Home > Biometrics, Facebook > Analysts raise alarm on Facebook expansion of facial recognition technology

Analysts raise alarm on Facebook expansion of facial recognition technology

June 9, 2011


The expansion of Facebook’s facial recognition technology into Australia has raised privacy and security concerns, with local analysts warning information accumulated by the social networking giant could be used for malicious purposes.

In a statement on its blog on Tuesday, Facebook announced it had launched the feature beyond North America into most countries. The feature uses facial recognition technology to speed up the process of tagging friends in photos.

The technology scans new uploaded photos and determines if the faces in these new images resemble those in previously tagged photos. If they do, the feature will suggest friends to tag. Previously, users had to tag photos manually, without suggestions.

According to the social network’s blog, more than 100 million photos per day are uploaded to Facebook and the goal of the feature is to make tagging and sharing photos easier.

The decision to automatically enable the feature for Facebook’s 500 million users has been criticised by digital media analysts.

Sam Yip, a senior researcher for independent telecommunications research company Telsyte, told SmartCompany that Facebook should take more care when introducing new features.

“Facebook is not obliged to tell anyone anything,” he says.

“But telling people that they can edit their settings would be the wise and responsible thing to do. There will be a large percentage of people who don’t check and it will be too late for people who don’t want to share every part of their lives on a digital space.”

Yip is wary of the security threat the facial recognition feature poses.

“I’ll be one of the savvy few that has a pulse on the industry and changes my settings,” he says.

“I would like to retain a sense of control of my identity. The feature suggests a selection of names and that’s already one piece of information that could be used for malicious purposes. The security concerns are not just online, but could now be brought offline as well.”

A senior technology consultant, Graham Cluley, wrote in his blog this week that Facebook enabled the feature without giving users any notice.

“Rather creepily Facebook is now pushing your friends to go ahead and tag you,” Cluley writes.

“Remember, Facebook does not give you any right to pre-approve tags. Instead the onus is on you to untag yourself in any photo a friend has tagged you in.”

“The onus should not be on Facebook users having to “opt-out” of the facial recognition feature, but instead on users having to “opt-in.”
“Many people feel distinctly uncomfortable about a site like Facebook learning what they look like, and using that information without their permission.”

Facebook then delivered this mea culpa on its blog: “We should have been more clear with people during the roll-out process when this became available to them”.

Facebook users can disable the feature by going into Facebook privacy settings, clicking on customise settings and editing the “suggest photos of me to friends” option.

Let’s take a look at some facts here:

– Facebook has 600 million members.

– Each day, Facebook’s members upload over 200 million photos, and Facebook currently hosts over 90 billion photos.

– Each time you “tag” a photo on Facebook, its facial recognition technology learns more about what that person looks like.

– Even if you happen to “opt out” of the facial recognition tagging, Facebook’s technology can surely use the tagged photos of you (hey, perhaps even the tagged photos of you that you end up un-tagging) to figure out what you look like.

– Right now Facebook is using this technology to help people tag photos. But once they have an accurate facial recognition database of several hundred million people? Hmm.

%d bloggers like this: