ai
Bipartisan bill proposes federal AI rules and three-year state preemption
A discussion draft introduced by two members of Congress would establish federal standards for large AI developers and block new state AI regulations for three years. The measure also creates reporting and audit requirements for frontier model companies.
urbanmilwaukee.comAudio version
Tap play to generate a narrated version.
Developing·Limited corroboration so far. This page will refresh as more sources emerge.
A bipartisan discussion draft called the Great American AI Act would set federal standards for large AI developers and impose a three-year ban on new state AI regulations. The bill, introduced by Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA), applies to companies with more than $500 million in revenue.
Those firms would have to publish a frontier AI framework covering catastrophic risks, cybersecurity, and deployment decisions, and submit safety reports before releasing new models.
Reporting and enforcement provisions Developers would report accidents to the Center for AI Standards and Innovation, an office inside the Commerce Department that the bill would codify. Independent verification organizations licensed by the center could audit companies, and violations could result in civil penalties of $1 million per incident.
The Department of Homeland Security would offer assistance to organizations seeking to reduce AI-related threats. The bill also directs the National Science Foundation to fund AI-related research, scholarships, and workforce studies.
State preemption and reactions The legislation would prevent states from enacting new AI rules for three years. It would override a California statute requiring disclosure of AI training data summaries. Rep. Chip Roy (R-TX) said federal preemption is a non-starter.
Patrick Hedger, director of policy for NetChoice, warned that auditing and data-sharing rules could expose trade secrets. Brendan Steinhauser, CEO of the Alliance for Secure AI Action, said the risk of not using independent verification organizations is greater than the risk of using them.
