An analysis by Wired magazine revealed that Meta secretly incorporated facial recognition technology into its smart glasses via an app downloaded on millions of phones worldwide.
The code, added through multiple updates to the "Meta AI" application, reveals an internal feature called NameTag, capable of identifying people captured by the glasses' camera and sending a notification to the glasses wearer when they are recognized.
This discovery comes at a time when Meta was publicly describing the technology as still under development, stating last April that it would not roll out the feature without a "very careful approach." However, Wired found that the core components of the system had already been integrated into software distributed to millions of users since January.
Although not yet activated, the feature is embedded within an app that has been downloaded more than 50 million times and is essential for using Ray-Ban and Oakley smart glasses. When activated, it will convert faces into facial fingerprints and scan them against a database on the user's phone that receives updates from Meta's servers.
This move marks a return to a technology that Meta halted in 2021 after widespread controversy, when the company paid $650 million to settle a class-action lawsuit in Illinois, and in 2024 agreed to a $1.4 billion settlement with Texas over the illegal collection of biometric data.
Privacy advocates warn that this technology will give anyone, from stalkers to immigration officials, a dangerous tool. One security researcher said, "Meta has created the ability to turn its clients into a distributed surveillance machine."
A researcher conducted successful tests on the feature and confirmed that "the key components for facial recognition are already in place, and there aren't many pieces that stand between this and a working feature."
Meta, for its part, said it is still exploring these features and that no final decision has been made, emphasizing that it is not building a centralized facial database. However, a review of the code reveals that the system is designed to pull facial scans from its servers and store them on users' devices.
