Face Recognition Threatens Privacy

Carnegie Mellon researchers have demonstrated how a photo of a person’s face can uncover private information stored in social networks. The research was presented at the Black Hat Security conference where thousands gathered last week to discuss the latest in hacking and security technologies. The purpose of the research was to show how face-recognition software freely available from Facebook, Apple, Google and others, may provide the next big tool for hackers, marketers, and snoops.

The researchers ran three experiments to illustrate their point. Each experiment utilized face-recognition software, an Internet connection, and social networks to uncover detailed private information about a person given only a photo of the person’s face. In the first experiment, they used profile photos from a popular dating site to track down the person’s profile information on Facebook. In the second experiment, they photographed students walking across campus, and were able to use the photos to find the students on Facebook. In the third experiment, the researchers were able to accurately predict personal interests and even the Social Security numbers of subjects based only on a photo of the subject’s face. They then went on to create a mobile app that takes a photo of a person’s face and gives back personal info about that person gleaned from the Internet. Each experiment illustrates how strangers can access personal information of individuals intended only for friends, in some cases with malicious intent.

Imagine photographing people on the street, and immediately accessing their name, home town, interests, lists of friends and other private information. More-over, imagine someone doing this to you? Now imagine the technology in the hands of hackers, businesses, marketers, the government, law enforcement agencies, and snoops. The technology will force us to “reconsider our notions of privacy,” the researchers say.

Leave a Reply

Your email address will not be published. Required fields are marked *