This week, Anthropic quietly introduced identity verification requirements for its AI platform, Claude, requesting some users provide a government-issued photo ID and possibly a live selfie—a measure not demanded by competitors.
“We are implementing identity verification in certain scenarios to ensure our platform’s integrity and safety,” the company stated. “This data is solely used for confirming user identities.”
After millions migrated from OpenAI to Anthropic in February following concerns over OpenAI’s Pentagon contract, users lauded Anthropic as a privacy haven. However, this new policy may require some of these users to prepare documents if they wish to continue using Claude. Critics argue this decision is voluntary rather than mandated by regulatory bodies.
Ryan SΞan Adams (@RyanSAdams) criticized on April 15, 2026: “New users of Claude need government ID for subscription—unlike ChatGPT and Gemini. Anthropic is giving its rivals an unintended advantage.”
Anthropic has chosen Persona Identities as its verification partner, necessitating a valid passport, driver’s license, or national identity card, excluding photocopies and mobile IDs. Verification may also entail taking a live selfie.
The policy applies to specific situations involving potential fraudulent behavior, according to Anthropic’s spokesperson who spoke with Decrypt. The company assures that user data is encrypted and not used for model training or shared with third parties for marketing purposes—a promise they’ve consistently made since their early commercial policies.
Despite these assurances, the history of breaches like Discord’s 2025 incident involving government IDs raises concerns about third-party data custody. Anthropic has also been tightening identity controls, previously implementing classifiers to detect self-identified minors and banning accounts from unserved regions—affecting particularly Chinese users who use intermediaries.
The article was updated post-publication with comments from Anthropic.