AI risk governance gaps, including biosecurity
Analysis warns that biological-weapons governance may not cover AI-designed protocols, sequence files, and cloud-lab access, with U.S.–China rivalry complicating coordination. New reporting on state officials’ demands for Anthropic IPO transparency connects existential-risk claims to disclosure, control, and release decisions; these remain demands, not binding rules. A claimed 2025 federal risk-assessment law and proposed AI-restraint frameworks warrant follow-up, with legal status, implementation, and verification unresolved.
Sources (4)
Updated Nov 18, 2026