AI risk thresholds

Anthropic Proposes AI Safety Bill to Block Models at High Risk

The plan includes independent evaluators and judicial oversight, but faces questions about independence and real-world applicability.

Read More »