Overview
- In mid-June, Anthropic paused access to its Mythos-class models including Fable 5 after a U.S. instruction to limit access to U.S. citizens left the company unable to enforce geographic controls.
- The company has announced it will require identity verification for a subset of Claude users using a photo ID and a short live‑selfie check run by vendor Persona, a choice that has raised user privacy concerns because of Persona’s investor links.
- Anthropic says the checks will first apply to a small group of accounts already flagged for possible policy violations and that verified users will be allowed to keep their accounts.
- The company denies any connection between the new ID checks and the recent model suspension and says it routes sensitive requests away from its most capable model to a more restricted model, Claude Opus 4.8.
- The move could shift how users interact with Claude by adding identity gates for enforcement, heighten scrutiny of third‑party verification vendors, and signal closer interaction between AI firms and regulators as Anthropic adjusts billing and access plans.