What did Meta add for face recognition?
Meta’s unreleased smart-glasses face-recognition code
Meta embedded an unreleased “NameTag” face-recognition feature in its smart glasses ecosystem, and the code appears to have been distributed widely through updates to the Meta AI app on millions of phones, according to analysis reviewed by WIRED.
Instead of being a public-facing product, the feature seems to have been quietly staged: the discovery centered on application code that would enable the glasses platform to identify people using biometric data stored on users’ devices. The key point is that this capability was not clearly communicated as a live feature to end users at the time the code was present.
From a product and privacy standpoint, this matters because face-recognition systems are categorically sensitive. They raise questions about:
- User consent and transparency (whether people knew identification features were being developed or enabled)
- Data handling boundaries (how biometric data is used and where it resides)
- Deployment risk (code reaching large numbers of devices before a feature is officially launched)
In practical terms, the rollout path described by the reporting suggests a common pattern in modern software delivery: features can be staged behind the scenes, then activated later—sometimes after code has already been shipped broadly. That approach can accelerate development, but it can also amplify privacy concerns if customers are not clearly informed about what is being tested or enabled.
The discovery also adds to ongoing scrutiny of consumer AI and wearables, where companies experiment with capabilities like identification, assistants, and on-device processing. Even when specific behavior is not confirmed as “active” across all devices, the presence of code for face recognition is still a notable signal for what’s being prepared for future glasses functionality.