The fight over state AI laws moves from Congress to the courts

An executive order sets up a task force to challenge state AI laws and ties broadband money to them. Founders face a contested period, not a clean rule.
State AI laws now face a federal legal team built to challenge them. The executive order on a national policy framework for AI that President Trump signed on 11 December tells the Attorney General to set up an AI Litigation Task Force within 30 days, aimed at state laws that conflict with federal policy.
The order does not settle the fight; it moves it. An executive order, like the 2023 order on frontier models, cannot repeal a state law. Preemption needs an act of Congress or a court ruling, so the order turns a legislative debate into litigation and funding pressure. The direction of federal policy is now clear, while the compliance cost of selling AI products across the US stays open until the first cases are decided.
Money is the sharper lever. Within 90 days the Commerce Department must publish a list of onerous state laws, and states on that list lose remaining non-deployment money from the BEAD broadband program as far as federal law allows. The order cites Colorado's ban on algorithmic discrimination and warns of a patchwork of 50 regimes that weighs hardest on start-ups. Washington reached for these tools after two failures in Congress: in July the Senate voted 99 to 1 to strip a ten-year moratorium on state AI laws from the budget bill, and this month the idea stayed out of the defense bill. The order also sets its own limits, since the federal framework it requests should leave state rules on child safety, data center infrastructure and state procurement of AI in place.
Pausing compliance work because of an executive order would be a mistake. Colorado's law and California's transparency law for frontier models stay on the books until a court or Congress says otherwise, and enterprise buyers will keep asking for the same documentation. The procurement carve-out leaves states free to set their own terms as buyers. The practical design is one compliance layer built around the strictest regime a company sells into, with testing and disclosure records that would also satisfy a federal standard. Teams that also serve the EU and Latin America go furthest with products designed to meet Europe's AI Act and a lighter US regime at once.
A quick statute would cut the contested period short, since only an act of Congress ends the patchwork. Whether Congress takes up the federal framework the order asks White House advisers to draft is the signal that settles it. Before that, the Commerce list will show whether Washington targets a handful of statutes or most state AI rules, and the task force's first suits will show which states stand and defend their laws.
US rules for AI are heading toward a lighter and more uniform standard, but the path runs through courts and appropriations rather than a single signature. Fanaura treats the stretch ahead as legal friction for companies selling into regulated US sectors and as an advantage for teams that build compliance into the product, the kind of founders we look for. While the law is contested, compliance that ships with the product sells.
Building an AI-native company? Apply in five minutes. →
More from Fanaura
Building AI-native products for regulated US sectors with compliance built in? Send us the project.
Fanaura invests its own capital and the capital of its limited partners (LPs), usually through a separate SPV for each deal. If we see a fit, we get in touch.
Get the news by email
One short note each weekday. No marketing, and you can unsubscribe at any time.