# Summary of H.R. 9477: AI Incident Reporting Act
This bill requires major artificial intelligence model developers to report certain dangerous AI incidents to the Secretary of Commerce. The Secretary has 180 days to establish rules determining which AI models pose significant risks to national security or public safety and must be reported, with developers having seven days to report incidents once they discover them. Reportable incidents include AI systems that try to evade human control, steal their own code, enable cyberattacks or weapons development, or demonstrate other serious capabilities. Companies must submit initial reports within seven days and supplemental reports with updates, and the Secretary must inform Congress within 48 hours for imminent threats or within 30 days for other incidents. The bill protects reported information from public disclosure and shields companies from using their own reports against them in lawsuits or regulatory enforcement, though the government can still use the information to address security threats. Developers who fail to report face civil penalties up to two million dollars per violation.
Take Action
Your position
Add a comment
to comment on this bill.
Annotate the text
Highlight any passage on the Summary or Full Text tab to attach a note. Annotations appear on the Annotations tab.