Two House lawmakers from opposite parties have produced what may be the most ambitious attempt yet to write a national rulebook for artificial intelligence. On June 4, Representatives Jay Obernolte (R-CA) and Lori Trahan (D-MA) released a 269-page discussion draft of the "Great American Artificial Intelligence Act of 2026" (GAAIA), a sweeping bid to set federal terms for how the country's most powerful AI systems are built, audited, and overseen — and to temporarily strip states of part of their authority to do the same.
The draft is not a bill on the floor, let alone law. It is a discussion draft, the legislative equivalent of a working blueprint circulated to invite reaction. But its scope is unusual. Organized into four titles — Frontier AI Governance, Workforce, Cybersecurity, and Research, Development, and International Cooperation — it is being described by its sponsors and outside analysts alike as the most comprehensive federal AI framework Congress has yet put forward.
"This discussion draft is an important step toward building a clear federal framework that promotes innovation, protects Americans from emerging risks, and ensures the United States continues to lead the world in AI," Obernolte, a former video-game developer who chairs the House AI task force, said in a statement accompanying the release.
Trahan framed the stakes in starker terms. "The threats AI poses to our national security, our safety, and our workforce are here and growing by the day," she said. "This bipartisan framework is designed to meet the challenges posed by this rapidly advancing technology without smothering American innovation. It protects workers, establishes real accountability for the most powerful frontier systems, and positions the United States to set the global standard on AI."
What the draft would do
At the center of the draft is a tiered system aimed at the largest model builders. It defines a "frontier developer" as an entity that has trained a frontier model and has more than $50 million in gross revenue, and a "large frontier developer" as one topping $500 million. The biggest obligations scale with size.
Under the draft, frontier developers would have to publish transparency reports describing their models and report critical safety incidents to the government (Section 111). Large frontier developers would face the heaviest lift: independent third-party audits (Section 112) and detailed incident reporting to federal authorities. The draft also bars retaliation against whistleblowers (Section 113) and, notably, extends whistleblower protections to employees and contractors at any employer who report "AI violations" — not just those working inside the marquee labs.
Beyond governance, the draft authorizes $100 million per fiscal year for a Center for AI Standards and Innovation, codifies the National Artificial Intelligence Research Resource to broaden access to compute and federal datasets for AI training, adds penalties for using AI to impersonate government officials, and directs the Census Bureau and the Bureau of Labor Statistics to add AI-adoption questions to federal surveys.
The preemption fight
The provision drawing the most heat is the one that pits Washington against the states. The draft would impose a three-year preemption of state laws that "specifically regulate the development" of AI models. States would keep the power to govern how AI systems are used and deployed inside their borders, and existing consumer-protection, civil-rights, and privacy laws would remain intact. But for three years, states could not write new rules dictating how the models themselves are built.
In practice, that would freeze or supersede AI-development transparency requirements in states such as California, New York, and Illinois. The timing is pointed. Colorado's AI Act — which gives residents the right to appeal certain decisions made about them by AI — was scheduled to take effect June 30, 2026, though a court recently granted a joint motion by xAI and the Colorado Attorney General to temporarily suspend its enforcement.
The draft also lands inside a broader federal push to wrest AI policy from the states. The Trump administration signed a June 2 executive order promoting advanced AI innovation and security, and a Justice Department AI Litigation Task Force has begun challenging state AI laws. Against that backdrop, the GAAIA's three-year sunset is being read by supporters as a more measured, negotiated alternative to a blanket federal override — and by critics as the same idea in friendlier packaging.
Supporters and skeptics
The draft arrived with bipartisan cover. Representatives Scott Franklin (R-FL), Suhas Subramanyam (D-VA), Erin Houchin (R-IN), and Scott Peters (D-CA) signed on in support, and industry voices have welcomed the prospect of a single national standard in place of a growing patchwork of state rules.
But the reception was far from unanimous. The co-chairs of the House Commission on AI and the Innovation Economy said the draft "cannot serve as the basis for productive dialogue," a blunt rebuke from within Congress itself. Consumer advocates and several state officials warned that pausing states' ability to regulate model development — even temporarily — removes a layer of protection at the precise moment frontier systems are advancing fastest, and some characterized the move as freezing state consumer protections for three years.
That tension — between a clean national framework and the states' role as policy laboratories — is the fault line the discussion draft now has to survive.
Why It Matters
For years, the absence of comprehensive federal AI law has left a vacuum that states rushed to fill, producing a thicket of overlapping and sometimes conflicting rules. The GAAIA is the most serious congressional attempt yet to replace that patchwork with a single regime, and it does so by coupling real obligations on the biggest developers — audits, transparency, incident reporting, whistleblower protection — with a partial, time-limited rollback of state power. How Congress resolves that trade-off will shape who gets to set the rules for the most consequential technology of the decade.
What to Watch
Watch whether the discussion draft attracts enough cross-party support to become formal legislation, or whether the preemption clause sinks it. Watch the fate of Colorado's suspended AI Act and other state laws as the Justice Department's litigation task force advances. And watch whether the three-year sunset survives negotiation intact, gets shortened, or is stretched toward the permanent preemption the administration has signaled it prefers.
Sources
- [Obernolte, Trahan release a discussion draft of the Great American AI Act](https://obernolte.house.gov/media/press-releases/obernolte-trahan-release-discussion-draft-great-american-ai-act) — obernolte.house.gov - [Trahan, Obernolte Unveil Federal AI Framework Discussion Draft](https://trahan.house.gov/news/documentsingle.aspx?DocumentID=3783) — trahan.house.gov - [Bipartisan AI draft proposes three-year preemption of state laws](https://rollcall.com/2026/06/04/bipartisan-ai-draft-proposes-three-year-preemption-of-state-laws/) — rollcall.com - [Unpacking the Great American Artificial Intelligence Act of 2026](https://www.techpolicy.press/unpacking-the-great-american-artificial-intelligence-act-of-2026/) — techpolicy.press - [Bipartisan 'Great American AI Act' draft proposes new federal AI governance framework](https://fedscoop.com/bipartisan-great-american-ai-act-draft-proposes-new-federal-ai-governance-framework/) — fedscoop.com - [Unpacking the Great American AI Act](https://www.dlapiper.com/en-us/insights/publications/2026/06/unpacking-the-great-american-ai-act) — dlapiper.com - [Colorado's AI Law Takes Effect June 30](https://www.techtimes.com/articles/318002/20260608/colorados-ai-law-takes-effect-june-30-it-gives-you-right-appeal-decision-ai-made-about-you.htm) — techtimes.com
"This bipartisan framework is designed to meet the challenges posed by this rapidly advancing technology without smothering American innovation."- Lori Trahan, U.S. Representative (D-MA)