With Congress stalled, many states have moved ahead, passing dozens of laws aimed at regulating artificial intelligence — setting rules for child safety, demanding more transparency from companies and creating whistleblower protections. That momentum, however, has sparked tension with the White House, which argues a single national framework is needed to give innovators certainty and avoid a patchwork of state rules.
The White House in March released a national AI legislative framework it wants Congress to adopt. Administration officials, including Michael Kratsios, head of the Office of Science and Technology Policy, say only Congress can create the uniform environment innovators need. “We want to create an environment where innovators have certainty about the way that they can develop their products,” Kratsios said, adding that a federal framework is the first step to avoid inconsistent state rules.
That message has reached state capitals — sometimes halting legislation. Utah state Rep. Doug Fiefia, a Republican and former Google employee, said a bill he proposed to improve transparency around company consumer protections never reached a vote after the White House circulated a short memo opposing it. Fiefia said the memo characterized the bill as “unfixable” and contrary to the administration’s AI agenda, though he received no detailed explanation. A White House official, speaking on background, told NPR the administration has not told a state it cannot pass child safety protections, but did not comment specifically on the memo to Fiefia.
Many state lawmakers — including Republicans — say they feel compelled to act because Congress is gridlocked. “Congress is in a gridlock and they not only will not act, they can’t act,” Fiefia said. “In states like Utah we see this as an opportunity to step forward and protect our constituents and our citizens, especially as it relates to child safety.”
Other GOP state lawmakers expressed similar views. Pennsylvania State Sen. Tracy Pennycuick, sponsor of the SAFECHAT Act that requires safeguards to stop chatbots from encouraging self-harm or violence, said states can move faster than the federal government. “States are the first ones to see when there’s a problem and they have the ability to pivot and act quickly,” she said.
In Texas, State Sen. Angela Paxton, also a Republican, warned against a confusing mix of laws but said states should retain the ability to legislate until Congress acts. “We don’t want to have the patchwork of regulatory structures … in general that’s a good rule of thumb,” she said. “I like the idea of there being strong federal legislation but until that exists, I think we have to preserve the ability of states to pass laws.”
The White House framework lists principles the administration would like Congress to adopt, including protecting children and guarding consumers from issues such as rising data center costs. But reactions from experts and some lawmakers have been mixed. Supporters say a single federal standard is preferable to 50 different regimes; critics say the framework lacks specificity and accountability measures.
Riki Parikh, policy director at the nonprofit Alliance for Secure AI, argued the framework doesn’t address crucial topics like job displacement or robust corporate accountability. “A federal standard is better than a 50-state patchwork,” Parikh said. “But what they are proposing here is not sufficient. It does not earn the right to replace the good work states are doing.”
Tennessee Attorney General Jonathan Skrmetti praised the framework as a positive step compared with an earlier White House push for a 10-year moratorium on state AI laws, a proposal last year that alarmed some critics and was favored by technology companies. Efforts to impose that moratorium — backed by the White House and allies such as Sen. Ted Cruz — ultimately failed. Skrmetti remains wary, however, of the administration’s ties to the AI industry and what that might mean for regulation.
Public opinion reflects some of that concern. A January survey by Morning Consult and the Tech Oversight Project found a majority believes the Trump administration is too close to Big Tech. Other polls, such as one from Vanderbilt University, show bipartisan support for AI regulation, with more Republicans than Democrats favoring restrictions in some surveys.
On Capitol Hill, GOP allies have expressed support for the White House principles, but major legislation has not yet moved. Sen. Marsha Blackburn of Tennessee said she is coordinating with the White House on her TRUMP AMERICA AI Act, a more expansive bill that would codify the administration’s agenda into law. The White House says it is holding “productive conversations” with lawmakers as it seeks congressional action.
For now, the tug-of-war continues: the White House urging Congress to craft a single national approach, and states — including many led by Republicans — pressing forward with their own laws to address immediate harms and gaps.