Six members of the House Energy and Commerce Committee released a 269-page discussion draft on June 4 that would stop states from regulating the development of AI models for three years. The bill is led by Reps. Jay Obernolte, a Republican from California, and Lori Trahan, a Democrat from Massachusetts. They are calling it the Great American Artificial Intelligence Act of 2026.
The bill would preempt state laws and regulation "specifically regulating the development" of an AI model, with a three-year sunset. It specifies that preemption would not apply to laws related to the use or deployment of AI models. The distinction matters: a state could still regulate how a chatbot is used in hiring. It could not require a developer to run safety tests before shipping the model in the first place.
The bill would require large frontier developers — those above $500 million in annual revenue — to publish safety frameworks, report critical incidents, and submit to semi-annual third-party audits. It would also codify the Commerce Department's Center for AI Standards and Innovation.
What gets preempted, specifically
The discussion draft would create a national standard that preempts recently passed state laws in New York and California that require companies to publish safety protocols and security incidents. An accompanying document released by Trahan's office said that California's AB 2013 law, which requires model developers to publicly post high-level summaries of their training data, would be preempted.
The bill does not operate in isolation. President Trump signed an executive order in December ordering the White House Office of Science and Technology Policy to recommend federal AI legislation preempting any state laws in conflict with the administration's framework. The discussion draft comes just days after Trump signed a separate executive order to establish voluntary federal agency reviews of AI.
The bill includes no mentions of protection for children, but a spokesperson for Obernolte said the congressman intends for child protections to be enacted in separate legislation. Sen. Marsha Blackburn's competing Senate draft takes the opposite approach — her version contains strict child safeguards built in.
Who's praising it and who isn't
The reaction broke along predictable lines. The Information Technology Industry Council, which represents major tech firms, praised the bill, saying "Congress must set a national standard that enables responsible AI development, deployment, and adoption across the American economy."
Brad Carson, president of Americans for Responsible Innovation and a former U.S. representative, called preemption a "generational mistake," saying the bill "takes the current floor on state AI legislation and turns it into a federal ceiling."
Public Citizen said the proposal would prohibit states and local governments from enacting or enforcing laws specifically regulating AI model development while leaving oversight largely to a federal government that has repeatedly failed to pass meaningful AI protections.
The draft is not a bill yet. A discussion draft is meant to solicit feedback from lawmakers and experts before it would be formally introduced. The sponsors have not set a timeline for formal introduction.
For small business owners, the bill's preemption logic cuts both ways. The three-year freeze on state development rules would mean fewer patchwork compliance obligations if a small firm is building AI tools — but the $500 million revenue threshold for the bill's audit and reporting requirements sits well above what most small operators generate. Existing state laws are already shaping vendor contracting practices and downstream compliance expectations, particularly through AI-specific addenda and third-party risk allocation — obligations that tend to flow downhill to smaller vendors regardless of whether a law technically applies to them. A federal preemption that clears out state development rules would simplify that picture, at least on paper, for three years.
Whether the bill survives the Senate — or finds the votes to formally launch in the House — remains an open question. More than 300 AI-specific standing orders in federal courts alone have failed to arrest the pace of AI-related legal disputes; Congress has been debating AI legislation since at least 2023 without passing a framework. The Great American Artificial Intelligence Act is the most detailed bipartisan House attempt yet.
