White House and Congress Relaunch a Push to Preempt State AI Laws

--- headline: "White House and Congress Relaunch a Push to Preempt State AI Laws" slug: white-house-hill-state-ai-preemption-push category: policy story_number: 17 date: 2026-06-09 ---

The clock on Colorado’s landmark AI law is ticking. In roughly three weeks, on June 30, the first real AI enforcement deadline on American soil was supposed to arrive — granting consumers the right to appeal automated decisions made about them. But Washington isn’t waiting. The White House and key congressional allies have quietly relaunched one of tech’s most coveted prizes: federal preemption of state AI laws.

According to an Axios scoop published June 8, the Trump administration is negotiating a package deal in which it would support Hill tech-policy priorities — particularly legislation protecting children online and combating AI-generated deepfakes — in exchange for Congress delivering federal preemption of at least some state AI regulations. The talks represent the most serious preemption push since a 99-to-1 Senate vote in July 2025 crushed a proposed 10-year moratorium on state AI rules buried inside the “One Big Beautiful Bill.”

Sen. Marsha Blackburn (R-Tenn.) is at the center of the new effort. “Senator Blackburn is spearheading the negotiation with the White House to finalize legislative text of an AI preemption package that includes protections for kids, creators, and communities through the Senate version of KOSA, the NO FAKES Act, and age verification requirements,” a Blackburn spokesperson told Axios. A White House official confirmed the administration’s role, saying it “continues to proactively engage across government and industry.”

The Blackburn-brokered framework is distinct from a parallel legislative vehicle: the Great American Artificial Intelligence Act, a 269-page discussion draft introduced June 4 by Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.). The Obernolte-Trahan bill would impose a three-year moratorium on new state laws “specifically regulating the development” of AI models, while leaving state authority over AI deployment and use intact. It would also mandate semi-annual third-party audits of frontier developers and formally enshrine the Center for AI Standards and Innovation in statute with $100 million per year in authorized funding for fiscal 2027 through 2029.

Axios reported, however, that the Obernolte-Trahan bill “isn’t the likely vehicle for AI policy in this Congress,” a signal that the Blackburn-White House track is where the real dealmaking is happening.

The trade on the table

The structure of the emerging deal reflects a classic Washington bargain: give tech companies relief from a patchwork of state regulations in exchange for popular consumer-protection wins that both parties can take home to voters. Linking preemption to kids-online-safety legislation and anti-deepfake rules gives Senate Republicans cover to support what has previously been a politically toxic ask.

That toxicity is real. When the Trump administration first tried to preempt state AI laws via the reconciliation bill in 2025, it was met with an avalanche of opposition from advocacy groups, state officials, and Republican governors. The 99-to-1 Senate rebuke was a rare and humiliating defeat. Blackburn’s explicit endorsement this time around is considered crucial — she had not backed the previous attempt.

The administration has also been laying groundwork through executive action. In March 2026, the White House released a National Policy Framework for Artificial Intelligence calling for preemption of state laws and steering away from “open-ended liability” for AI firms. A December 2025 executive order went further, directing federal agencies to challenge conflicting state AI laws through litigation and establishing an AI Litigation Task Force — which later joined an xAI lawsuit seeking to block Colorado’s original AI Act.

Colorado: the deadline that defused itself

Colorado’s June 30 deadline had been the most concrete symbol of state-level AI enforcement. But under a combination of legal pressure, White House signaling, and industry lobbying, Colorado Governor Jared Polis signed Senate Bill 26-189 on May 14, 2026, repealing the original AI Act and replacing it with a narrower disclosure-and-rights framework focused on automated decision-making. The new law does not take effect until January 1, 2027, effectively defusing the June 30 flashpoint. The DOJ, alongside xAI, had filed suit in federal court seeking to block the original law on constitutional grounds.

California’s frontier-AI transparency law remains in effect, and the Obernolte-Trahan draft explicitly identifies California’s AB 2013 and a portion of SB 942 as laws that would be preempted under its three-year freeze.

The opposition holds firm

Critics have not softened. A bipartisan coalition of 36 state attorneys general sent a letter to Congress in late 2025 opposing any federal ban on state AI consumer-protection laws, arguing states have historically been the backstop for tech accountability. Connecticut Attorney General William Tong led that effort, and more than 260 state legislators from all 50 states had separately called the reconciliation-bill preemption provision “reckless.”

Civil liberties groups were swift to condemn the Obernolte-Trahan draft. “This draft bill fails to learn from Congress’s previous attempts to block state AI regulations,” said Jina John, senior policy counsel for AI, privacy, and technology at the ACLU, on June 4. “The federal government must not grant tech giants a blank check to deploy these technologies without the oversight of states.”

Brad Carson, president of Americans for Responsible Innovation and a former Democratic representative from Oklahoma, called the approach “a generational mistake.” “This bill takes the current floor on state AI legislation and turns it into a federal ceiling, preventing state lawmakers from addressing emerging AI harms in an era of fast-moving technology,” Carson said.

Even within the industry’s orbit, the reaction was mixed. Brendan Steinhauser, CEO of the safety-focused Alliance for Secure AI, praised the bill’s bipartisan focus on catastrophic risk but opposed its preemption clause: “A national AI standard should protect at least as much as it preempts, and states across the country have already enacted laws that do far more for Americans on issues like child safety and consumer protections than this proposal.”

Why it matters

The renewed push reveals a fundamental tension in American AI governance: the absence of any comprehensive federal AI law has created a proliferating, contradictory web of state regulations that industry argues stifles innovation, while consumer advocates say is filling a vacuum Washington refuses to occupy. Pairing preemption with child-safety and deepfake legislation is a shrewd political construction — it forces opponents to vote against popular protections in order to defend state authority.

But the opposition coalition is formidable and bipartisan. The 36-AG letter, the 260-plus state legislators, the ACLU, and AI-safety groups represent a durable constituency that has already beaten back this fight twice. And as Axios noted, the pressure of August recess in a midterm election year could compress the legislative window to near-zero.

The Obernolte-Trahan bill remains a discussion draft — not yet formally introduced — meaning the Blackburn-negotiated Senate package is the more credible near-term path, even as its final contours are still being hammered out behind closed doors.

What to watch

- Whether the Blackburn-White House package crystallizes into legislative text before August recess. - How narrowly “subject-matter preemption” is defined — preempting AI-development rules is far narrower than preempting all AI-related state laws. - Whether state AGs escalate legal challenges to executive-branch preemption efforts already underway. - The fate of California’s remaining AI transparency laws, which industry has specifically targeted. - How child-safety and NO FAKES Act provisions are crafted, since their specifics could either broaden or collapse the coalition needed for passage.