AI Behavior

Anthropic Asks Government Power to Block Risky AI and Mandate Independent Tests

Published June 10, the company's policy essay urges legally required third‑party audits to give government veto power over models that pose cybersecurity, biological or loss‑of‑control risks.