Meta Deletes Face-Recognition System From Its Smart Glasses App After WIRED Report

Overview

Meta deleted a hidden facial-recognition system from Meta AI, the companion app for its Ray-Ban and Oakley smart glasses, one day after WIRED found the dormant code in June 2026. The system, internally called Name Tag, was built to turn face photos into biometric identifiers stored on the device and match them against new scans to flag a recognized person to the wearer. Meta shipped a June 5 update stripping the code, the machine-learning models, and the alert system, and declined to say why the code was there or whether the feature returns.

Key Takeaways

Meta removed an unreleased facial-recognition feature, known inside the company as Name Tag, from its smart glasses app on June 5, 2026, one day after WIRED reported the code.
The dormant code was built to convert faces into biometric signatures stored on the device and cross-reference them against new scans to identify people in a wearer's view.
The affected Meta AI app had been installed on more than 50 million phones, so the inactive code already sat on tens of millions of devices before removal.
Meta said the work was a pilot effort with no final decision made, and its communications VP Andy Stone criticized the WIRED report for understating that the feature was never switched on.
More than 70 advocacy organizations had urged Meta to drop facial-recognition plans for its glasses, and the Electronic Frontier Foundation called the removal a win driven by public pressure.
The system had reportedly sat dormant in the app since at least January 2026, and a New York Times report in February 2026 first described Meta developing face recognition for the glasses.

Stats & Key Facts

#More than 50 million phones had the Meta AI app installed, putting the dormant code on tens of millions of devices.
#Meta deleted the code 1 day after WIRED published its report, shipping the update on June 5, 2026.
#More than 70 advocacy organizations had pushed Meta to abandon facial-recognition plans for its smart glasses.
#Three machine-learning models were found in the code: one for face detection, one for cropping, and one for encoding biometric data.
#The feature had reportedly sat dormant in the app since at least January 2026 before its discovery.

What the Name Tag Facial-Recognition Code Was Built to Do

The hidden system was designed to identify people a wearer looked at through the glasses.

›Converted photos of faces into biometric signatures, sometimes called faceprints, that act as a numeric fingerprint of a face.
›Stored those identifiers on the device rather than only in the cloud, and cross-referenced each new scan against them.
›Triggered a person-recognized alert to the wearer when a face matched, surfacing information through Meta's AI assistant.
›Cropped and indexed faces the system did not recognize so they could be processed later.

How WIRED Found the Hidden Code and How Fast Meta Pulled It

Reporters located the inactive software inside the app that runs the glasses' core features.

WIRED reporters found dormant facial-recognition code buried inside Meta AI, the companion app that connects the smart glasses to a phone and runs their core features. The code was labeled Name Tag inside the company and included recognition libraries, a processing pipeline, and a person-recognized alert mechanism.

One day after the report, on June 5, 2026, Meta shipped an app update that deleted the facial-recognition code, the machine-learning models, and the alert system. Reporters noted that a few minor traces, such as a debug menu label and a dormant profile link, remained in the latest version.

Why Dormant Code on 50 Million Phones Raised Privacy Alarms

The feature was never switched on, but its reach concerned privacy advocates.

The Meta AI app had been downloaded onto more than 50 million phones, so the inactive code already sat on tens of millions of devices before it was removed. Even unused, the presence of an identification system designed to name strangers in public spaces drew strong criticism.

Privacy and civil-liberties groups warned that a tool letting a wearer identify people in their field of view could turn everyday glasses into a distributed surveillance device. More than 70 advocacy organizations had urged Meta to abandon facial-recognition plans for the glasses, citing risks to privacy and public safety.

Meta's Defense: A Pilot Effort and Criticism of the Report

The company framed the work as exploratory and pushed back on coverage.

›Andy Stone, Meta's vice president of communications, said the work was a pilot effort and that no final decision had been made on what to do, if anything.
›Stone described the feature as exploratory and criticized the WIRED report for understating that the feature was never enabled for users.
›Meta declined to explain why the code was present or whether the feature returns in a future build.

The New York Times Report and Meta's Super Sensing Ambitions

Reporting earlier in 2026 set the stage for the discovery.

The New York Times reported in February 2026, citing internal Meta documents, that the company was developing face recognition for its smart glasses and weighing a launch as soon as that year. One internal memo reportedly described releasing the feature during a dynamic political environment when privacy and civil-liberties advocates would be focused on other concerns.

The same reporting described a Meta model referred to internally as super sensing, designed to keep the glasses' cameras and sensors running continuously to build a running record of a wearer's day. The Name Tag code fit that broader direction toward always-on perception.

EFF Calls the Removal a Win but Warns the Ambition Remains

The Electronic Frontier Foundation tied the deletion to public pressure.

The Electronic Frontier Foundation framed the removal as a win driven by public backlash after the report spread. The group argued that there are strong reasons not to turn customers into a distributed surveillance machine and credited the speed of the change to public attention.

EFF also cautioned that the deletion does not settle the matter. The group called for enforceable privacy laws, a private right of action so individuals can sue over biometric misuse, and transparency about any data gathered during internal testing, warning that Meta's facial-recognition ambitions are unlikely to disappear.

Frequently Asked Questions

What was Meta's Name Tag feature?

Name Tag was the internal name for an unreleased facial-recognition system inside the Meta AI smart glasses app. It was built to turn faces into biometric identifiers stored on the device and match them against new scans to identify people a wearer looked at.

Was the facial-recognition feature ever turned on for users?

No. Meta said the feature was never enabled and described it as a pilot effort. The code sat dormant in the app, but its presence on tens of millions of phones still raised privacy concerns.

Which smart glasses does the Meta AI app support?

The app connects to Meta's smart glasses sold under the Ray-Ban and Oakley brands and runs their core features, including the AI assistant.

Why did Meta remove the code so quickly?

Meta deleted the code one day after WIRED published its report, on June 5, 2026. The company declined to explain why the code was there or whether the feature returns, and the EFF credited the fast removal to public pressure.

How many devices were affected?

The Meta AI app had been installed on more than 50 million phones, so the inactive facial-recognition code already sat on tens of millions of devices before Meta removed it.

Meta stripped the Name Tag facial-recognition system from its smart glasses app within a day of being exposed, but with no explanation for why the code existed and no commitment to drop the feature, privacy advocates say the underlying ambition is still in play.

Why It Matters for Business

Real business deployments are the most reliable signal of where AI is generating measurable ROI. Watching which sectors operationalize AI, what they pay for it, and how it changes their P&L tells you more than any vendor demo. These case studies are what serious buyers and investors triangulate on.