On Thursday evening, with White House AI and crypto czar David Sacks looking over his shoulder, Donald Trump signed an executive order aiming to grab unilateral power over regulating artificial intelligence for the federal government. The order can’t by itself unilaterally override state AI laws, but it directs federal agencies to take steps to reduce or eliminate their influence, and discourage states from passing laws that the federal government might challenge, or put crucial funding for other programs at risk.
It specifically calls out Colorado’s recently passed consumer protection law, making the claim that “banning ‘algorithmic discrimination’ may even force AI models to produce false results in order to avoid a ‘differential treatment or impact’ on protected groups.”
The final order is largely the same as the draft copy we obtained and reported on last month. It directs the creation of an “AI Litigation Task Force” that the attorney general will direct, suing states over any AI laws that it determines are inconsistent with a goal to “sustain and enhance the United States’ global AI dominance through a minimally burdensome national policy framework for AI.”
The FTC is directed to issue a policy statement explaining “circumstances under which State laws that require alterations to the truthful outputs of AI models are preempted by the Federal Trade Commission Act’s prohibition on engaging in deceptive acts or practices affecting commerce.”
It also directs Commerce Secretary Howard Lutnick to, in the next 90 days, issue a report on states with laws deemed to be in conflict with the order, and determine which ones may become ineligible for rural broadband funding from the Broadband Equity Access and Deployment (BEAD) program. Meanwhile, FCC Chairman Brandon Carr is tasked with working on proceedings to “adopt a Federal reporting and disclosure standard for AI models that preempts conflicting State laws.”
Section 8 of the order contains a notable, and vague, carveout, claiming its recommendations don’t propose preempting “otherwise lawful State AI laws” covering child safety, AI compute and data infrastructure building, use of AI by state governments, and “other topics as shall be determined.”
In the past year alone, a mounting pile of bills submitted and laws passed by state governments across the country have attempted to place guardrails on artificial intelligence — a patchwork of laws, as Trump framed it, that the AI industry claims makes it extremely difficult to operate and innovate. In an ideal situation, Congress would handle this confusing patchwork by passing AI regulatory laws at the federal level, which would then preempt, or automatically override, any conflicting state laws.
The approach they’ve taken, however, has proven controversial. Citing the slow pace of Congress and the speed at which they need to pursue innovation, the AI industry and their political allies have instead pursued a ban, or moratorium, on states writing or enforcing their own AI laws. In the past year alone, Congress has twice attempted to pass a moratorium, and failed both times — first during debate over Trump’s Big Beautiful Bill, where it ended up dying in the Senate, and then failed during negotiations over the National Defense Authorization Act, which sets yearly defense spending.
But while Congress’s moratorium attempts drew criticism, it would have been constitutional had it been signed. The White House, however, is wading into murky territory. Several weeks ago, a draft executive order leaked that illustrated how the Trump administration would approach preemption: instead of trying to put guardrails on AI, they would instead attempt to punish states with “onerous” laws that contradicted whatever the White House wanted.
Policywatchers and political insiders noted that the draft order would have handed an exorbitant amount of power to billionaire venture capitalist David Sacks, who serves as the White House AI and Crypto czar, while cutting out crucial federal agencies and offices that would have normally been involved with setting tech policy. Though he is technically a provisional government employee, Sacks serves as a direct conduit between the president and Silicon Valley’s elite, and has gained a massive amount of influence over Trump’s tech policy — so much that Trump’s statements on AI, H1B visas, and chip sales have horrified the MAGA base.
In the past few weeks, however, Trump has begun signaling his willingness to sign an executive order that would override state laws. On Monday, he stated on Truth Social that the order would, theoretically, create “only One Rulebook” so that AI companies would not have to follow “50 states, many of them bad actors, involved in RULES and the APPROVAL PROCESS. THERE CAN BE NO DOUBT ABOUT THIS! AI WILL BE DESTROYED IN ITS INFANCY!”.
Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.
- AIClose
AI
Posts from this topic will be added to your daily email digest and your homepage feed.
FollowFollow
See All AI
- LawClose
Law
Posts from this topic will be added to your daily email digest and your homepage feed.
FollowFollow
See All Law
- PolicyClose
Policy
Posts from this topic will be added to your daily email digest and your homepage feed.
FollowFollow
See All Policy
- PoliticsClose
Politics
Posts from this topic will be added to your daily email digest and your homepage feed.
FollowFollow
See All Politics
