What Meta’s Hidden Code Reveals About Smart Glasses Privacy
Meta’s undisclosed “Name Tag” feature in Ray-Ban smart glasses shows how wearable devices can silently add facial recognition capabilities that turn everyday cameras into persistent surveillance tools without clear user awareness or consent. At the center is a Meta AI companion app installed on tens of millions of phones for Ray-Ban smart glasses owners. WIRED’s investigation found that substantial, dormant Meta facial recognition code was already shipped, designed to identify people the wearer encounters. This discovery has intensified Ray-Ban smart glasses privacy debates, because the feature was never announced, displayed no interface, and yet appeared close to launch-ready. Although Meta says “nothing has shipped to consumers,” the presence of near-complete code blurs the line between internal experiment and live deployment, and raises new smart glasses privacy risks for bystanders who never agreed to have their faces scanned at all.
Inside ‘Name Tag’: A Personal Face Database on Your Phone
Name Tag was engineered to turn Ray-Ban smart glasses into a real-world tagging engine, quietly assembling a personal biometric database. According to Gadget Review, the buried system would detect faces in photos or video, crop them, and convert them into biometric signatures, or “faceprints,” stored on the user’s phone. When a wearer met the same person again, the app would notify them that a “person was recognized,” enabling ongoing tracking of people previously captured. Faces that could not be matched were still saved locally “for future processing,” expanding the database over time. PetaPixel reports that Meta engineers also explored versions of the Name Tag feature that could identify people connected to the user on Meta platforms or anyone with a public social media account, deepening wearable surveillance concerns about how easily strangers might be identified in public spaces.
A Rapid Deletion That Raises More Questions Than It Answers
Once WIRED publicised the dormant Name Tag system, Meta moved fast. Within 24 hours, an updated Meta AI app stripped out the face recognition libraries, biometric processing workflow, and storage locations for cropped faces, leaving only inert debug fragments. This rapid cleanup, reported by both Gadget Review and PetaPixel, has prompted fresh Ray-Ban smart glasses privacy questions. If the feature “does not exist,” as Meta spokesperson Andy Stone told WIRED, why did so much working infrastructure ship to consumers and then vanish overnight? Meta declined to answer detailed questions on how long photographs and biometric data of unrecognized people would have been stored, or whether any data would ever sync back to company servers. The quiet removal suggests the system was more than a theoretical prototype, and that public exposure, not proactive transparency, drove its disappearance.
Consent, Bystanders and the New Wearable Surveillance Normal
The Name Tag feature highlights a growing gap between what smart glasses can do and what users and bystanders have meaningfully agreed to. Always-on cameras tied to Meta facial recognition can turn everyday encounters into indexed records, even if bystanders never look at the device or know they are being scanned. More than 70 advocacy organizations warned Meta that such tools could help stalkers and abusers identify strangers in public, a core part of today’s wearable surveillance concerns. The ability to store unknown faces “for future processing” makes encounters searchable long after they occur. For Ray-Ban smart glasses owners, the episode shows that invisible software updates may quietly expand smart glasses privacy risks without clear notice, while people around them carry the surveillance burden without any option to opt out.
A Pattern of Pushing Privacy Boundaries in Consumer Wearables
Meta describes Name Tag as “purely exploratory,” yet its secret inclusion and rushed removal fit a pattern of testing privacy limits until public scrutiny forces retreat. Shipping extensive, dormant facial recognition code into a mass-market wearable ecosystem signals that such capabilities are not far-off concepts but real options waiting for the right moment. For regulators and privacy advocates, the case shows why relying on voluntary restraint is risky when biometric tracking can be activated in software. For consumers, it underlines that wearable surveillance concerns are not abstract: the hardware on their faces may be able to do more than advertised. As smart glasses become more common, meaningful consent, strict limits on biometric data, and enforceable oversight will determine whether they remain useful tools or slip into normalized, always-on surveillance.
