If you want into Anthropic's Claude club, you may have to show ID

What Anthropic is rolling out

Anthropic quietly updated its support pages this week to say some Claude features will require identity verification — rolled out “on a case-by-case basis,” the company says. Expect prompts at odd times: when accessing certain capabilities, as part of “routine platform integrity checks,” or for “other safety and compliance measures.” Short version: you might be asked for ID, and you might be asked suddenly. Why? “Identity verification helps us prevent abuse, enforce our usage policies, and comply with legal obligations,” Anthropic states, adding it won’t use ID data to train its models and will only collect the “minimum information required.”

Persona, paranoia and chain-of-trust worries

Anthropic says it’s using Persona to handle selfies and scans of identity documents, and that Persona is contractually limited in use and retention. But it has been reported that Persona’s involvement has spooked parts of the community — recall Discord’s aborted age-verification experiment that featured Persona and a security researcher’s claim that a Persona front end appeared on a government server. Persona denied any illicit scheme, but the uproar was enough to delay Discord’s plans, and now some Claude users on Reddit are allegedly threatening to cancel subscriptions.

There’s a sting here: verification rarely happens in a vacuum. Persona lists subprocessors — AWS, Confluent, Google, Stripe, Twilio, even other AI vendors — which means your selfie could touch a whole chain of providers. Anthropic says it can set retention limits and controls, but it hasn’t published the retention period for IDs processed by Persona. That omission matters. If any link in that chain is sloppy, your data might end up in places you never intended.

So should you worry? Depends on how much you value convenience versus control. Verification can curb abuse and unlock features. It can also make privacy-conscious users grind their teeth. This is the new trade-off of AI platforms: less friction for more oversight. And yes, the debate will only heat up as more companies — from chat apps to OS vendors — ask to prove who you are before they let the fun begin.

Sources: The Register