Congress is unlikely anytime soon to pass an artificial intelligence governance bill that would preempt state regulation, says Ryan Hauser, research fellow at the free-market Mercatus Institute, who argues lawmakers should focus on advancing implementation of the Trump administration’s artificial intelligence action plan and on improving AI cybersecurity.
“If the U.S. can't get comprehensive, pro-innovation, preemptive federal AI legislation done, my hope is that lawmakers can at least scuttle the costliest bills and focus on empowering agencies like the Cybersecurity and Infrastructure Security Agency , the NIST Center for AI Standards and Innovation and others to execute the [Trump] AI action plan and promote public-private cooperation to protect Americans from malign foreign influence,” Hauser told Inside AI Policy.
Hauser noted that the end of the 43-day government shutdown last week “doesn't greatly increase the odds of getting preemptive federal AI legislation before Jan. 30, when federal funding runs dry again.”
In fact, he said, “As of [late last week], Polymarket shows an 8% chance of the U.S. enacting an AI safety bill by December 31st, 2025,” meaning a legislative breakthrough is unlikely.
“That is right where I'd put my own subjective estimate,” the AI policy analyst said.
“It's tough to see comprehensive AI legislation with state preemption getting passed,” Hauser explained, “given the lingering distaste of many state GOP lawmakers and U.S. senators like Marsha Blackburn of Tennessee. Her critical withdrawal helped scuttle moratorium negotiations with [Senate Commerce Chairman] Ted Cruz [R-TX] in July. GOP governors led by Arkansas Gov. Sarah Huckabee Sanders also weighed in against the moratorium.”
Hauser said, “Preemptive legislation would need strong presidential leadership -- not just Senator Cruz and Representative [Jay] Obernolte [R-CA] leading the charge. Otherwise, I don't see how you get the votes when Republican lawmakers are often just as worried as Democratic lawmakers when it comes to AI.”
He said, “Congress will have just two and a half months over the holiday season to get a deal done while navigating competing legislative priorities in the House and Senate. Given the resistance we saw this summer and a lack of appetite for negotiating new, comprehensive regulatory frameworks, I'd say it's a long shot.”
Industry leaders are pledging to continue pushing for legislation in the near term that would pause regulatory activity at the state level.
Paul Lekas, executive vice president of the Software & Information Industry Association, last week told Inside AI Policy “we’re also looking to Congress to make progress on AI legislation and define the parameters for federal and state lawmaking -- ideally ahead of what’s expected to be a packed AI agenda during the 2026 state legislative sessions.”
But Hauser said, “I would expect more of the same -- continued executive branch implementation of the White House's AI action plan and other agency-based moves, possibly alongside narrower bills on watermarking, deepfakes, adversarial AI bans within the U.S. government.”
“There's also the GAIN AI Act,” he said, “which would prioritize U.S. chip access while limiting exports under certain conditions. NVIDIA has been massively boosting its presence in DC over the past year and is looking to quash that bill. It already lost the fight on its ability to sell [the advanced AI chip] Blackwells to China and likely is not eager to see a repeat.”
“Moreover,” he said, “advancing exports for the American tech stack has been a consistent goal articulated by the White House at least since the start of the President Trump's current term.”
Hauser also emphasized that cyber “should be front and center, especially in the wake of news that Chinese state actors reportedly used Anthropic's Claude Code tool to conduct an ‘AI-orchestrated cyber espionage campaign.’ These attacks targeted 30 or so global institutions. … This is an issue of bipartisan concern, and these kinds of attacks will only increase in quality and quantity as new advances are made by frontier labs and in open-source AI.”