Meta Quietly Added Face Recognition Code to Millions of Phones

Meta has been quietly embedding facial recognition code into its smart glasses platform through routine app updates, a new investigation has revealed. The AI feature, internally called NameTag, was discovered by Wired buried within the Meta AI app — software already installed on millions of phones worldwide. The finding has reignited concerns about how tech companies handle biometric data.

How NameTag Works

The system relies on three separate artificial intelligence models working together. The first model detects faces captured by the smart glasses camera. A second model crops those faces from the surrounding image. The third converts facial features into biometric data — unique measurements that can identify a specific person, much like a digital fingerprint.

Once the system captures someone's face, it could later notify the wearer when that person is seen again. Past versions of the app even contained interface elements like a "Connections" menu that prompted users to "remember the people you met." While none of this functionality is currently active or sending data to Meta's servers, the core code has been distributed to devices since early 2026.

Privacy Concerns Run Deep

Wearable face recognition is widely viewed as far more invasive than traditional photo tagging on social media platforms. Unlike uploading a photo online, smart glasses operate in real time and could potentially identify strangers in public spaces without their knowledge or consent.

Meta spokesperson Ryan Daniels said the findings are "just evidence of exploration" and that no final decision has been made. The company also stated it is "not building a central face database." However, internal discussions reportedly included timing a potential launch during periods when privacy advocacy groups would have their attention focused elsewhere.

Meta previously retired facial recognition from Facebook in 2021 after widespread backlash but quietly reintroduced similar technology to Instagram and Facebook in 2024 for safety-related features. The New York Times first reported in February 2026 that Meta was actively exploring face recognition capabilities for its Ray-Ban smart glasses line. The discovery raises critical questions about transparency and consent in consumer technology. With AI-powered wearables becoming increasingly common, the line between helpful features and surveillance tools continues to blur — and users deserve to know exactly what code is running on their devices.