digg and pages belonging to CNN. Most of the commonly used methods of protection are failing to the new tool created in Stanford University…

The program removes background noise from the image and then splits the text into individual letters, which are easier to identify than a whole words. After the publication of the document, Visa and Digg started using reCAPTCHA on their websites.

via Software in Action


Geeks are Sexy needs YOUR help. Learn more about how YOU can support us here.