Facebook's facial recognition software might not be so harmless. A recent study from Carnegie Mellon finds that the technology can help researchers locate people's social security numbers, just from the information on their Facebook profiles and their photo. When first released, Facebookers noted the inherent creepiness of facial recognition, but now that the tool can double as personal identification software, the technlogy is looking even more iffy. With this increased ability to ID its users, Facebook is slowly morphing from a social network to a possible identification service.
Facebook first released its facial recognition software last winter, sidestepping criticism by slowly rolling out the feature and allowing users to opt out of the service. But this study finds that the technology can be used in harmful ways. By mining Facebook profiles, the researchers could identify at least one personal interest of each subject and the first five digits of a social security number about 30 perecnt of the time. Sure, that's not a huge percentage, but it will only improve, lead researcher Alessandro Acquisti told Network World's Richard Power.
As of today, automated face recognition isn't too accurate, but it keeps improving. If you look at the technological trends in cloud computing, the accuracy of face recognizers, and online self-disclosures, it is hard not to conclude that what we present today as a proof-of-concept in our study; will tomorrow become as common as everyday text-based searches on a search engine.
This isn't the first time Facebook has breached privacy concerns with users' identities. Facebook faces legislation for giving user data to advertisers. The applications transmitted Facebook IDs, which contain identifying information, to Internet tracking companies reported the Wall Street Journal, last October.