Latest AI Safety news

What's new with AI Safety? (as of Oct 20, 2026)

Oct 20, 2026 Anthropic updates its Frontier Safety Roadmap: automated red-teaming done early
Oct 16, 2026 Anthropic × METR: employee-like access without a veto?
Oct 15, 2026 What should an independent audit of a frontier lab look like? METR's terms
Oct 11, 2026 AMIE in The Lancet: 98 real patients, zero safety stops

↑