Budget reconciliation bill H.R.1, the “One Big Beautiful Bill Act,” currently pending in the U.S. Senate, includes not only tax and spending provisions but also a proposed five-year pause on state and local enforcement of laws or regulations “limiting, restricting, or otherwise regulating” AI models, AI systems, or automated decision systems (“AI safety laws”). It is not yet clear whether these provisions will survive a full Senate vote or subsequent scrutiny in the House.
Please note H.R. 1 was in active debate in the Senate at the time this article was published, and the status and content of the bill may evolve in the coming days.
What Would H.R.1 Mean for State and Local AI Regulation?
As currently drafted, H.R.1 would require states and local governments to stop enforcing AI safety laws for the next five years (the “AI Enforcement Pause”). The U.S. Senate Committee on Commerce, Science, and Transportation (the “Committee”) and its Senators have proposed tying the AI Enforcement Pause to states’ receipt of federal broadband funding. Specifically, the Committee’s proposal would deny states $500 million to build and deploy infrastructure for AI tools under the Broadband Equity, Access, and Deployment (BEAD) grant program if they enforce any AI safety laws.
On June 26, the Ranking Member of the Committee, Senator Maria Cantwell (D-WA), released a memo warning that the proposed text of the AI Enforcement Pause purportedly “includes a built in backdoor to apply [the] new AI [Enforcement Pause] requirements to the entire $42.45 billion program, not just the new $500 million” conditional funds.
What Laws Would be Impacted by the AI Enforcement Pause?
The AI Enforcement Pause would not apply to, among other things, state enforcement of federal law or any “generally applicable law or regulation … that may address, without undue or disproportionate burden, artificial intelligence models, artificial intelligence systems, or automated decision systems to reasonably effectuate the broader underlying purposes of the law or regulation, such as a body of common law,” or laws that support the deployment of AI tools or that streamline licensing to promote the adoption of AI tools.
However, the AI Enforcement Pause could disrupt the patchwork of AI-specific legislation enacted by the states in the absence of federal legislation. For example, the AI Enforcement Pause could impact states’ ability to enforce AI-specific consumer protection laws, like the California AI Transparency Act, the Colorado AI Act, or the recently enacted Texas Responsible AI Governance Act. The AI Enforcement Pause could also impact sector-specific AI legislation, including in real estate, insurance, and healthcare.
Will the AI Enforcement Pause Survive Challenge?
As of June 27, the AI Enforcement Pause once again cleared the Byrd Rule, a procedural Senate rule that limits the inclusion of provisions unrelated to spending or revenue in reconciliation bills. 2 U.S.C. § 644. In a process called the “Byrd bath,” the Senate Parliamentarian typically removes or “cleans” extraneous content from a draft bill. The AI Enforcement Pause has now cleared that hurdle twice.
The Senate will vote on H.R.1 following a limited 20 hours of debate, during which Democrats will prioritize their concerns with the bill. Rejecting the AI Enforcement Pause would require a simple majority in the “vote-a-rama” that follows, and several Republican senators and congressmembers have already raised concerns with the provisions.
As a result, the fate of H.R.1’s proposed AI Enforcement Pause remains uncertain, with procedural hurdles ahead. Stakeholders should closely monitor developments, as the outcome will likely shape the U.S. AI regulatory landscape for years to come.
Domenic Boresta contributed to this article.
[View source.]
