Claude now requires ID verification for some features

Key Takeaways

• Anthropic is requiring government-issued photo ID verification for certain Claude capabilities and safety compliance
• Persona Identities handles verification; images are not stored on Anthropic's systems or used for AI training
• Failed verifications can be retried; accounts may be banned for policy violations, underage use, or unsupported locations

Anthropic is rolling out identity verification for Claude users. If you try to access certain features, you may now see a prompt asking you to submit a government-issued photo ID and a live selfie. The company frames this as a safety measure: knowing who uses powerful AI, it argues, is the first step in preventing misuse.

This is a notable shift for consumer AI. While enterprise AI tools routinely verify customers through contracts and procurement processes, asking individual users to hand over a passport or driver's license to chat with an AI assistant is new territory. Anthropic's move signals that as AI capabilities grow, so do the compliance expectations around who gets access to them.

What does Claude identity verification require?

The process is straightforward, taking under five minutes in most cases. You'll need a valid, physical government-issued photo ID. Passports, driver's licenses, state or provincial ID cards, and national identity cards from most countries are accepted.

You'll also need a phone or computer with a camera. Anthropic may ask for a live selfie to match against your ID photo. No photocopies, screenshots, or digital mobile IDs. No student cards, employee badges, or temporary paper documents.

The verification partner is Persona Identities, a company that handles billions of identity checks for platforms including Square and various fintech firms. Persona collects and holds your ID images; Anthropic can access verification records through Persona's platform but doesn't copy or store the images itself.

How is verification data protected?

Anthropic is explicit about several boundaries. Your ID and selfie are not used to train AI models. The company says it collects only the minimum information needed to confirm your identity. Verification data stays between you, Persona, and Anthropic unless law enforcement presents valid legal process.

Persona operates under contractual limits: they can use your data only for verification services and fraud prevention improvements. All data is encrypted in transit and at rest. Retention periods follow limits set by Anthropic and applicable law, though the company hasn't published exact timeframes.

The privacy architecture here is intentional. By keeping images off Anthropic's own systems, the company reduces its exposure to potential data breaches. If Anthropic's servers were compromised, attackers wouldn't find a database of user passports.

When will you see a verification prompt?

Anthropic lists several triggers: accessing certain capabilities, routine platform integrity checks, and unspecified "safety and compliance measures." The company hasn't detailed exactly which features require verification. Based on Anthropic's Responsible Scaling Policy and recent product launches, advanced capabilities like computer use, agentic workflows, and certain coding tools are likely candidates.

Age assurance is another factor. Anthropic separately maintains age verification systems, and accounts flagged for potential under-18 usage face bans. Identity verification likely serves as a backstop for these checks.

What if verification fails or your account is banned?

Failed verifications happen for mundane reasons: blurry photos, unreadable documents, expired IDs, or technical glitches. You get multiple attempts within the verification flow. Better lighting and a different ID often solve the problem.

Account bans are a different matter. Anthropic lists several reasons: repeated usage policy violations, account creation from an unsupported location, Terms of Service violations, and underage use. If you're banned, you can appeal by logging into claude.ai with the banned account and submitting an appeal form. The Safeguards team reviews these cases.

The appeal process requires you to log in with the banned account. If you've lost access entirely, you'll need to contact Anthropic directly through their support form.

Why is Anthropic doing this now?

The stated reasons are abuse prevention, usage policy enforcement, and legal compliance. The unstated context is that Claude's capabilities are expanding rapidly. Agentic features that can browse the web, execute code, and operate computers autonomously present different risks than a basic chatbot.

Regulatory pressure is building too. Governments worldwide are considering AI regulations that may require providers to know their users. Implementing KYC-style verification now positions Anthropic ahead of potential mandates rather than scrambling to retrofit compliance later.

There's also competitive positioning. Anthropic markets itself as the safety-focused AI company. Identity verification is consistent with that brand, even if it adds friction that competitors may not impose.

“Being responsible with powerful technology starts with knowing who is using it.”

— Anthropic's identity verification announcement

The friction vs. security tradeoff

Not everyone will welcome this change. Hacker News discussions around AI verification typically surface predictable objections: privacy concerns about submitting government ID to a private company, questions about whether determined bad actors will simply find workarounds, and frustration that legitimate users bear the compliance burden.

These concerns aren't unreasonable. Identity verification creates a paper trail that some users may find uncomfortable, even if they have nothing to hide. And verification alone won't stop sophisticated abuse. Someone willing to use Claude maliciously could potentially obtain fraudulent credentials.

But verification does raise the cost and complexity of abuse at scale. A bad actor who previously could spin up dozens of accounts with disposable email addresses now needs multiple valid IDs. That's a real barrier, even if not an absolute one.

Frequently Asked Questions

Does Claude identity verification use my data to train AI models?

No. Anthropic explicitly states that verification data is used only to confirm identity and meet legal and safety obligations, not for model training.

Which government IDs does Claude accept for verification?

Physical passports, driver's licenses, state or provincial ID cards, and national identity cards from most countries. Digital mobile IDs, photocopies, and temporary paper IDs are not accepted.

Where is my ID photo stored after Claude verification?

Persona Identities, Anthropic's verification partner, holds the images. Anthropic can access records through Persona's platform but doesn't store the images on its own systems.

Can I appeal a Claude account ban after identity verification?

Yes. Log into claude.ai with your banned account and submit the appeal form. Anthropic's Safeguards team reviews appeals.

How long does Claude identity verification take?

Typically under five minutes. You'll need your physical ID and a device with a camera for a potential live selfie.

Source: Hacker News: Best