Smart glasses are having a weird — and, in some ways, exciting — moment. After years of awkward prototypes and products that screamed “tech demo,” we’ve finally landed on a formula that actually works: a frame that looks like regular Ray-Bans, with a camera, microphones, speakers, and a layer of AI that can (more or less) understand what you’re looking at. From a technology standpoint, the timing is right.
The problem is that Meta seems tempted to take a much more delicate leap: facial recognition on its glasses. And even as a hypothetical, that changes what the product is. According to reports circulating in the press, Meta has been evaluating an internal feature nicknamed “Name Tag,” designed to identify people through the built-in assistant, potentially drawing from public profiles across Meta’s platforms.
At that point, we’re no longer talking about a handy gadget. We’re talking about a decision that determines whether the world becomes, by default, scannable.
Why these glasses are different (and therefore more sensitive)
The usual comeback is predictable: “Your phone can already record video.” True — but a smartphone involves obvious gestures. You take it out, point it, put it away. Glasses live on your face. Capture becomes frictionless, nearly invisible, especially when the product is intentionally designed to blend in as a normal pair of frames.
In theory there’s a social guardrail: a tiny LED light that turns on when you’re recording. In practice, that’s where things get messy. Because that light can be hard to notice in real life, and because we’ve already seen — again and again — how easy it is to undermine “signals” like that with cheap covers or small modifications. Meta, for its part, leans heavily on the idea of responsibility and keeps pointing to the LED as the main safeguard. In its official guidance, the company also recommends turning the glasses off in sensitive places (doctor’s offices, locker rooms, public bathrooms, schools, places of worship).
The core issue remains: the system relies a lot on good faith… and a tiny dot of light.
“Name Tag”: a simple idea, a Pandora’s box
From a product perspective, the temptation is obvious: you pass someone and the glasses whisper their name. In certain contexts, that could even be genuinely useful — and the conversation often brings up accessibility, for example for people with low vision or for anyone who struggles with face recognition.
But shift the slider by one millimeter and everything changes: identifying a stranger because they have a public profile, linking a face to a social account, turning public space into a roaming directory. This isn’t sci-fi. We’ve already seen how quickly you can end up in doxxing territory when you combine wearable cameras with public databases. If Meta bakes it into a mainstream product, the fear isn’t “is it possible?” — it’s “does it become normal?”
And once something becomes normal, it’s very hard to walk it back.
Meta’s reputation makes this far more combustible
Even if another company launched consumer facial recognition, it would still be an earthquake. With Meta, it’s even more complicated because trust isn’t exactly its signature strength. The harshest critique isn’t “Meta invented evil,” but rather “Meta has already shown what happens when growth beats caution.”
There’s also a concrete detail that matters: in recent years Meta has pushed harder on AI integration across its products, and default settings are powerful. If a feature is enabled by default, it becomes “the standard” in the real world — even for people who never stopped to think about it. This is where the debate stops being technical and becomes social: everyday life isn’t a lab. It’s full of people who didn’t ask to be part of the experiment.
The law isn’t ready, and biometrics change the rules
In Europe, biometrics already sit in a special category: when you process biometric data to uniquely identify a person, you’re stepping into highly sensitive territory under GDPR. And with the EU’s AI Act, the idea of building or expanding facial recognition systems through large-scale harvesting of public images has been placed under even stricter scrutiny.
More than anything, facial recognition destroys the ambiguity of “I’m just recording a scene.” It turns an image into a query: “Who is this person?” Culturally, we’re not ready to live in a world where that question is always one voice command away. Technically, we’re almost there already.
FAQ
Do Ray-Ban Meta glasses already do facial recognition?
Based on what’s been communicated so far, not as a public, consumer-facing feature. The current conversation revolves around an evaluated concept (“Name Tag”) described in press reporting rather than a confirmed mainstream launch.
What is “Name Tag,” in plain terms?
As reported, it would be a feature that lets the AI assistant identify people — potentially by drawing from linked contacts and/or publicly available profile information within Meta’s ecosystem.
Is the recording LED a real safeguard?
It’s a signal, not a guarantee. If it’s hard to notice — or easy to bypass — it loses much of its value as a social consent mechanism.
Why does facial recognition change the conversation so much?
Because it shifts the act from “recording” to “identifying.” And biometric identification has huge implications: privacy, safety, stalking risks, profiling, plus a far stricter legal framework (especially in the EU).
Would this be legal in Europe?
It depends heavily on implementation, legal basis, limitations, and whether the system processes biometric data for unique identification. If it does, the compliance bar — and the risk — rises sharply.
Final Toughts
What’s almost tragic here is that Meta might have finally found the right smart-glasses format: useful, wearable, not embarrassing. And yet facial recognition could reset the story back to the Google Glass era — only with far more power and far more social infrastructure behind it. My take is straightforward: if Meta truly pushes “Name Tag,” it won’t be “just another feature.” It’ll be a cultural rupture. And honestly, Meta isn’t the company I’d choose to manage a rupture like that with restraint.
