The finish line is still Sharp glasses’ facial recognition technology has been built into an app downloaded on millions of phones, according to software analysis by WIRED.
Code discreetly added to Meta’s AI app as part of multiple updates this year shows that the feature, internally called “NameTag,” identifies people captured by the glasses’ camera and, when activated, alerts the user when it recognizes someone.
The discovery of a NameTag element in a running Meta AI app shows that Meta has started sending facial recognition code to users’ phones, publicly describing it as something the company is still “thinking about.” In April, Meta said that if it were to employ facial recognition, it would not be implemented without a “very thoughtful approach.” However, WIRED found that as early as January, core system components were integrated into software distributed to millions of people.
While not yet enabled, NameTag is included in the Meta AI companion app, which has been downloaded over 50 million times and is crucial for using key features of sharp glasses, including Ray-Ban and Oakley models. If activated, it will convert the faces captured by Meta’s glasses into unique biometric signatures, commonly known as faceprints, and compare each of them to the faceprints stored on the user’s phone – a database that is currently set up to receive updates from Meta. Recognized faces will trigger notifications, and the rest will be cropped, indexed and saved to a folder marked as “pending”.
NameTag will resurrect a certain type of technology. Meta said it would end in 2021 when the company announced it would delete more than a billion facial prints belonging to Facebook users after years of controversy over its photo-tagging system. Meta ultimately paid $650 million to settle a class-action lawsuit brought by Illinois users, and in 2024 it agreed to a separate $1.4 billion settlement with Texas over allegations that it unlawfully collected biometric data from users.
Its renewed efforts come amid growing opposition to consumer-level facial recognition, which privacy advocates say will give everyone from stalkers to immigration agents basic access to the perilous technology. Internal meta documents published by New York Times in February showed that the company planned to implement the feature in a “dynamic political environment” when Meta thought its biggest critics would be busy.
Three NameTag-enabled AI models have already been deployed from Meta’s servers and are currently residing on its customers’ phones, according to a WIRED analysis that was independently reproduced by third-party experts. One model detects faces, the second crops them, and the third encodes them into biometric data.
Currently, only traces of the UI are present to indicate how this feature may eventually work. In the May version of the app, the feature was renamed “Calls,” encouraging people to “remember people they meet.” It is unclear whose faces will be included in the system’s recognition database, how these profiles are created, or how many people will ultimately be identified using them.
WIRED shared its findings with two third-party security researchers who separately reviewed the app and reproduced key aspects of the analysis: Cooper Quintin, a security researcher and senior public interest technologist at the nonprofit Electronic Frontier Foundation’s Threat Lab, and an independent security and privacy researcher who goes by the pseudonym Buchodi, who has spent more than a decade reverse engineering consumer software and surveillance technologies.
“This feature hasn’t been rolled out to consumers yet, but it appears to be almost ready to go,” says Quintin. “Despite billions of reasons not to do so, Meta appears to have built the ability to turn its clients into a distributed surveillance machine.”