petapixel.com via Reddit

Meta Removes Ray-Ban Name Tag Facial Recognition Code

meta surveillance ai ethics surveillance privacy facial-recognition

Key insights

  • Meta deleted the 'Name Tag' code from its Ray-Ban app after WIRED reported the dormant facial recognition feature on June 4.
  • The code converted face photos into biometric signatures and stored unrecognized faces locally, with two deployment variants designed.
  • More than 70 advocacy organizations had urged Meta to abandon facial recognition plans before the code was quietly removed.

Why this matters

If the Name Tag code processed real users' faces before removal, Meta's 'exploratory' framing offers no shield from biometric privacy laws like Illinois BIPA that attach liability at the point of collection, not at product launch. Andy Stone's simultaneous claims that the feature was 'purely exploratory' and 'does not exist' set up a direct contradiction that regulators and potential litigants can exploit when probing what data was actually generated. Consumer-facing AI hardware is now a front line for biometric privacy enforcement, and this incident establishes that dormant, unactivated code in a shipped app is fully subject to press investigation, advocacy pressure, and regulatory scrutiny.

Summary

Meta quietly deleted the dormant 'Name Tag' facial recognition code from its Ray-Ban smart glasses companion app after WIRED reported discovering the unactivated feature on June 4. The code converted face photos into biometric signatures and compared them against databases stored on users' devices, with unrecognized faces cropped, indexed, and held locally. Engineers had designed two versions: one limited to existing Meta contacts, and a broader build capable of identifying anyone with a public social media account. Essentially: Meta and WIRED are in a disclosure standoff over undisclosed biometric data practices in consumer wearables. - Andy Stone, Meta's vice president of communications, called the feature 'purely exploratory' and said 'no final decision has been made on what to do here, if anything,' while also claiming 'the feature does not exist.' - More than 70 advocacy organizations had already urged Meta to abandon the plans, warning of serious risks to privacy and public safety, including risks to stalking and abuse victims. - Meta declined to answer ten questions from WIRED about database creation, data retention, and whether biometric data would be transmitted to Meta's servers. Deleting the code without explanation leaves unresolved whether any facial signatures already generated from real users were purged or retained.

Potential risks and opportunities

Risks

  • Meta faces potential enforcement under biometric privacy laws in states like Illinois if real user facial data was generated and not purged after the June 4 WIRED report and subsequent code deletion
  • Andy Stone's contradictory statements, 'purely exploratory' versus 'the feature does not exist,' could become a litigation liability if discovery in any future suit shows the code processed actual biometric data from real Ray-Ban users
  • Other wearable AI camera makers and hardware startups now face preemptive advocacy pressure and regulatory scrutiny for features still in internal development, raising compliance costs before products ship

Opportunities

  • Biometric data governance platforms (OneTrust, BigID) are positioned to sell compliance frameworks to AI hardware companies seeking to prove dormant code does not constitute active data collection
  • Privacy-first smart glasses competitors can differentiate by publishing explicit no-biometric-data commitments before launch, directly capitalizing on the trust gap Meta created with the Name Tag disclosure
  • State legislatures without existing biometric privacy statutes gain political momentum to introduce BIPA-style coverage for AI-enabled wearables, creating advisory and compliance consulting demand in those markets

What we don't know yet

  • Whether biometric facial signatures already generated by the Name Tag algorithms before code removal were deleted from user devices or retained by Meta
  • What Meta would reveal if required to answer the ten specific questions WIRED posed about database creation, data retention, and whether biometric signatures were transmitted to Meta's servers before removal
  • Whether the New York Times internal memo from February captured the full scope of Name Tag functionality or only a subset of what was implemented in the dormant code

Originally reported by petapixel.com

Read the original article →

Original headline: Meta Removes Hidden NameTag Facial Recognition Code From Ray-Ban Smart Glasses App After Wired Exposure, Calls Coverage 'Intellectually Dishonest'