What the NameTag Discovery Is and Why It Matters
The NameTag discovery refers to dormant face-recognition code found inside Meta’s Meta AI companion app, silently shipped to tens of millions of phones and later removed after public scrutiny highlighted the privacy risks. WIRED’s reverse-engineering work uncovered internal “NameTag” modules, three AI models, and user interface traces that could have turned faces into local faceprints and triggered recognition alerts on connected smart glasses. According to WIRED, core NameTag components were not confined to lab builds but appeared in public Meta AI updates beginning in January 2026. By the time this came to light, the app had surpassed 50 million installs, meaning a large installed base carried a nearly ready biometric identification pipeline. Even though Meta describes the work as “exploratory,” privacy groups and security researchers see the episode as a warning sign for face recognition privacy and Meta AI security.
Dormant Code Risks: Why ‘Off’ Does Not Mean Harmless
Dormant code risks arise when powerful features are bundled inside apps but kept disabled, creating silent technical capacity without user awareness or consent. In the Meta AI case, NameTag’s face-recognition libraries and models were present even though the feature was not publicly launched. That matters because any future toggle, misconfiguration, or security flaw could activate a full biometric pipeline on devices that users believed were running a simple AI companion. Privacy advocates highlight that shipping such components “before public debate risks normalizing surveillance,” especially when they relate to face recognition privacy in public spaces. The existence of debug menus and links to person profiles shows the system was closer to deployment than an abstract experiment. For regulators, the concern is not only what an app does today but what it is technically able to do the moment a company changes its settings or policy.
How Meta Responded: Code Removal Without Full Answers
After WIRED reported the hidden NameTag components, Meta pushed a June 2026 Meta AI update that stripped the main face-recognition libraries from the app. Analyses of that release show that the building blocks for device-side identification were largely removed, including modules that could create local faceprints and link them to notifications on Ray-Ban and Oakley smart glasses. While this quiet code purge reduced immediate dormant code risks, it did not resolve key questions. Meta spokespeople declined to specify whether any images or biometric data had been retained, how long any data might have lived on devices, or whether NameTag could return in a different form. Civil liberties groups welcomed the removal but argued it proves the need for stronger oversight of Meta AI security and clearer app store disclosures when biometric features, even experimental ones, are bundled into consumer apps.
What This Reveals About Smart-Glasses and Privacy
NameTag was designed for Meta’s smart glasses ecosystem, turning the Meta AI app into a hub for real-time recognition of people in a wearer’s field of view. Code reviewed by WIRED indicated that faces could be detected, cropped, encoded into faceprints, and matched locally on the phone, then used to notify wearers when someone was recognized. With more than 50 million downloads of the companion app and three AI models already delivered to devices, smart-glasses users were closer to everyday AR surveillance than most realized. Critics warn this can enable stalkers and normalize constant identification in public spaces. Supporters of assistive uses argue that some of the same tools might help low-vision users identify acquaintances. The clash shows how one codebase can support both accessibility and intrusive tracking, making transparent limits and consent central to face recognition privacy.
Practical Steps: How to Protect Yourself Now
For users, the immediate step is to update the Meta AI app to the latest version so that removed NameTag components are no longer present on your phone. Then run an app permissions audit: check which permissions Meta AI has, especially camera, microphone, local storage, and nearby device access, and disable anything that is not essential to how you use the app. Consider whether you need the app at all if you own compatible smart glasses; uninstalling remains the simplest way to cut dormant code risks. Review privacy settings on any connected Ray-Ban or Oakley frames and turn off logging or media-saving options you do not need. Finally, stay alert for future Meta AI security updates and policy changes. If Meta or any other developer adds new biometric features, look for clear opt-ins, readable explanations, and the option to decline without losing core functionality.
