Meta removes face-recognition NameTag from smart glasses app after WIRED exposé

One day after WIRED exposed that Meta had quietly embedded an unreleased face-recognition system called NameTag into the Meta AI app—installed on over 50 million phones—the company removed it. The latest version of the app strips out code libraries explicitly named for face recognition, along with the recognition process, a 'Person recognized' alert, and the folder that stored cropped images and biometric signatures of unrecognized faces. Meta's VP of communications Andy Stone described the feature as purely exploratory, stating that no final decision had been made. However, WIRED's analysis found the code had been present since January, even as Meta publicly denied the feature's existence. Chief Technology Officer Andrew Bosworth called the reporting 'incredibly misleading.'

The removal leaves only faint traces: an internal debug menu label and a dormant profile link. Privacy advocates, including the ACLU of Massachusetts, argue the initial decision to ship the code is evidence that stronger legal protections for consumer privacy are urgently needed. Meta declined to answer ten specific questions from WIRED, including whether it had already created a faceprint database, how long data from unrecognized people is retained, and whether users would be allowed to opt in or out. The incident highlights ongoing tensions between Meta's ambitions for augmented reality and pervasive surveillance concerns, especially as the Massachusetts House unanimously passed a privacy bill last week.