House AI Regulation Deal Faces Collapse Over State Preemption Fight

A sweeping bipartisan effort to establish federal oversight of artificial intelligence appears headed for failure this year, as lawmakers from both parties reject key provisions in a 269-page draft bill released Thursday by Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.).

The proposal would impose nationwide safety and transparency requirements on major technology companies developing advanced AI systems, while also funding research, education, and workforce development programs. But the framework has drawn fire from Democratic leadership and Republican skeptics alike, with a shrinking legislative calendar threatening to delay any congressional AI action until 2027 at the earliest, according to details first reported by POLITICO.

Why it matters

The collapse of this negotiation reveals the deep partisan divide over AI regulation strategy. Democrats want to preserve state-level authority to address AI harms, while Republicans seek to shield tech companies from a patchwork of state rules. Without federal consensus, AI development will continue largely unregulated even as the technology reshapes business operations, content creation, and workforce dynamics across industries.

Democrats reject state law restrictions

The most contentious element is a three-year moratorium on state laws regulating AI development. Three House Democrats leading their party's AI policy—Reps. Josh Gottheimer (N.J.), Valerie Foushee (N.C.), and Ted Lieu (Calif.)—issued a statement saying the draft "cannot serve as the basis for productive dialogue."

AI safety advocates warn the preemption language is broad enough to let tech companies challenge existing state protections for children and content creators. "It says all state laws that regulate 'development,' and development is defined pretty broadly in the bill," said Charlie Bullock, a senior research fellow at the Institute for Law and AI.

Adam Kovacevich, founder of tech industry group Chamber of Progress, suggested Democrats are holding out for potential electoral gains in November that would give them greater leverage over technology policy.

Republican leadership remains skeptical

House Majority Leader Steve Scalise expressed doubts about the compromise approach, telling POLITICO he had heard the innovation protections were "narrow" and warning against "some big regulatory structure at the federal level trying to control algorithms of private companies." Scalise indicated he supports broader preemption of state AI laws than the current proposal offers.

Speaker Mike Johnson declined to commit to bringing any AI legislation to the floor before the election, saying only that action would come "as soon as we are able to build consensus around a package."

Obernolte acknowledged he has been communicating with the White House but said he doesn't know whether the Trump administration supports the bipartisan framework. "We haven't asked them that specific question," he said.

Tech industry response muted

While some technology lobbyists praised the attempt to balance federal oversight with state preemption, major AI companies including OpenAI, Anthropic, Google, and Microsoft either declined to comment or did not respond to requests for reaction.

Tech lobbyists across Washington appear resigned to the proposal's dim prospects. "I think it's been made pretty apparent that House leadership isn't going to just get on board with whatever Jay Obernolte decides should be the federal framework for AI," one lobbyist told POLITICO on condition of anonymity.

Spokespersons for both Obernolte and Trahan emphasized that Thursday's release is a discussion draft and their offices remain open to feedback as they seek support from colleagues.

POLITICO first reported these developments.