A WIRED investigation found that Meta has incorporated facial recognition technology into applications connecting to its smart glasses, raising questions about the company's privacy practices and regulatory compliance ahead of any public launch.
The feature, identified as "NameTag," was integrated through several app updates in 2026. It uses three AI models implemented through Meta's servers: one detects faces, a second crops them, and a third encodes them into biometric data stored on users' phones. While the feature has not yet been activated for consumers, independent security researchers told WIRED it appears nearly ready to deploy.
Meta spokesperson Ryan Daniels said the company is merely "exploring" the technology and has made no final decision. "Nothing has shipped to consumers," Daniels said. "If we do decide to roll something out, we will take a thoughtful approach and do so with full transparency." He added that Meta is not building "a central face database."
In February, a Meta executive told the New York Times that while the company had heard interest in features like NameTag, it was still "thinking through options" before any rollout.
What the Left Is Saying
Privacy advocates argue the discovery demonstrates why federal and state regulators should move quickly to restrict facial recognition on wearable devices. In April, more than 70 advocacy groups called on Meta to abandon NameTag entirely, warning it could enable stalkers and abusers to covertly identify strangers in public.
"Despite the billions of reasons not to, Meta seems to have created the capacity to turn their customers into a distributed surveillance machine," said Cooper Quintin, a security researcher and senior public interest technologist at the Electronic Frontier Foundation's Threat Lab. "This technology, if activated, would fundamentally change the power dynamics between strangers in public spaces."
Democratic lawmakers have proposed legislation that would require explicit consent before any biometric data collection on wearable devices. Senator Elizabeth Warren of Massachusetts has called for expanded FTC authority to investigate pre-deployment surveillance features as potential consumer fraud.
Civil liberties organizations argue that Meta's actions—distributing recognition code while publicly claiming uncertainty about deployment—may constitute deceptive practices under existing consumer protection law, regardless of whether the feature is ultimately activated.
What the Right Is Saying
Industry analysts and some Republican lawmakers have pushed back against what they characterize as regulatory overreach. They note that facial recognition technology has legitimate applications in accessibility, security, and commerce.
"Before we ban technologies, we should understand what they're actually being used for," said Senator Josh Hawley of Missouri. "Facial recognition can help the blind navigate public spaces—it was one of the reasons Mark Zuckerberg supported research in this field. Premature regulation could foreclose beneficial innovations."
Tech industry groups argue that Meta's transparency about exploring the technology, combined with its stated commitment not to build a centralized face database, demonstrates appropriate corporate responsibility without government intervention.
"Companies should be allowed to develop features and gauge consumer interest before committing to deployment," said Carl Szabo, vice president and general counsel at NetChoice, a tech industry trade group. "What Meta is doing—exploring options publicly and seeking input—is exactly the kind of engagement Congress should encourage."
What the Numbers Show
Meta previously deployed facial recognition on Facebook beginning in 2010. The system analyzed photographs and suggested tags for users appearing in images, becoming one of the largest consumer face-recognition systems ever implemented.
European regulators raised concerns about that system's legality in 2011. In 2019, Facebook agreed to pay a $5 billion settlement to the Federal Trade Commission as part of a broader privacy case that included scrutiny of facial recognition practices.
According to WIRED's analysis, NameTag relies on three separate AI models working through Meta's servers: one detection model, one cropping model, and one encoding model. The feature stores biometric data locally on users' phones rather than in a centralized database—a technical architecture that could complicate regulatory oversight.
More than 70 advocacy groups signed the April letter calling for NameTag's abandonment, citing concerns about potential misuse by stalkers and abusers.
The Bottom Line
The discovery raises questions about what obligations technology companies have when distributing surveillance capabilities to consumer devices before activating them. Meta maintains it is engaged in legitimate exploration of emerging technology; critics argue that distributing recognition code without activation represents a form of deceptive preparation for privacy-invasive features.
Lawmakers on both sides of the aisle have expressed interest in examining the issue, though approaches differ. Progressive Democrats favor explicit consent requirements and expanded FTC authority. Some Republican lawmakers warn against premature restrictions that could limit beneficial applications.
The feature remains inactive, according to Meta, with no final decision made about its future. However, security researchers note that NameTag appears technically functional and ready for activation through a software update. What happens next likely depends on regulatory pressure, public backlash, and whether Congress moves to restrict biometric data collection on wearable devices.