Meta’s smart glasses draw privacy concerns after code discovery

Hidden code in your pocket

It was reported that Meta had embedded unreleased facial recognition code, internally called NameTag, in the Meta AI app used with Ray-Ban and Oakley smart glasses. The app had been installed on more than 50 million phones.

Meta later removed the main NameTag components from the latest Meta AI app, according to a follow-up WIRED code review. The feature was not publicly enabled, but it included models that could detect faces, crop them, and encode them into biometric signatures.

What is a faceprint, exactly?

You have heard of fingerprints, right? A faceprint works the same way for your face. The system maps your unique features like the distance between your eyes and the shape of your jawline, then turns all that data into a long string of numbers. 

If Meta turns on NameTag for your Meta Smart Glasses, every face the camera sees gets turned into one of these digital codes. Then the glasses check that code against a database stored right on your phone. If it finds a match, the wearer gets a notification telling them who you are.

Meta’s billion-face promise

In November 2021, Meta announced it would shut down Facebook’s Face Recognition system and delete more than one billion facial recognition templates. The company said facial recognition could be useful in some cases, but needed to be weighed against broader societal concerns and unclear regulations.

It was reported in June 2026 that Meta had built unreleased NameTag facial recognition components into the Meta AI app for smart glasses before removing the main code after the report. The reporting has renewed scrutiny from privacy advocates because Meta had previously scaled back broad facial recognition on Facebook.

Meta says relax, it’s nothing

Meta said the feature was exploratory and that no final decision had been made. Meta also said it was not building a central face database, while WIRED reported that the code it reviewed compared faceprints against a database stored on the user’s device.

Privacy advocates argue that local storage does not resolve the consent problem for people captured by someone else’s glasses. WIRED also reported that an outside researcher tested the pipeline by adding a faceprint of French philosopher Michel Foucault and triggering a Person recognized notification.

Little-known fact: Security researchers independently tested the NameTag code by adding a faceprint of French philosopher Michel Foucault. When they triggered the system, the app produced a Person recognized notification, proving the pipeline actually works.

What are the privacy groups saying?

Privacy groups say Meta’s explanations do not answer the core civil liberties concerns. WIRED quoted EFF senior public interest technologist Cooper Quintin saying the feature seemed nearly ready and that Meta had created the capacity to turn customers into a distributed surveillance machine.

A coalition led by the ACLU and joined by 75 organizations urged Meta to cancel the planned facial recognition feature for Ray-Ban and Oakley glasses. The groups warned that real-time identification through smart glasses could create risks for stalking, harassment, doxxing, and government misuse.

Little-known fact: Meta has already paid roughly $2 billion in settlements for violating biometric privacy laws, including $650 million to Illinois users and $1.4 billion to Texas

A creepy internal memo leaked

Here is where this story gets really troubling. The New York Times obtained an internal Meta memo from May 2025. In that document, company leaders reportedly wrote that they planned to launch NameTag during a dynamic political environment. 

Why would politics matter? Because Meta believed civil rights groups would be too distracted by other crises to fight back. The memo said that many civil society groups that we would expect to attack us would have their resources focused on other concerns. That sounds like a company trying to sneak something through.

Bystanders never get a say

The central concern is bystander consent. When people walk down a street, sit in a coffee shop, or stand in a grocery line, they have not agreed to have their faces turned into biometric identifiers by someone else’s wearable camera.

Privacy and civil liberties groups say facial-recognition-enabled smart glasses would be different from posting your own photo online. The wearer could control the device, while bystanders may not know they were captured or have a meaningful way to opt out.

The recording light can be killed

Meta includes a capture LED that signals when Ray-Ban Meta glasses are taking photos, recording video, or going live. Meta’s privacy guidance says the glasses notify the wearer to clear the LED before capturing if the light is covered.

Reports have shown that modders can physically disable the LED while leaving recording functions intact, with services advertised for roughly $50 to $100. Without the normal indicator light, people nearby may have a much harder time knowing whether the glasses are recording.

A nightmare for abuse survivors

Privacy experts warn that facial recognition on glasses could create serious risks for people escaping dangerous situations. An abusive ex-partner or stalker could use facial-recognition-enabled eyewear to identify someone in a crowd, at a bus stop, or outside a grocery store.

The ACLU-led coalition asked Meta to disclose any known reports of its wearables being used for stalking, harassment, or domestic and sexual violence. The groups also asked whether Meta has had discussions with law enforcement agencies, including ICE and CBP, about wearable technology.

Your face cannot be reset

Biometric data carries a different risk than a password or credit card number. A password can be changed, and a payment card can be replaced, but a person cannot simply reset their face.

That is why privacy experts treat faceprints as especially sensitive. If biometric templates are exposed or misused, the risk can last much longer than a typical credential leak because the underlying physical trait is permanent.

Other companies are watching

Meta is not alone in building smart glasses. Google said at I/O 2026 that intelligent eyewear with Gemini is coming this fall through Android XR partnerships with Samsung, Gentle Monster, and Warby Parker. Reuters has also reported that Apple is developing smart glasses.

That broader competition is why privacy advocates want rules before facial recognition becomes a common wearable feature. Their concern is that decisions made by major platforms now could shape how smart glasses handle biometric data, consent, and public recording across the industry.

How does this break state laws?

Illinois, Texas, and several other states have strong biometric privacy laws. These laws generally require companies to get written permission before collecting face geometry. But how can a stranger wearing smart glasses get your permission before scanning your face in public? They cannot.

That legal problem is huge. A feature that might be legal in one state could get the company sued in another. Some experts think this is why NameTag exists in the code but has not been switched on yet. The legal risks are just too high, at least for now.

If you’re curious how privacy concerns fit into the broader changes happening inside the company, check out how Meta’s job cuts spread across multiple divisions during restructuring for another look at the challenges Meta is navigating.

What can you do right now?

You do not need to throw away smart glasses to follow this story closely. The key point is that NameTag was not publicly enabled, and WIRED later reported that Meta removed the main facial recognition components from the latest Meta AI app after public scrutiny.

People concerned about the issue can review app permissions, watch for future Meta announcements, and follow privacy groups that are pushing the company to abandon facial recognition in smart glasses. The ACLU is also running a campaign asking consumers to contact Meta directly about the feature.

If you’re curious about how Meta’s broader push into AI and infrastructure is expanding behind the scenes, check out how Nebius and Meta struck a massive $27 billion AI infrastructure deal for a closer look at what’s powering these new technologies.

This slideshow was made with AI assistance and human editing.

Don’t forget to follow us for more exclusive content right here on MSN.

Read More From This Brand:

• Meta and Arm strike a partnership to expand global AI capabilities
• IREN scores a $9.7B partnership with Microsoft to fuel global AI expansion
• Adobe expands AI reach through new Google partnership and YouTube integration