The White House wants to write one rulebook for AI and erase the other 50.

The Summary

  • The Trump administration asked Congress to preempt state AI laws, pushing a single federal framework it calls "One Rulebook" to replace what it sees as a patchwork of state regulations.
  • The plan covers child safety, data center energy offsets, and agency AI capacity, but its real aim is killing state-level guardrails before they scale.
  • Getting Congress to pass sweeping AI legislation in an election year is unlikely, but federal preemption could still freeze state action while the debate drags on.

The Signal

This is not about simplifying compliance. It is about who gets to set the rules for the agent economy before it fully materializes. States like California, Colorado, and New York have been moving fast on AI regulation, targeting everything from algorithmic bias to disclosure requirements for AI-generated content. Some of these laws are clumsy. Some are thoughtful. All of them represent local experimentation with governance models for technology that Washington has mostly ignored.

The White House framework, pushed by AI czar David Sacks, argues that 50 different regulatory regimes will "stifle innovation" and hurt U.S. competitiveness. That is the industry's preferred framing. What it actually means is that companies building autonomous agents, deploying models at scale, and racing toward AGI do not want to navigate different rules in different markets. They want one set of rules, ideally light ones, written at the federal level where industry has more concentrated lobbying power.

The framework includes provisions on child safety and energy offsets for data centers, issues with genuine bipartisan traction. But those are sweeteners. The core ask is preemption, the legal doctrine that lets federal law override state law. If Congress grants it, states lose the ability to regulate AI more strictly than Washington does. Given how slowly Congress moves and how captured it is by Big Tech money, that means the baseline becomes the ceiling.

The timing matters. Data center buildout is now a political flashpoint. Communities are pushing back on power grid strain, water use, and the economic disruption of AI infrastructure landing in their backyards. State lawmakers have been responsive because they are closer to the consequences. A federal preemption play pulls that power away, right as the agent economy starts consuming real resources at scale.

The Implication

If you are building AI products, watch whether Congress actually moves on this. Preemption language could show up in must-pass bills as a rider, not a standalone law. If you are in a state experimenting with AI governance, this is a direct challenge to local control. The White House is betting that "innovation" rhetoric will override concerns about accountability, and in Washington, that bet has historically paid off. The question is whether states fight back or fold.

Source: Fast Company Tech