Meta has removed dormant facial recognition code from the companion app used with its smart glasses, according to a WIRED report.
The update came one day after WIRED reported that it had found the inactive code inside the Meta AI app. The app handles key features for Meta’s smart glasses and is required to pair the glasses with a user’s phone over Bluetooth.
WIRED said the removed feature was internally called NameTag.
WIRED found the dormant tool on June 4 while reviewing code for the Meta AI app. The code reportedly included algorithms that could turn photos of faces into biometric identifiers stored on a user’s device. The system could then compare new facial scans with those stored identifiers.
According to WIRED, the system was not active for users.
Meta released an app update on June 5 that removed the code.
WIRED reported that the latest version of the Meta AI app removed the face recognition software, the code that ran the NameTag recognition process, and a “Person recognized” alert that would have appeared if someone was identified.
The update also removed a folder where the app would have stored cropped images and biometric signatures of faces that the system captured but did not identify.
In February, The New York Times reported that Meta was working on facial recognition for its smart glasses.
That report also referred to the internal name Name Tag. The code later found by WIRED appears to be linked to those reported efforts.
The tool appeared to be designed to help users identify people they had previously met. Such a feature could help users remember names or recognize people. However, it also raises privacy concerns because it would involve face scans through a wearable camera. People near a smart glasses user may not know that their faces are being analyzed by a device.
Meta makes its smart glasses in partnership with Luxottica brands, including Ray-Ban and Oakley. The glasses have already raised privacy concerns because they can record people in public or private settings.
The report also noted that some social media influencers have used Meta smart glasses to record and harass women. In December, a woman was accused of breaking a man’s Meta glasses on the New York City subway.
Meta also faced a class action in March after a Swedish newspaper investigation reported that Kenyan workers reviewed footage from the company’s smart glasses. The reported footage included sexual intimacy and bathroom use, and some clips appeared to have been recorded without the owners’ knowledge.
Meta vice president of communications Andy Stone told WIRED on Monday that the feature was only a pilot effort.
Stone said Meta had not made a final decision on what to do with the feature, if anything.
Meta has previously said it is not building a central face database. The company has also said that nothing has shipped to consumers and that it would take a thoughtful and transparent approach if it decides to release such a feature.
The quick removal of the code is likely to add to concerns around facial recognition in smart glasses.
Although the feature was not active, its presence in a live companion app raised questions about how close Meta was to deploying the tool and how the company would handle consent from people being scanned.
For users, the concern is whether their glasses could become tools for collecting face data. For people around them, the concern is whether their faces could be analyzed without their knowledge.