White House nudges GOP on AI rules
The White House is pressing Republican-led states to avoid creating their own AI regulations, seeking fewer state-level guardrails and a more unified federal approach. That push shifts regulatory risk for companies from complying with many different state laws to navigating a politically charged national debate about pre-emption and future enforcement. (axios.com)
The fight over artificial intelligence rules is moving to an unusual place: Republican statehouses are writing bills, and a Republican White House is telling them to slow down or drop them. Axios reported on April 9 that the pressure campaign has focused on Nebraska and Tennessee. (axios.com) The White House is not arguing that artificial intelligence needs no rules at all. It is arguing that 50 different state rulebooks would be harder for companies to follow than one national standard, and a December 2025 executive order said the federal government should pursue a “minimally burdensome” framework instead. (whitehouse.gov) That same executive order told the Justice Department to create an Artificial Intelligence Litigation Task Force within 30 days to challenge state laws the administration sees as inconsistent with federal policy. It also said some state rules could be attacked under the Federal Trade Commission Act or the Constitution’s limits on states regulating interstate commerce. (whitehouse.gov) The administration then turned that broad goal into a legislative pitch on March 20, 2026. The White House said Congress should build a national framework for artificial intelligence, and White House-aligned summaries said preemption of conflicting state laws was one of the central pieces. (whitehouse.gov, (dlapiper.com)) States moved first because Congress still has not passed a comprehensive artificial intelligence law. Law firms and state trackers show a fast-growing patchwork instead, with states like California, Colorado, Texas, and Utah adopting their own rules in 2025 while dozens of other proposals spread across the country. (venable.com, (ai-law-center.orrick.com), (transparencycoalition.ai)) That patchwork is not one single kind of law. Some state bills target election deepfakes, some cover hiring tools, some cover healthcare uses, and some try to stop what Colorado law calls “algorithmic discrimination,” which is the use of automated systems in ways that create unlawful unequal treatment. (venable.com, (whitehouse.gov)) For technology companies, the old risk was operational: one product might need one set of disclosures in California, another testing regime in Colorado, and a different liability rule somewhere else. The new risk is political and legal: if Washington blocks state rules without replacing them quickly, companies get fewer immediate guardrails but more uncertainty about what Congress, regulators, or courts will do next. (whitehouse.gov, (politico.com)) Congress is the weak link in the White House plan. Politico reported on April 3 that the administration’s push to curb state artificial intelligence laws is already running into skepticism from Democrats, which makes a broad national bill much harder even with Republicans controlling Washington. (politico.com)) That leaves Republican governors and legislators in a bind. If they keep writing state rules, they risk crossing a White House from their own party; if they stop, they may have to explain why they shelved bills on deepfakes, children, hiring, or healthcare while Congress is still deadlocked. (axios.com, (transparencycoalition.ai)) The immediate story is Nebraska and Tennessee. The larger one is a race between two clocks: state legislatures are writing rules now, while the White House is betting it can freeze that state-by-state scramble long enough to force a single federal deal later. (axios.com, (whitehouse.gov))
