TL;DR: On June 4, 2026, the Electronic Frontier Foundation published “Move Fast, Surveil Things,” confirming through static code analysis that Meta has already deployed facial recognition code to millions of Ray-Ban smart glasses. The internal “Name Tag” feature converts every face in view into a 2,048-number faceprint array and matches it against a stored database. An EFF researcher activated the feature in debug mode and watched the glasses detect a known face in real time. Meta’s spokesperson says this is just “evidence of exploration.” Exploration doesn’t ship to millions of devices. This is staging for launch, and internal documents show Meta planned to flip the switch while privacy groups were too busy to fight back.
What EFF Found in the Code
Cooper Quintin, a security researcher with EFF’s Threat Lab, didn’t need a leak or a whistleblower. He needed a decompiler [1].
Through static analysis of Meta’s app, Quintin confirmed that facial recognition code is present, active, and functional on the software shipped to millions of Ray-Ban Meta glasses owners. The feature, internally called “Name Tag,” works like this:
- The glasses’ cameras capture a face
- The software converts it into a faceprint: a series of 2,048 numbers that uniquely map the positioning of a person’s facial features
- That faceprint gets compared against every existing faceprint in the user’s database
- Match found? The glasses tell you who you’re looking at
Quintin went further. He connected a phone running Meta’s app to a computer in debug mode and manually activated the feature. The glasses subsequently detected a previously scanned face when it came into view [1].
This isn’t a prototype on a whiteboard. It’s working facial recognition running on consumer hardware. The only thing between you and a stranger identifying you on the street is a feature flag Meta hasn’t flipped yet.
Wired Got There Too
EFF’s findings landed the same week as a Wired investigation that independently found the same code [2]. Wired reported that past versions of Meta’s app included interface elements like a “Connections” menu suggesting users “remember the people you met,” the consumer-facing skin for what the code does underneath.
Meta spokesperson Ryan Daniels responded with the kind of non-denial that says everything: “We’ve said before we’re exploring these types of features, and what you’re seeing is just evidence of that exploration.” He added: “We are not building a central face database” [2].
That second claim needs context. You don’t need a central database when every user builds their own local faceprint collection. The faces still get scanned. The biometric data still gets created. Meta just outsources the storage to your phone, and the legal liability to you.
$7 Billion in Settlements. And They’re Doing It Again.
Meta has a specific, documented, expensive history with facial recognition:
- 2010: Facebook rolls out facial recognition for photo tagging. Nobody is asked for consent.
- 2019: $5 billion FTC settlement over privacy violations, including deceptive biometric practices [3].
- March 2021: $650 million settlement under Illinois’ Biometric Information Privacy Act, one of the largest privacy payouts in history [3].
- November 2021: Meta kills facial recognition on Facebook. Says it will delete over 1 billion face templates. Gets credit for doing the right thing.
- July 2024: $1.4 billion Texas biometric privacy settlement for the same system they supposedly killed [3].
- Late 2024: Meta quietly reintroduces facial recognition on Instagram and Facebook as a “scam detection tool” [2].
- June 2026: EFF confirms facial recognition code shipped to millions of glasses.
The pattern is: deploy first, apologize later, pay the fine, do it again somewhere else. Face scanning on photos became face scanning on glasses. “Scam detection” became “Name Tag.” The technology never went away. It just changed addresses.
The Launch Strategy: Wait Until Nobody’s Watching
An internal Meta document from May 2025, first reported by The New York Times in February 2026, laid out the company’s plan in plain English [4]:
“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.”
That’s not a privacy policy. That’s a war plan. Meta knows exactly who will object (the ACLU, the EFF, the CDT, state attorneys general), and their strategy is to time the launch for when those groups are stretched thin fighting government surveillance, AI regulation, and everything else.
EFF’s title says it all: “Move Fast, Surveil Things.”
76 Organizations Say No
Meta’s bet that privacy groups would be too distracted hasn’t panned out. In April 2026, the ACLU organized a coalition letter signed by 75 organizations (domestic violence advocates, worker rights groups, reproductive freedom organizations, immigrant rights advocates, and civil liberties groups) demanding Meta “immediately halt and publicly disavow” plans to add facial recognition to its glasses [5].
The coalition includes the Electronic Privacy Information Center, Fight for the Future, the National Organization for Women, and dozens of state-level civil liberties chapters. Their argument is simple: glasses that identify strangers on sight are a threat at protests, at medical clinics, at domestic violence shelters, and on the street [5].
Meanwhile, Texas Attorney General Ken Paxton launched a formal investigation in May 2026, issuing a Civil Investigative Demand to Meta over the glasses’ privacy practices. The investigation targets Meta’s data collection, the always-on camera mode, and the facial recognition plans specifically [6]. Paxton’s office also flagged reports that Meta subcontractor workers in Kenya had access to intimate footage from users’ glasses, including bathroom visits [6].
What You Can Do
Assume You’re Being Scanned
If someone near you is wearing Ray-Ban Meta glasses, treat it like a running camera, because it is. The small LED indicator that’s supposed to signal recording can be covered with tape, and Meta’s “always enabled” mode processes video without activating it [6]. The Nearby Glasses app can detect Meta Ray-Bans via Bluetooth.
Lock Down Your Meta Profiles
Name Tag matches faces against Meta platform data. If you have Facebook or Instagram accounts, set your profile photo to something that isn’t your face. Make your accounts private. Better yet, consider whether you need them at all. Every photo of your face on Meta’s platforms is training data for identifying you later.
Support the Legal Pushback
The EFF, ACLU, and EPIC are fighting this on legal, legislative, and technical fronts. Texas’s investigation proves state-level enforcement works: Meta paid $1.4 billion last time Texas came knocking over biometrics. Push for biometric privacy laws in your state. Illinois and Texas showed the way.
Don’t Buy the Glasses
Every pair sold expands the surveillance network. Meta has shipped millions already. Each one is a camera pointed at everyone else. If you own a pair, understand that the facial recognition code is on your device right now, and Meta can activate it with a software update whenever they decide the political moment is right.
The Gap Between “Exploring” and “Deployed”
Meta wants you to believe there’s a wide gap between having facial recognition code in an app and actually using it. But EFF just proved that gap is one debug-mode toggle. The faceprint arrays are defined. The matching logic is built. The glasses already process faces.
A company that paid $7 billion in biometric privacy settlements, whose internal documents describe timing a surveillance feature launch for maximum political cover, has shipped that feature’s code to millions of devices. The question isn’t whether Meta will turn it on. It’s whether anyone will stop them before they do.
References
- EFF: Move Fast, Surveil Things (June 4, 2026)
- Engadget: Wired Found Code for an Unreleased Facial Recognition Feature in Meta’s AI App (June 4, 2026)
- EFF: Seven Billion Reasons Facebook Should Abandon Its Face Recognition Plans (February 13, 2026)
- TechCrunch: Meta Plans to Add Facial Recognition to Its Smart Glasses, Report Claims (February 13, 2026)
- ACLU: 75 Organizations Sound Alarm on Meta’s Plans to Add Facial Recognition to Ray-Ban and Oakley Eyeglasses (April 2026)
- Texas Attorney General: Ken Paxton Launches Investigation Into Meta Glasses (May 2026)