ANTHROPIC WANTS YOUR DRIVER'S LICENSE FOR SOME CLAUDE FEATURES

Anthropic is rolling out identity verification for a handful of use cases on Claude, asking some users to upload a government-issued photo ID and a selfie before they can access certain capabilities. The company says the program is small and targeted, but the move places it alongside an increasingly divided set of AI companies: those that verify users and those that do not.

The process runs through Persona Identities, a third-party verification provider. Anthropic says it chose Persona on the strength of its privacy controls and security safeguards. The actual verification takes under five minutes. Users need an original, physical government-issued photo ID from most countries. Driver's licenses and state or provincial ID cards are common examples. Anthropic does not accept photocopies, screenshots, scans, photos of a photo, digital or mobile IDs, or non-government-issued cards such as student IDs or employee badges. The ID must be clearly legible, undamaged, and include a photo of the user.

you may like

ANTHROPIC LAUNCHES CLAUDE REFLECT

UNO REVERSE CARD: ALIBABA BANS CLAUDE CODE

ANTHROPIC ACCUSES ALIBABA OF STEALING CLAUDE SECRETS

The rollout is not universal. Users might see a verification prompt. The language is broad enough to cover a lot of scenarios, and Anthropic is not specifying exactly which features trigger the requirement. The company frames the move as a security and accountability measure.

WHO HAS TO VERIFY AND WHY

Anthropic's stated rationale is straightforward, the company says. Identity verification helps prevent abuse, enforce usage policies, and comply with legal obligations. The company says it is not collecting more than it needs and asks for the minimum information required to verify identity.

The policy has teeth. As part of the safety process, Anthropic may ban an account for repeated violations of its Usage Policy or account creation from an unsupported location. If an account gets wrongly suspended, users can log in to claude.ai with the banned account and fill out an appeal form. The Safeguards team investigates why the account was disabled.

But failing verification can prevent access to the capabilities that require it, though Anthropic provides multiple attempts within the flow and a recourse form if problems persist. Most failures are resolved by retaking the photo in better lighting or using a different ID.

THE DATA FLOW IS WORTH UNDERSTANDING

The ID and selfie are collected and held by Persona, not stored on Anthropic's own systems. Anthropic says it can access verification records through Persona's platform when needed, for example, to review an appeal, but it does not copy or store the images itself. Anthropic is the data controller and sets the rules for how the data is used and how long it is kept.

Persona is contractually limited in what it can do with the data: only to provide and support verification and to improve its ability to prevent fraud. The company says Persona must protect the data with industry-standard security controls and delete it according to retention limits Anthropic sets. All data passing between Anthropic and Persona is encrypted in transit and at rest.

Anthropic explicitly states that identity data is not used to train the company's models. Verification data is used solely to confirm who someone is and to meet legal and safety obligations. The data stays between the user, Persona, and Anthropic, except where the company is legally required to respond to valid legal processes. It is never shared with third parties for marketing or advertising.

WHAT FAILURE LOOKS LIKE

Verification can fail for several reasons: a blurry photo, an unreadable document, an expired ID, or a technical issue. Users get multiple attempts within the verification flow. If they exhaust their attempts, they can reach out to Anthropic via a web form. The appeals process requires the user to be logged into the banned account on claude.ai to access the appeal form itself.

For users in unsupported locations, account creation from such a location can itself be grounds for a ban. If banned, they can appeal by logging into claude.ai with the banned account.

what to read next

ANTHROPIC’S NEW FABLE 5 MIGHT BE SILETNTLY NERFING YOU

1PASSWORD LETS CLAUDE USE YOUR PASSWORD WITHOUT EVER SHOWING THEM

ANTHROPIC IPO COULD ECLIPSE EVERY SOFTWARE COMPANY IN HISTORY

THE LANGUAGE IS CAREFUL

Anthropic is positioning the identity verification as a limited rollout for specific use cases. The language is careful, the privacy disclosures are thorough, and the company has taken steps to limit its own exposure to the data. The cost is user anonymity. For a tool that competes on being helpful and conversational, requiring a government ID to access certain features is a friction point that will shape who uses Claude and for what. The companies that figure out how to balance safety and privacy without requiring a driver's license will have a real advantage. For now, Anthropic is betting that the tradeoff is worth it.