What the Hidden Face Recognition Code Actually Does
Meta facial recognition in smart glasses refers to experimental software, codenamed “NameTag,” embedded in Meta’s AI companion app that can detect, crop, and convert faces captured by wearable cameras into biometric identifiers stored on a user’s phone, then notify the wearer when a previously recorded face is recognized, raising new questions about smart glasses privacy, consent, and everyday surveillance in public spaces. According to WIRED’s reporting, which Engadget and Techloy summarize, three AI models in the app now handle face detection, cropping, and creation of faceprints. The feature is not active for consumers, and Meta says no biometric data is being sent to its servers. Yet earlier versions of the app exposed interface hints, such as a “Connections” menu promising to “remember the people you met,” suggesting Meta has been testing how a face recognition code layer might look in daily use.
Meta’s Shifting Relationship With Facial Recognition
The discovery lands in the context of Meta’s long, uneasy history with facial recognition. Back in 2010, Facebook launched a large tagging system that scanned billions of photos and later cost the company over $2 billion in settlements. Meta shut that system down in 2021 and said it deleted more than a billion faceprints, portraying the move as a reset on biometric tracking. Yet the company has since reintroduced facial recognition on Facebook and Instagram as a safety tool to spot faces in scam ads. Engadget notes that Meta has also previously explored facial recognition for smart glasses, and anonymous staff told The New York Times the internal tool was called “Name Tag.” Together, these moves show a pattern: Meta retreats when public pressure peaks, then slowly brings the technology back in narrower, less visible forms.
The ‘NameTag’ Vision: Convenience, Accessibility and Surveillance
NameTag is designed to let Ray-Ban and Oakley smart glasses users capture faces, store them locally, and receive alerts when those faces reappear. If enabled, faces that do not match a stored profile would sit in a “pending” list until the wearer labels them. Meta sources cited by The New York Times described ambitions to launch this during a “dynamic political environment,” when major critics might be distracted. Supporters argue there are real benefits, especially for people with visual impairments who struggle to identify acquaintances in social settings. But critics warn that blending faceprints with always-on cameras turns ordinary users into a networked sensor grid. Cooper Quintin of the Electronic Frontier Foundation warns that Meta has built “the capacity to turn their customers into a distributed surveillance machine,” and many privacy advocates see this as normalization of everyday biometric tracking.
Meta’s Assurances and the Question of Trust
Meta insists the discovery does not mean face recognition will ship to users. “Nothing has shipped to consumers and no final decision has been made on what to do here, if anything,” spokesperson Ryan Daniels told Engadget, adding that the company is “not building a central face database.” Meta frames the NameTag code as exploration and promises a “thoughtful approach” and “full transparency” if it moves ahead. Yet two security researchers told WIRED the feature appears close to deployment, and internal documents reported by The New York Times show Meta once timed a launch for when public scrutiny might be lower. That tension fuels Meta surveillance concerns: the company is asking people to trust early assurances even as its own product code, past reversals on biometrics, and internal planning memos point toward repeated attempts to expand facial recognition in new forms.
What Users Need: Transparency, Consent and Real Control
For smart glasses privacy to be credible, users and bystanders both need more than vague promises. More than 70 civil society groups, including the ACLU and Fight for the Future, have already urged Meta to drop facial recognition from its smart glasses entirely. At minimum, any launch would need clear, up-front consent, visible recording indicators, strong limits on retention, and easy ways to opt out or delete faceprints. People recorded by wearers should have rights too, not only device owners. Joseph Jerome, a former Meta Reality Labs policy official, told WIRED, “I don’t know how Meta can responsibly deploy a technology like this,” reflecting doubts from inside the company’s own orbit. The discovery of embedded face recognition code shows this debate is no longer hypothetical. It is a design decision waiting to be made in software updates.
