Meta Quietly Strips Face-Recognition Code From Smart Glasses App After Public Scrutiny

Meta removed facial-recognition software components from its Meta AI companion app on June 8, one day after WIRED published an investigation revealing the company had covertly embedded an unreleased system called NameTag into the application. According to WIRED’s code analysis, the June 8 version stripped out code libraries explicitly named for face recognition that were present in earlier builds. However, Meta’s response has raised new questions about the system’s scope and the company’s transparency with users.

The NameTag Architecture and Embedded Code

According to WIRED, NameTag was designed to capture faces through Meta’s smart glasses, convert them into biometric signatures (commonly called faceprints), and match them against a database of face profiles stored locally on a user’s device. The system would also crop, index, and store on-device any faces it failed to recognize for later processing. Wired’s investigation revealed that substantial portions of this machinery were integrated into the Meta AI app as early as January 2026—months before Meta publicly said it had reached no final decision on facial recognition.

The app, which is installed on more than 50 million phones, contained multiple code libraries with explicit facial-recognition naming conventions until Meta removed them following WIRED’s report.

Meta’s Evasive Response and Unanswered Questions

Meta’s public response has been notably defensive. According to WIRED, Andy Stone, Meta’s vice president of communications, initially dismissed the findings by stating the company “couldn’t answer questions about how the system would work because the feature does not exist.” Andrew Bosworth, Meta’s chief technology officer, called the reporting “incredibly misleading” and “absolutely dishonest.”

Despite this dismissal, Meta did not deny the presence of facial-recognition code in the app—it removed it instead. The company declined to answer 10 specific questions WIRED posed before publication, including whether it has already created the facial-profile database NameTag relies on, how long biometric data of unrecognized people is retained, and whether users will receive opt-in or opt-out controls over the system.

The Broader Context: NameTag’s Origins and Timeline

NameTag first became public in February 2026, when The New York Times reported—citing internal Meta documents—that the company was developing facial recognition for smart glasses with a potential launch as early as that year. One internal memo reportedly described releasing the feature during a “dynamic political environment,” when privacy advocates might be distracted.

This timeline suggests Meta had been building and deploying facial-recognition infrastructure while publicly maintaining that no final decision had been made.

Why This Matters

The removal of code does not resolve the fundamental transparency gap. Meta has not clarified whether the facial-profile infrastructure NameTag requires has already been constructed, whether on-device biometric data could eventually be transmitted to company servers, or what privacy controls—if any—users will have over this capability.

For organizations evaluating Meta’s smart glasses platform and the Meta AI app, this incident underscores the risk of embedded, unarticulated capabilities in widely distributed software. Privacy advocates have warned that facial-recognition systems embedded in glasses could facilitate stalking and abuse by enabling identification of strangers in public spaces. Until Meta addresses the unanswered questions about data retention, database status, and user consent mechanisms, the risk profile of the Meta AI app remains unclear.