EVERYTHING
According to Techtimes, Anthropic will begin requiring consumer users of Claude Free, Pro, and Max tiers to undergo rigorous identity verification starting in July. The new policy allows the company to demand government-issued photo IDs, live selfies, and facial geometry scans. This represents a notable shift for the AI industry, as it is the first time a major U.S. frontier AI lab has codified such extensive biometric data collection within its consumer terms of service.
Scope of data collection and verification methods
The updated policy introduces a specific "Verification Data" category that outlines exactly what information Anthropic may collect. Depending on the chosen verification method, this includes images of physical government documents—such as passports, driver's licenses, or national identity cards—alongside all personal details printed on them, including names and birth dates. Furthermore, the company will collect facial geometry templates, which Anthropic admits may be considered biometric data in certain legal jurisdictions.
While business customers on Team, Enterprise, and API plans remain exempt from these requirements, individual subscribers must comply or risk losing access. The verification process is handled through specific third-party vendors depending on the goal of the check:
Users must provide original physical documents, as the system will reject screenshots, photocopies, or digital IDs. Verification may be hindered by poor lighting, expired documents, or low-quality camera hardware.
Security concerns regarding third-party processing
The data processing is managed by Persona Identities, a San Francisco-based firm with ties to Founders Fund and Peter Thiel. While Anthropic acts as the data controller, Persona serves as the processor, meaning ID images and selfie data are stored on Persona's servers rather than Anthropic's infrastructure. However, this partnership has faced scrutiny following a February 2026 security incident where Persona's codebase was found exposed on a public endpoint.
The exposure revealed that Persona possesses the capability to perform extensive screening, including monitoring adverse media and filing Suspicious Activity Reports with financial regulators. Furthermore, Persona is authorized to retain biometric data, government ID numbers, and device fingerprints for up to three years. Despite these capabilities, Persona CEO Rick Song has denied any involvement in direct government surveillance lines.
This policy shift arrives amidst a changing regulatory landscape, following recent comments from the White House regarding Anthropic's national security status. As the company moves toward stricter identity protocols, users face an increasing trade-off between accessing advanced AI tools and maintaining personal data privacy.