Meta silently pushed face recognition code for its smart glasses to millions of Android and iOS devices. According to Wired, the update included functionality for a feature called "Nametag," which would let Ray-Ban Meta smart glasses identify people by their faces and display their names and social media profiles.
Here's what makes this significant: Meta distributed the capability before announcing the feature, before getting meaningful public input, and before clarifying what safeguards would exist. The code is sitting on phones right now, waiting to be activated.
From a technical perspective, this is straightforward. The smart glasses capture video, send frames to the phone app, which runs facial recognition against Meta's database of profile photos, then returns identity information. It's the same technology that powers photo tagging, just applied in real-time to everyone you look at.
The privacy implications should be obvious. Walking around with face recognition glasses turns every interaction into a database query. You see someone at a coffee shop, glance at them, and instantly know their name, where they work, their relationship status—anything public on their Facebook profile. They have no idea this is happening.
Meta's defense will likely emphasize consent and control. Users opt into making their profile photos searchable. People choose what information is public. Technically true. But there's a difference between information being publicly available and someone being able to access it instantly by looking at you.
This isn't hypothetical technology. Two Harvard students built a similar system using Ray-Ban Meta glasses in 2024, demonstrating exactly this capability. They called it I-XRAY and showed how easily it could identify strangers, find their addresses, and more. Meta didn't announce Nametag features in response—they quietly embedded the code anyway.
The regulatory landscape is scattered. Some jurisdictions have strict biometric privacy laws that might restrict this use case. Others have no relevant regulations. Meta presumably included the code now to enable faster rollout when they decide to launch, rather than waiting for app updates to propagate.
From a product perspective, I understand the appeal. Imagine never forgetting someone's name at a conference. Sounds useful. But that use case requires normalizing constant facial recognition surveillance, which is a massive trade-off.
The technology is impressive. The question is whether anyone needs AR-enabled facial recognition badly enough to accept a world where every glance is a potential privacy invasion.
What bothers me most is the deployment approach. Distribute first, announce later, deal with backlash if it happens. That's not how you build trust with technology that fundamentally changes social norms around privacy and anonymity.
Meta hasn't officially enabled Nametag yet, which means there's still time for public debate about whether this feature should exist at all. But the code is already out there, on millions of devices, waiting for a server-side flip of a switch.
For users who don't want their faces in Meta's recognition database, the options are limited. You can make your profile private, but that doesn't necessarily opt you out of recognition—just limits what information is displayed. The more comprehensive option is not having a Facebook account, which increasingly feels like the only real privacy choice Meta offers.
