Senate bill sets a 72-hour AI incident window

U.S. Senators Brian Schatz and Mark Warner introduced the Artificial Intelligence Risk Management and Security Act of 2026, which would require frontier-AI companies to report serious incidents within 30 days, or within 72 hours when an incident threatens national security, critical infrastructure, or public safety. The senators’ bill summary also proposes 45 days of pre-release access for evaluators and civil penalties up to $250,000 per violation, per day. It is only a proposal, but builders shipping agents should start retaining incident evidence, authority boundaries, and release-gate decisions—the same records the agent-evaluation governance case says turn testing into operational control.