Anthropic's ID Verification Goes Live: Claude Users Now Face Government ID and Selfie Checks

Anthropic's identity verification regime is now live. As of July 8, Claude Free, Pro and Max subscribers can be required to submit a government-issued photo ID, a live selfie and a facial geometry scan through identity provider Persona before accessing the company's most advanced capabilities — the most sweeping identity requirement yet imposed by a frontier AI lab.

What Changed on July 8

Two policies converged this week. Anthropic's updated privacy policy took effect, formalizing the collection of identity documents, personal information and what the company acknowledges "may be considered biometric data in some jurisdictions." Simultaneously, Fable 5 — restored globally on July 1 after its three-week export-control suspension — shifted to usage-credit billing at $10 per million input tokens and $50 per million output tokens.

Verification is one-time per account and handled by Persona, which stores the data rather than Anthropic. The Claude API, Team and Enterprise plans are exempt; the checks target consumer tiers, where Anthropic says it has the least visibility into who is actually using its models.

Born From the Fable Shutdown

The system is a direct answer to the crisis that took Fable 5 and Mythos 5 offline globally in June: under US export controls, Anthropic had no way to selectively block foreign nationals from its most capable models at scale. Identity verification closes that compliance gap — and the company told senators in June that it had already disrupted 28.8 million distillation attacks routed through some 25,000 fraudulent accounts, many linked to relay services used to funnel Claude access into China.

Privacy Pushback

Privacy advocates have objected to the collection of facial geometry — data regulated under biometric privacy laws such as Illinois' BIPA — and to the precedent of tying frontier AI access to government identity. For users across Asia, the practical effect varies by market: verified US users get immediate access, while availability in other countries depends on how each jurisdiction maps onto Anthropic's export-control obligations.

With OpenAI's GPT-5.6 launching publicly today under its own government-coordinated framework, the era of anonymous access to the world's most capable AI models appears to be ending — one verified selfie at a time.