TL;DR
- New Requirement: Anthropic now requires some Claude users to verify their identity with a government-issued photo ID and a live selfie.
- Access Gap: Chinese users without passports appear to be locked out, since China-issued national ID cards are not accepted.
- Privacy Concerns: Third-party provider Persona handles all ID images and biometric data, though Anthropic has not disclosed exact retention timeframes.
- User Backlash: Critics argue the checks hand a competitive advantage to ChatGPT and Gemini, which impose no such requirement.
To access certain Claude features, users must now hold up a government-issued photo ID and complete a live selfie on camera. Anthropic confirmed the new identity verification system on April 16, making Claude the first major AI chatbot to require such checks.
Checks are not being applied universally but are being introduced gradually in specific scenarios, including access to advanced capabilities and safety checks, routine platform integrity checks, and compliance measures. Backlash followed quickly, with some users arguing Anthropic has handed a competitive advantage to ChatGPT and Gemini by adding friction that rivals do not require.
How Claude’s ID Checks Work
Verification takes around five minutes and requires a camera-enabled device. Users must present a government-issued photo ID, such as a passport, driver’s license, state or provincial ID, or national identity card. Each document must be physically present, legible, undamaged, and feature the holder’s photograph.
Anthropic uses Persona Identities as its third-party verification partner to handle checks. Photocopies, screenshots, digital or mobile IDs, non-government credentials like student IDs or employee badges, and temporary paper IDs are not accepted.
Persona already handles age verification for other platforms. In December 2025, it began processing selfie and ID checks for users who were incorrectly placed in under-18 experiences on other services.
Despite the range of document types listed in Anthropic’s identity verification requirements, the restrictions create a notable accessibility gap.Chinese users without passports appear to be excluded, since passports are the only Chinese government ID the system recognizes. National ID cards issued by China are not on the accepted list, potentially locking out millions of users who lack passport access.
Privacy and Data Handling
Under the verification framework, Anthropic serves as the data controller, while Persona processes data under Anthropic’s instructions. ID images and selfies are handled by Persona, not on Anthropic’s own systems.
All data passing through Persona is encrypted in transit and at rest. Anthropic states that verification data will not be used to train AI models and will remain confidential except in response to valid legal processes.
Persona deletes verification data per Anthropic’s retention limits and applicable law, though Anthropic has not publicly specified exact timeframes for how long ID images and biometric data are stored before deletion.
For international users, submitting passport images and live facial data to a US-based company raises particular concern amid geopolitical tensions. Anthropic’s support documentation does not address how jurisdiction-specific privacy regulations apply to verification data collected from users outside the United States, leaving a gap in its public-facing privacy assurances.
User Backlash and Access Barriers
Since Anthropic’s confirmation, reports of unexpected account locks have unsettled existing users. Multiple outlets have framed the system as KYC-style verification, drawing comparisons to know-your-customer requirements in the financial industry. Such parallels highlight how unusual identity checks remain in the consumer AI space.
Anthropic is not the first AI company to face this friction. OpenAI encountered similar developer resistance in October 2025 when it introduced mandatory ID verification for API access, suggesting that identity checks may be a recurring challenge across the industry. Early user reactions also suggest older Claude accounts could become valuable assets as new users face growing hurdles to registration.
Whether other AI providers follow Anthropic’s lead remains an open question. Anthropic’s identity verification documentation confirms the current requirements but offers no timeline for expanding or revising the accepted document list.
For users in countries where accepted ID types are limited, the immediate concern is more concrete: without expanded document support, they face permanent exclusion from a leading AI assistant at a time when competitors impose no such barrier.
