Anthropic begins rolling out identity verification for Claude users
Verification will be required for certain capabilities as part of platform integrity and compliance checks, with data handled by third-party partner Persona.
1 source · cross-referenced
- Anthropic is introducing identity verification for some Claude users, citing abuse prevention, policy enforcement, and legal compliance.
- Verification uses government-issued photo IDs and live selfies, processed by Persona Identities under contractual privacy and security controls.
- Anthropic states verification data is used only to confirm identity and is not shared or used for model training.
- Users may see prompts during routine platform integrity checks or when accessing certain capabilities.
Anthropic is rolling out identity verification for some Claude users to confirm who is using the platform, prevent abuse, enforce usage policies, and meet legal obligations. Verification may appear during routine platform integrity checks or when accessing certain capabilities.
The process requires a valid government-issued photo ID and a live selfie taken via phone or webcam. Verification typically takes under five minutes. Accepted IDs include passports, driver’s licenses, state or provincial IDs, and national identity cards, provided they are original, physical, clearly legible, undamaged, and include a photo of the user.
Anthropic selected Persona Identities as the verification partner, citing the strength of its technology, privacy controls, and security safeguards. Persona processes verification data on Anthropic’s behalf under contractual instructions, with all data encrypted in transit and at rest.
Anthropic states it does not use verification data to train its models and limits data collection to what is necessary to verify identity. Verification data is not shared with third parties except where legally required, and Persona is contractually restricted to using data only for verification, fraud prevention, and support.
Users may encounter verification prompts as part of routine checks or when accessing certain capabilities. If verification fails due to issues like blurry photos or expired IDs, users can retry within the flow or contact Anthropic for assistance.
Anthropic notes that accounts may still be banned for violations such as repeated policy breaches, unsupported location usage, or under-18 usage. Users can appeal account bans via a form linked on the verification page.
- Aug 5, 2026 · The Verge — AI
AI agents from OpenAI and Anthropic displayed deceptive behavior in UK safety tests
Trust74 - Aug 5, 2026 · Wired
Chinese researchers demonstrate self-replicating behavior in AI models under adversarial prompts
Trust74 - Aug 5, 2026 · OpenAI — News
OpenAI details third-party cybersecurity evaluations of its models and new safeguards
Trust72