A study into the face recognition tools used in social media websites has highlighted the alarming ease with which private information can be exploited.
While the world has been gleefully uploading out-of-focus holiday snaps onto Facebook, there are growing concerns about the use of automated tools to identify those in the pictures.
Back in June, Facebook announced that it would be using the technology to ‘tag’ account holders automatically.
It’s a default feature, which means you’ll have to go out of your way to turn the functionality off.
While new kid Google+ appears to offer a more customisable approach to information sharing, a study proves that there is still cause for concern.
A team of researchers at Carnegie Mellon University managed to identify subjects and work out their social security number, armed with just a web cam and Google’s facial recognition software.
Using the PittPatt face recognition program, the team was able to compare pictures available on the majority of Facebook user profile pages. It was then possible to almost instantly work out the identity of 30 percent of those photographed, says the Wall Street Journal.
The study showed that in 27 percent of cases it was also possible to, very simply, work out the first five digits of a social security number – solely from harvested information.
This was made possible by taking simple user-submitted information often available on profile pages, like the date of birth.
So then, not quite as harmless as some would have you believe.
It’s not beyond the realms of possibility for a burglar to snap someone leaving their house and use Facebook as a tool to figure out working schedules or when they are on holiday.
Of course, Facebook doesn’t force you to have a profile picture, as a spokesperson pointed out. Neither does it force you to keep the face recognition enabled. But it’s not something a lot of users will consider.
To prove the point, the lead researcher Professor Alessandro Acquisti has created a smartphone app that can identify anonymous members of the public on the go – as well as part of their social security number.
Of course he doesn’t plan to post it to the Android Market. But it paints a disturbing picture of how easy it is to manipulate face recognition technology.
According to spokesperson for privacy watchdog Big Brother Watch, Maria Fort, more should be done to ensure that the public is aware of the potential danger of social media misuse.
“Advances in the accuracy of facial recognition software are rapidly outpacing public awareness on internet privacy,” she told us. “The increasing ease with which people can match the image of an individual to a Facebook profile, education, employment history and National Insurance number means that anyone with the right technology has access to your most personal information.
“These kinds of programmes are fascinating advances on their own, but the potential for misuse is incredibly great, particularly when the public is unaware.
“Google and Facebook must only use these technologies on an ‘opt in’ basis. They have a duty to provide users with information risks they pose by releasing personal information and images.”