Meta rejected the use of face scanning technology in its AI smartglasses and subsequently discreetly removed any evidence of it.

      Meta’s face-recognition past resurfaces, complicating its AI smart glasses goals

      Meta's smart glasses privacy concerns have evolved once more. Following a discovery by WIRED of inactive references to face identification within the Meta AI app, that same code has reportedly disappeared after a subsequent app update.

      The smart glasses app revealed remnants of face-ID development.

      The code was linked to an internal initiative named “NameTag.” WIRED noted that the system was not active for users, but its existence indicated that Meta had moved past the initial concept stage and had commenced testing the potential for face identification within its smart glasses framework.

      WIRED reported that this dormant system seemed to convert faces into on-device identifiers that could be correlated with previously stored data. While this does not signify an imminent public implementation of facial recognition on Ray-Ban Meta smart glasses, it highlights why the finding has attracted attention.

      The sensitivity of the situation is heightened as this feature wasn't concealed within a research demonstration or a developer-exclusive version. It emerged in the app that typical smart glasses users regularly engage with. Given that this device includes a camera designed for public use, even inactive references to face recognition raise concerns regarding consent and the extent to which users are aware of what is being evaluated behind the scenes.

      Civil rights organizations had already raised alarms.

      This was not the initial warning regarding Meta’s aspirations for smart glasses. Prior reports indicated that civil rights groups expressed dissatisfaction with Meta’s supposed intentions to introduce facial recognition in its AI glasses. Advocates for civil rights contended that a feature capable of identifying individuals through wearable cameras could pose privacy threats to unsuspecting bystanders and further extend surveillance in everyday settings.

      The worry has intensified following the code’s removal. Meta’s communications executive Andy Stone informed WIRED that the feature was part of a pilot program, and the company had not yet made a decision on its implementation. This may clarify why the feature was inactive, yet it does not explain why face-ID code was present in an app designed for standard smart glasses users.

      Given Meta’s previous experiences with this technology, it becomes challenging to dismiss such concerns. In 2021, Facebook announced it would discontinue its facial recognition system and delete recognition templates for over a billion users, due to privacy and regulatory issues. The recent report does not confirm that facial recognition is on the horizon for Meta glasses. However, the emergence of dormant face-ID code in a consumer app, followed by its removal after being reported, complicates the perception of Meta’s interest as merely hypothetical.