A bipartisan US AI bill has been introduced in the House of Representatives, aiming to create a nationwide framework for regulating artificial intelligence while balancing innovation, safety and state oversight. The proposal seeks to address growing concerns about advanced AI systems by introducing federal safety requirements and clearer rules for developers.
The legislation, sponsored by Republican Rep. Jay Obernolte of California and Democratic Rep. Lori Trahan of Massachusetts, follows weeks of consultations with technology companies, researchers, advocacy groups and fellow lawmakers after an earlier discussion draft generated significant debate.
US AI bill narrows state law preemption
One of the biggest changes in the revised US AI bill is its approach to state regulations. An earlier version proposed temporarily overriding certain state AI laws for three years, a move that drew criticism from AI safety advocates and state officials.
The updated proposal significantly narrows those provisions. Instead of broadly limiting state authority, the legislation focuses on specific areas involving frontier AI systems, including catastrophic-risk transparency, independent audits, third-party verification and incident reporting.
At the same time, the bill preserves the ability of states to enforce laws covering AI deployment, child safety and other areas considered outside the scope of the federal framework.
Lawmakers said the revised language is more precise, providing greater clarity on which regulations would fall under federal oversight while addressing concerns raised during public consultations.
Safety standards for advanced AI models
The US AI bill introduces several new obligations for developers of advanced artificial intelligence systems.
Under the proposal, major AI developers would be required to:
• Publish safety reports outlining the capabilities and potential risks of newly developed AI models.
• Adopt widely accepted risk management frameworks to reduce potential harm.
• Undergo independent third-party audits to verify compliance with federal safety requirements.
The legislation is designed to increase transparency while creating consistent national standards for the development of powerful AI technologies.
Commerce Department gains oversight powers
A key provision of the US AI bill gives independent auditors an important role in identifying potentially dangerous AI systems.
If auditors determine that an AI model presents an imminent catastrophic risk, they would be required to notify the US Department of Commerce. The Commerce Secretary would then have the authority to suspend or restrict the deployment and use of those systems until safety concerns are addressed.
Supporters say these measures would provide an additional layer of oversight without unnecessarily slowing responsible AI innovation.
Congress faces a challenging timeline
Despite bipartisan backing, the US AI bill faces an uncertain legislative path.
Congress has limited time before the upcoming midterm elections, while lawmakers must also prioritize government funding legislation and other pressing matters. These scheduling pressures could make it difficult for the proposal to receive a House vote this year.
However, supporters argue that establishing clear national AI regulations has become increasingly urgent as artificial intelligence continues to evolve rapidly and influence industries ranging from healthcare and finance to education and national security.
Lawmakers behind the proposal remain optimistic that continued bipartisan cooperation will improve the chances of advancing comprehensive AI legislation before the end of the current congressional session.
