Detailed Analysis
A user of Anthropic's Claude platform has reported being caught in an unresolvable loop after their account was suspended on suspicion of belonging to a minor, with Anthropic's third-party age verification partner Yoti failing to process any of their submitted verification attempts. The user describes repeated "unknown error" responses and "photo may still be unclear" messages across multiple methods — including facial age estimation and government ID submission — despite careful adherence to Yoti's instructions, use of multiple devices and browsers, and submission of what they describe as properly aligned, high-quality images. The account in question is used across multiple projects, making the suspension a significant functional disruption rather than a minor inconvenience.
A critical complicating factor in the user's case appears to be geographic. Yoti's dedicated mobile application is not available in the user's country, which likely constrains the available verification pathways and may explain the persistent technical failures. Age verification systems that rely on app-based biometric analysis or specific regional data infrastructure can produce silent failures when accessed from unsupported territories, often without surfacing a meaningful error message to the end user. This creates a structurally inequitable outcome: users in regions not formally supported by Yoti are effectively unable to complete a mandatory remediation step, regardless of their actual age or identity.
Anthropic's reliance on Yoti as its age verification mechanism reflects a broader industry push to implement child safety compliance following regulatory pressure, particularly in jurisdictions like the United Kingdom and European Union, where age assurance requirements have become increasingly codified in law. Yoti is a well-established provider in this space and holds certifications under several national frameworks. However, the outsourcing of account reinstatement to a third-party system introduces a dependency that Anthropic does not directly control, meaning users who fall through Yoti's technical gaps have no clear escalation path within Anthropic's own support infrastructure — a problem the user explicitly encountered when the "submit appeal" option simply redirected back to the suspended account screen.
The support experience described in the post — automated responses linking to inapplicable help center articles, circular UI redirects, and no apparent human escalation pathway — points to a gap in Anthropic's account remediation design. When a core access-restoration mechanism fails, users require a fallback that bypasses the broken system, not a loop that returns them to it. This is particularly acute in cases where the failure is systemic rather than user error, as appears to be the situation here. The user's expressed uncertainty about what is even causing the failure underscores how little diagnostic transparency the current system provides.
The incident connects to a wider tension in AI platform governance: the increasing imposition of age verification and identity assurance requirements on AI services, which are being applied at scale to large user bases with highly varied geographic and technical circumstances. Anthropic, like other frontier AI developers, is navigating genuine child safety obligations while managing a global user population that does not map cleanly onto the compliance infrastructure designed for specific regulatory markets. Until age verification systems either achieve broader geographic coverage or platforms like Anthropic implement robust human-review fallbacks for verification failures, users in unsupported regions will continue to face effective permanent account loss through no fault of their own.
Read original article →