A few weeks ago, an internal Meta memo leaked to the New York Times. The plan: add facial recognition to their Ray-Ban and Oakley AI glasses so the wearer can look at a stranger and see who they are. Meta calls it a "name tag" feature.
Stop and picture that for a second. Someone walks past you on the street, glances over for a beat too long, and now they know your name. Maybe your job. Maybe whatever else they decide to look up next. You did nothing. You wore a face.
We started FacePrivacy because the legal frameworks people assume will protect them (GDPR, CCPA, state biometric laws) only work if somebody actually files the requests. The right to be forgotten isn't automatic. It's a form. It's nine forms, in different formats, in different jurisdictions, that have to be re-filed every time your face gets re-scraped. Most people don't do it. The databases count on that.
This wasn't an isolated week
The Meta news broke alongside a few other things you should know about:
Amazon Ring's "Familiar Faces" rollout
Senator Markey's office disclosed last week that Ring's new facial recognition feature scans every face in front of every Ring doorbell. Device owners get privacy controls. Delivery drivers, kids walking past, and neighbors don't.
The OkCupid / Clarifai settlement
The FTC just confirmed Match Group shared roughly three million OkCupid users' photos with the facial recognition firm Clarifai back in 2014 without consent. Clarifai used them as training data. Eleven years later, the company is finally agreeing to delete the models.
Kimberlee Williams
Wrongfully arrested by Maryland police based on a bad facial recognition match. Six months in jail. The fourteenth case of this kind that's been made public.
None of these stories are about FacePrivacy. We're not on any of them. But they're all, in different ways, the same story: your face is being used in systems you didn't consent to, by people who don't know you, with consequences you can't predict and can't easily reverse.
What we actually do (and don't)
We won't tell you we can stop Meta from shipping name tag glasses. We can't. We're not going to pretend a subscription to our service will keep your face out of every camera in every airport. It won't.
What we do is narrower than that, and more useful. We submit and re-submit removal requests on your behalf to the facial recognition databases that honor opt-outs:
We use the legal rights you already have under GDPR and CCPA. When your face reappears in those databases (which it does, because new scrapes keep happening), we file again. And we offer a way to alter the facial embeddings on photos you upload going forward, so future training runs and search engines don't get a clean match. The pictures still look like you. The math underneath stops cooperating.
That's it. No magic. Just paperwork done at scale, and one technical trick that buys you a little ground on new uploads.
Why we think this year matters
The arguments for waiting are getting weaker. Regulation is real, but slow. The EU AI Act has teeth, but enforcement scales with budgets and political will. State laws are a patchwork. Federal action in the US, on a timeline that would actually matter, isn't happening.
Meanwhile, every quarter the consumer rollout of facial recognition gets one notch more aggressive. Last year it was Faceboarding at Linate. This year it's Ring on every porch and Meta on every bridge of every nose at brunch. Whatever your mental model of "places where I get scanned" was eighteen months ago, it's already out of date.
If you're reading this on our site, you probably don't need the convincing part. You need the practical part.
So: search yourself on PimEyes and Precheck.ai first, before doing anything else. See what's there. Then decide how much paperwork you want to do alone, and how much you want us to do for you.
Start Removal →We'll be here either way. ♡