Federal Government Pushes to Override State AI Laws
President Trump signed Executive Order 14365 on December 11, 2025, directing the federal government to establish a national AI policy framework. The order explicitly targets state regulations, proposing federal preemption of laws the White House deems burdensome to AI development and national competitiveness.
The White House released its draft framework on March 20, 2026. It calls for Congress to preempt state AI laws that impose what it characterizes as "undue burdens," arguing that a single federal standard would prevent a "fragmented patchwork" of state rules.
What the Federal Framework Would and Wouldn't Preempt
The White House proposal carves out limited exceptions. States would retain authority over:
- Laws protecting children, preventing fraud, and safeguarding consumers
- Zoning and placement of AI infrastructure
- Their own use of AI in procurement, law enforcement, and public education
But the framework would prevent states from regulating AI development itself, citing national security and foreign policy concerns. States could not penalize AI developers for how third parties misuse their models, nor impose restrictions on lawful uses of AI.
Congress Moves Quickly
Two competing bills have emerged from Republican leadership. Senator Ted Cruz's SANDBOX Act creates a regulatory waiver system allowing AI developers to modify or suspend rules blocking their work. Senator Marsha Blackburn's TRUMP AMERICA AI Act takes a harder line, preempting state laws on frontier AI and digital replicas while enabling the FTC and state attorneys general to sue developers for product defects.
House Republicans issued a joint statement pledging bipartisan work on a national framework. White House adviser David Sacks told Bloomberg Government that Congress could pass AI legislation within months.
States Aren't Backing Down
State activity contradicts the federal push for uniformity. Over 40 states introduced roughly 250 AI-related bills in 2025, covering government use of AI systems. California, New York, and Texas have already enacted comprehensive laws on AI transparency and consumer safety. Colorado and Utah passed legislation addressing algorithmic discrimination and mental health AI applications.
The National Governors Association launched a working group in January 2026 to develop a roadmap on AI and workforce issues. After the White House released its framework, the NGA signaled governors intend to use executive orders and legislation tailored to their states, citing concerns about centralizing AI regulation and the importance of local governments addressing emerging technology risks.
Industry Wants One Rule, Not Fifty
Technology companies broadly support federal preemption, but their positions diverge sharply. Large incumbents like Microsoft and Google often support some regulation, including licensing for powerful AI tools. Startups favor minimal rules to avoid compliance costs that could drain resources before profitability. Open-source advocates argue for accessible AI to prevent power concentration; others warn that releasing powerful models poses safety risks.
All sides agree on wanting regulatory certainty, a risk-based approach, and global competitiveness. Lobbying on AI surged to more than 640 companies engaging federally in 2024, a 141% jump from the prior year.
The Political Reality
Congress faces a genuine tension: overriding state laws means contradicting decisions made by state legislatures and governors. This is an election year with a compressed legislative calendar, making passage uncertain despite White House urgency.
The outcome will depend on whether enough members of Congress prioritize national AI competitiveness over state authority-and whether they can move fast enough to pass legislation before the session ends.
Your membership also unlocks:
