The White House Wants Federal Rules to Override State AI Laws

The Trump White House released a national AI legislative framework on March 20 with a clear ask to Congress: let federal rules govern AI and stop states from setting their own. The framework recommends preempting state laws that regulate AI model development, impose liability on developers for third-party misuse, or "unduly burden" AI activity that would be legal without AI.

States would still be allowed to enforce fraud and consumer protection laws, zone data centers, and regulate how their own governments use AI. But the preemption push, if enacted, would invalidate a substantial portion of what states have been doing for the past two years.

That is a lot of legislation to sweep aside. More than 600 state AI bills were introduced in the current session. New York's RAISE Act, signed into law in December 2025, requires large AI developers to publish safety protocols and report critical incidents within 72 hours. California's SB 53 imposes similar obligations. Both would fall directly in the crosshairs of the framework's preemption recommendation, since they regulate AI development, the category the framework targets for federal override.

The framework also explicitly rules out creating any new federal AI regulatory body. Existing agencies with relevant expertise would handle oversight in their own domains: the FTC for consumer protection, the FDA for medical AI, and so on. The administration's reasoning is that AI does not require new bureaucracy; it requires existing institutions to adapt.

One area where the framework goes further than most state bills is protection for children. It recommends age verification requirements for AI services accessible to minors, stronger parental controls, and tighter limits on using children's data to train models. That part of the framework has drawn less attention but is likely to find broader bipartisan support.

The framework is currently a recommendation, not law. Congress would have to act, and no implementing bill has been introduced. But it sets the administration's position clearly ahead of what is shaping up to be a significant federal-versus-state fight over who governs AI in the United States.