← Back to feed

WIRED reviewed code in Meta’s smart glasses platform and found an unreleased facial recognition system already embedded in apps running on millions of phones. The feature, internally called “Name Tag,” would let wearers of Meta’s Ray-Ban and Oakley smart glasses identify people in real time by matching their faces against Meta platform profiles.

The code has not been activated for users. But it exists, it is already on devices, and it can be switched on without a new install.

What Name Tag Does

When active, a glasses wearer looks at a person; Meta’s AI identifies them by matching the face against Facebook and Instagram profiles; the system surfaces the person’s name and profile details. Meta’s internal documents describe the feature as matching against people connected on Meta platforms and those with public Instagram accounts.

Meta has 3.3 billion monthly active users across Facebook and Instagram. “Limited to connected users” covers nearly half the planet.

The glasses hardware already captures continuous video. EssilorLuxottica, Meta’s manufacturing partner, sold more than 7 million pairs in 2025. The indicator light that signals recording is dim enough in daylight that most bystanders do not notice it.

The Internal Memo

An internal Meta document from May 2025, reviewed by The New York Times earlier this year, stated the company planned its launch strategy explicitly around regulatory distraction:

“We will launch during a dynamic political environment where many civil society groups that we would expect to attack us would have their resources focused on other concerns.”

This is not a company uncertain about whether the feature raises privacy concerns. It is a company that knows the feature raises concerns and is timing the release around when those concerns are least likely to be acted on.

The Settlement History

Meta has paid approximately $7 billion to settle biometric privacy lawsuits:

• 2019: $5 billion FTC settlement, partly over facial recognition
• 2021: $650 million Illinois class action over Facebook photo tagging
• 2024: $1.4 billion Texas biometric privacy settlement

As part of the FTC settlement, Meta agreed to review every new product for privacy risks before launch. In January 2025, Meta restructured its privacy review process: privacy teams lost authority over product releases, and time limits were imposed on how long reviews could run. Andie Millan, a director of risk review in Meta’s Reality Labs, told employees in a recorded internal meeting that the new structure would “push the bounds” of the FTC agreement.

What Changes With Shipped Code

The difference between an investigation revealing Meta’s plans for facial recognition (the NYT story from February) and WIRED finding the code already embedded in production apps is significant.

Plans can be walked back. A privacy review can pause a launch. Shipped code that reaches millions of devices silently, without user disclosure, is a different kind of action. The code is already distributed. Meta controls the flag that activates it.

Seven million smart glasses are in circulation. The camera does not announce itself. The LED indicator is small. Most people walking past a Ray-Ban wearer have no idea they are being recorded, and would have no idea they are being identified if the feature were switched on overnight.

The Regulatory Position

The FTC’s 2019 consent decree applies. The EU’s biometric data rules under GDPR apply in Europe. Illinois BIPA and Texas CUBI have already cost Meta nearly $2.1 billion combined and cover data collection on residents of those states regardless of where Meta is based.

The question is not whether the legal exposure exists. It is whether enforcement moves faster than deployment.

Meta’s statement: “We’re still thinking through options and will take a thoughtful approach if and before we roll anything out.” The code on millions of phones does not wait for thinking to finish.