A federal effort to stop states from passing their own AI laws has stalled again, this time after Republicans dropped the proposal from the National Defense Authorization Act.
The measure, backed by Donald Trump, would have blocked state AI laws for a decade. Its removal does not end the fight, but it shows how difficult it has become for Republicans to unite around a national preemption strategy while states, advocacy groups, and industry-aligned organizations battle over who should write the rules for artificial intelligence.
Why the NDAA path failed
House Majority Leader Steve Scalise (R-La.) told reporters Tuesday that some Republicans are now “looking at other places” to move the proposal. The reason, according to Scalise, was not that interest had disappeared. He said the National Defense Authorization Act “wasn’t the best place” for the measure “to fit.”
The proposal was part of a broader push by Trump to stop a state-by-state approach to AI regulation. Trump has argued that a patchwork of state rules would slow innovation by forcing AI firms to spend time and resources complying with different legal regimes.
That argument has not been enough to hold Republicans together. The source article notes that Republicans first voted against including a similar measure in the “Big Beautiful” budget bill, then failed this week to settle on a way to include the AI preemption in the NDAA.
Among the Republican figures pushing back this week were Rep. Marjorie Taylor Greene (R-Ga.), Arkansas Gov. Sarah Huckabee Sanders, and Florida Gov. Ron DeSantis, according to The Hill.
The fight is about speed, control, and risk
The core dispute is simple: supporters of federal preemption want one national standard, while opponents want states to keep the power to regulate AI risks as they emerge.
Trump made the national-standard argument directly on Truth Social last month:
“We MUST have one Federal Standard instead of a patchwork of 50 State Regulatory Regimes,” Trump wrote. “If we don’t, then China will easily catch us in the AI race. Put it in the NDAA, or pass a separate Bill, and nobody will ever be able to compete with America.”
Scalise offered a similar economic rationale. He said AI is an area where “a lot of new massive investment is going” and that “we want that money to be invested in America.” He also argued that state-level limitations had drawn federal attention because of concerns about an open marketplace and innovation.
Opponents see the issue differently. They argue that states can move faster than Congress and can respond to risks involving children, workers, families, fraud, crime, national security, and frontier-model dangers.
Senate Majority Leader John Thune (R-SD) called the measure “controversial,” The Hill reported. He also suggested that a White House compromise could leave some state authority intact because “you know, both sides are kind of dug in.”
A broad coalition pushed back
Americans for Responsible Innovation, a bipartisan group that lobbies for AI safety laws, celebrated the measure’s removal from the NDAA. The group said the proposal had faced opposition from a nationwide, bipartisan coalition that included state lawmakers, parents, faith leaders, unions, whistleblowers, and other public advocates.
Brad Carson, ARI’s president, described the opposition as a “widespread and powerful” movement and said critics had pushed back against a “rushed attempt to sneak preemption through Congress.” He argued that “Americans want safeguards that protect kids, workers, and families, not a rules-free zone for Big Tech.”
The source article also notes that the fight may not stay in Congress. If lawmakers do not find another way to pass the measure, Trump will likely release an executive order to enforce the policy. Republicans in Congress had previously dissuaded Trump from releasing a draft order while they looked for legislation where an AI moratorium could pass.
The lobbying battle around state AI laws
The struggle over AI preemption is also being shaped by a major lobbying fight. Forbes profiled a $150 million AI lobbying war last month, with ARI and Leading the Future on opposing sides.
ARI uses funding from “safety-focused” and “effective altruism-aligned” donor networks to support state AI laws. The group expects those laws can pass faster than federal regulations aimed at emerging risks.
Leading the Future, or LTF, is on the other side. Forbes reported that LTF is “backed by some of Silicon Valley’s largest investors” who want to block state laws and prefer a federal AI framework.
The priorities described in the source show why the conflict has become so intense:
- Protecting kids from dangerous AI models
- Preventing AI from supercharging crime
- Protecting against national security threats
- Addressing “long-term frontier-model risks”
Some Republicans have supported compromises that would preserve state authority in areas such as protecting children or preventing fraud. But the source article says Trump’s opposition to AI safety laws like New York’s “RAISE Act” seems unlikely to fade as the White House considers weakening federal preemption.
Why New York’s RAISE Act matters
The New York RAISE Act has become a central example in the larger fight. New York lawmakers passed the RAISE Act this summer, but it is still waiting for New York’s Democratic governor, Kathy Hochul, to sign it into law.
If signed, potentially by the end of this year, big tech companies like Google and OpenAI would have to submit risk disclosures and safety assessments. Companies that fail to comply could face fines up to $30 million.
Alex Bores, a Democrat and author of the RAISE Act, has become LTF’s prime target to defeat in 2026, Politico reported. CNBC reported that LTF plans to invest many millions in ads to block Bores’ Congressional bid.
LTF leaders Zac Moffatt and Josh Vlasto accused Bores of “pushing “ideological and politically motivated legislation that would ‘handcuff’ the US and its ability to lead in AI,” Forbes reported.
Bores told Ars that even tech industry groups spending hundreds of thousands of dollars opposing his law had reported that tech giants would only need to hire one additional person to comply. He said that showed how “simple” compliance with many state laws could be.
Bores also said he agrees that a federal law would be better than a patchwork of state laws. But he argued AI is moving “too quickly,” and that “New York had to take action to protect New Yorkers.”
He described the RAISE Act as “a limited bill that is focused on the most extreme risks” and “be a first step that deals with the absolute worst possible outcomes” while Congress works on a federal framework. Those risks include bad actors using AI to build “a bioweapon or doing an automated crime spree that results in billions of dollars in damage.”
For now, state AI laws have survived another attempt at federal preemption. But the proposal is still alive, and the next stage may come through another bill, a compromise, or an executive order.