Meta quietly embedded face-recognition technology into its AI app — downloaded more than 50 million times — months before publicly acknowledging the feature was under consideration, according to a WIRED analysis of the company’s software published in June 2026.
The feature, internally called “NameTag,” is designed to identify people captured by Meta’s smart glasses cameras, including Ray-Ban and Oakley models. When activated, it would convert faces into unique biometric signatures called faceprints, match them against a database stored on the user’s phone, and alert the wearer when someone is recognized. Unrecognized faces would be cropped, indexed, and saved to a folder marked “pending.” A May 2026 version of the app rebrands the feature as “Connections,” with the prompt “remember the people you met.”
Though NameTag is not yet enabled, WIRED found that core components had been integrated into the app as early as January 2026 — well before Meta said in April that it was still “thinking through” whether to use face recognition and would not roll it out without “a very thoughtful approach.” Three AI models that power NameTag, handling face detection, cropping, and biometric encoding, have already been deployed to users’ phones from Meta’s servers.
WIRED’s findings were independently reproduced by two outside researchers: Cooper Quintin of the Electronic Frontier Foundation’s Threat Lab and an independent researcher known as Buchodi. “Despite the billions of reasons not to, Meta seems to have created the capacity to turn their customers into a distributed surveillance machine,” Quintin said.
The development revives a technology Meta said it retired in 2021, when it deleted over a billion faceprints from Facebook following years of legal controversy. Meta paid $650 million to settle a class-action lawsuit in Illinois and, in 2024, agreed to a separate $1.4 billion settlement with Texas over allegations of unlawful biometric data collection.
Internal Meta documents reported by The New York Times in February showed the company had planned to launch the feature during a period when it believed its biggest critics would be occupied elsewhere. It remains unclear whose faces would be included in the recognition database or how user profiles would be created.
Source: WIRED