Anthropic updates privacy policy to require government ID and facial biometric data from Claude users, effective July 8

Anthropic has updated its privacy policy to include identity and age verification provisions that take effect on July 8, 2026, requiring consumer Claude users — those on Free, Pro, and Max plans — to provide a government-issued photo ID, a live photo or video of their face, and facial geometry data the company acknowledges "may be considered biometric data in some jurisdictions."

What's new

The updated privacy policy, published June 8 and effective July 8, adds a "Verification Data" category to the personal information Anthropic may collect. When Anthropic requests identity verification, users must submit:

• A government-issued identity document (passport, driver's license, state ID, or national ID card)
• Name, date of birth, and ID number as printed on the document
• A photo or video of the user's face
• Facial geometry templates — which the policy notes "may be considered biometric data in some jurisdictions"
• The result of the verification process

The checks have been running in limited form since April 14, 2026, through third-party vendor Persona Identities, a San Francisco-based KYC platform. Images and facial data are stored on Persona's servers, not Anthropic's, under contractual data use limits. The policy applies to claude.ai Free, Pro, and Max plans; enterprise and commercial API customers are not affected.

Anthropicstates the biometric processing is consent-based. The company's identity verification support documentation is explicit: "We are not using your identity data to train our models. Verification data is used solely to confirm who you are and to meet our legal and safety obligations."

Context

Anthropicprevious privacy policy made no provision for biometric identity checks. The addition of government ID and facial geometry requirements puts its consumer platform in line with financial services and crypto exchanges that routinely run Know Your Customer (KYC) flows — a significant departure from how most large language model providers handle user identity.

Verification is triggered "when accessing certain capabilities, as part of our routine platform integrity checks, or other safety and compliance measures," according to Anthropic's help documentation. The company has not publicly specified which capabilities will require it after July 8. Accepted documents are physical and government-issued; photocopies, screenshots, and digital or mobile IDs are explicitly rejected.

The June 8 privacy policy update drew attention this week when users surfaced the change on Hacker News and Reddit, with 120+ upvotes on the HN thread.

Why it matters

Biometric facial templates are regulated as sensitive personal data under U.S. state laws — Illinois BIPA, Texas CUBI, and Washington's biometric privacy law — and as special-category data under GDPR in the EU. Anthropic's reliance on consent as the legal basis for biometric processing means users who decline verification could lose access to certain Claude capabilities.

For the broader AI industry, this is the first time a frontier AI lab has implemented a government-ID-plus-facial-biometric gate for consumer users at scale. It signals a shift in how top labs may approach access control for high-capability features: not just account flags or usage limits, but verified real-world identity.

Enterprise and API-focused customers face no immediate change. But the July 8 effective date marks a meaningful new term of service for the tens of millions of claude.ai consumer subscribers, setting a precedent other labs may face pressure to follow — or explicitly reject.