White House pushes states to pause AI rules

The White House has been quietly telling Republican lawmakers in Nebraska and Tennessee to weaken or drop state artificial intelligence bills, according to Axios, even though those bills came from the president’s own party. The pressure campaign landed just as states were moving faster than Congress on rules for chatbots, deepfakes, and automated decisions. (axios.com) What the administration wants instead is one national approach. A White House artificial intelligence policy framework released on April 9 says Congress should create a single federal set of rules rather than leave companies to navigate 50 different state systems. (fabbs.org) That sounds simple if you run a big company. A bank, hospital chain, or software firm would rather fill out one form and follow one playbook than rebuild the same product for Tennessee, Nebraska, Illinois, and California one state at a time. (fabbs.org) The problem is that Congress has not passed that one national rulebook. Politico reported on April 3 that the White House push to block state rules was already running into skepticism on Capitol Hill, which means the federal substitute may not arrive soon. (politico.com) So the real choice is not “state rules or federal rules.” Right now the real choice is “state rules now” or “a gap where companies write their own standards while politicians argue.” (politico.com) (fabbs.org) That gap is not small. The National Conference of State Legislatures said state lawmakers introduced more than 1,000 artificial intelligence measures in 2025, after all 50 states had already introduced bills and 38 states adopted or enacted around 100 measures that year. (ncsl.org 1) (ncsl.org 2) States moved because artificial intelligence problems show up in very ordinary places. Legislatures have been writing bills on elections, hiring, health care, education, housing, government use, and fake media, which is why statehouses have become the first responders for this issue. (ncsl.org 1) (ncsl.org 2) The White House argument is that a patchwork of state laws can choke development before the United States settles on a national strategy. The administration framework says federal agencies should use existing powers, keep barriers to innovation low, and expand access to artificial intelligence tools. (fabbs.org) But lighter regulation from Washington does not erase risk inside a company. A Crain’s Chicago Business commentary published on March 30 argued that firms still need boards, executives, and product teams to decide where artificial intelligence is too risky, because the damage from a bad model can arrive before any regulator does. (chicagobusiness.com) That means a company can win the policy fight and still lose the real one. If its hiring tool screens out older workers, if its customer service bot invents legal advice, or if its image generator spreads a fake campaign ad, the first blow may come from lawsuits, clients, or headlines rather than from a state agency. (ncsl.org) (chicagobusiness.com) So this White House push is not just about who writes the rules. It is also about who carries the risk during the months or years before Congress acts: governors and state attorneys general on one side, or corporate boards and in-house lawyers on the other. (axios.com) (politico.com)