The strategy of “disappearing into the bushes like Homer Simpson” is a bold choice.
Only a day after discovering a dormant piece of code that appeared to be a facial recognition algorithm in a companion app for its smart glasses, Meta released an update that removed that code, Wired reported. The publication first discovered the suspicious code, internally dubbed Name Tag in Meta, while reviewing the code of a Meta AI application that manages some key functionality of the glasses. In other words, the same app needed to pair the Meta smart glasses to a user’s phone via Bluetooth was also ready to start harvesting every face a user passed while wearing them.
Wired discovered the dormant tool on June 4. It contained algorithms that would have converted photos of faces into biometric identifiers stored on the device and referenced with each new facial scan. On June 5, an update was released that removed it entirely. In February, The New York Times had reported that Meta was working to bring facial recognition to its glasses. Since the Times heard about the internal nickname Name Tag at that time, the code discovered by Wired was probably the result of these efforts.
The way the tool works suggests that it might have been intended as a way for users to more easily identify people they had already met. A handy feature for the forgetful, no doubt, but also an extremely scary and invasive solution to a very common interpersonal dilemma. Most people would probably rather someone simply admit to forgetting their name than have their image ingested by a front-facing camera.
Meta smart glasses are made in partnership with popular Luxottica brands including Ray-Ban and Oakley. They are already causing concern, with social media influencers close to the manosphere using them to harass and record women. In December, a woman was accused of breaking a man’s Meta glasses on the New York City subway. Meta was also the subject of a class-action lawsuit in March after a Swedish newspaper investigation revealed that Kenyan workers were reviewing images of the company’s smart glasses — including scenes of sexual intimacy and bathroom use — that appeared to have been taken without the owners’ knowledge.
In a statement made to Wired On Monday, Meta’s vice president of communications, Andy Stone, was quoted as saying that the feature was only a pilot project and that the company had not made a “final decision on what, if anything, to do here.” That may be true, but real Meta employees were paid real money to spend their time writing, reviewing, and shipping that code into a live product. The fact that it was never activated is likely to be little comfort not only to owners who don’t want to turn into mobile data collection tools, but also to the people in those users’ lives who don’t want their faces unknowingly analyzed. The very fact that the code was so quickly removed and the PR statements released suggests that Meta knows it’s walking a tightrope with these kinds of invasive features.
