The EO, which Trump signed Thursday, has been much anticipated although widely criticized for infringing on states’ rights. Under the Trump administration, the federal government has sought to limit states’ regulatory authority more than once, through a now-struck moratorium in the budget bill and via a measure in defense policy legislation, both aimed at making the U.S. a leader in AI technology.
Experts say the EO, the latest such attempt, is illegal.
WHAT’S IN THE ORDER?
The order states that it is intended to prevent a patchwork of AI laws. However, some have argued that state-level AI laws actually clarify the rules for businesses that want to work with government.
The EO calls for a “carefully crafted national framework,” while also indicating that no such standard currently exists.
Next, it calls for the creation of an AI Litigation Task Force, to be established within 30 days. This task force is charged with challenging state AI laws that are inconsistent with the national policy framework for AI, although this does not yet exist. Other state AI laws that may be challenged are those that are believed to impact interstate commerce, those that are pre-empted by federal regulations or those that are “otherwise unlawful in the Attorney General’s judgment.”
Within 90 days of the EO, federal officials are required to publish an evaluation of existing state AI laws, highlighting those deemed to be in conflict with the yet-to-be-created national standard. Specifically, the EO calls for the identification of any laws that require AI models to “alter truthful outputs” or that may compel industry to disclose information that may violate the Constitution.
The EO also threatens to withhold congressionally approved funding for states that is intended to support broadband deployment through the Broadband Equity, Access and Deployment (BEAD) Program. BEAD was enacted through the bipartisan infrastructure law. A previous Government Technology analysis of state-level AI legislation revealed which states may be at risk of being impacted by the EO.
More specifically, the EO states that within 90 days, federal officials should enact a policy notice specifying the conditions under which states may be eligible for remaining BEAD funding. The policy notice must detail that states with “onerous AI laws” are ineligible for non-deployment funds — a subject of much contention recently, as the distribution of these funds is required by law.
The EO also calls for federal officials to start a proceeding — within 90 days — that would determine whether to adopt a federal reporting and disclosure standard for AI models.
It calls for a policy statement, to be issued within 90 days, explaining circumstances in which state laws requiring “alterations to the truthful outputs of AI models” are pre-empted by the Federal Trade Commission Act’s prohibition on engaging in deceptive acts. The EO calls out a Colorado law that aims to prevent algorithmic discrimination in high-risk systems, although this law does not include any provisions requiring the alteration of AI model outputs.
Finally, the EO calls on the special adviser for AI and crypto and the assistant to the president for science and technology to prepare a legislative recommendation establishing a federal policy framework for AI that pre-empts conflicting state AI laws. This pre-emption, the EO says, would not impact “lawful state AI laws” related to child safety protections, AI compute and data center infrastructure, state government procurement and use of AI, and “other topics as shall be determined.”
INDUSTRY REACTIONS
Initial reactions to this action are revealing.
U.S. Rep. Ted W. Lieu, co-chair of the U.S. House of Representatives’ bipartisan Task Force on AI, issued a statement underlining the related, bipartisan opposition to enacting such a ban, describing the EO as “unconstitutional”: “An executive order cannot create law and cannot preempt state authority.” This action, Lieu argued, will be challenged in courts.
“President Trump’s executive order doubles down on a dangerous policy that the Republican-led Congress has rejected not once, but twice: displacing states from their critical role in ensuring that AI is safe, trustworthy, and nondiscriminatory,” Cody Venzke, senior policy counsel with the American Civil Liberties Union, said in a statement. He emphasized the bipartisan opposition to these efforts. “Moreover, the executive order is not just dangerous, it’s unconstitutional.”
The unconstitutionality of this EO, as Venzke stated, lies in retroactively changing the conditions on federal grants to states, which the Supreme Court has clarified is not allowed.
Florida is a state with a Republican governor who has recently introduced legislation to protect consumers from AI costs. Gov. Ron DeSantis stated his opinion on the EO in a social media post Monday: “An executive order doesn’t/can’t preempt state legislative action.”
In a statement, Michael Kleinman, head of U.S. policy at the Future of Life Institute, argued that AI companies are “operating with impunity”: “No other industry operates without regulation and oversight, be it drug manufacturers or hair salons; basic safety measures are not just expected, but legally required.”
Common Sense Media’s founder and CEO James P. Steyer issued a statement noting that some of the U.S.’ “strongest federal consumer protections,” such as seat belt requirements, initially started as state laws.
A November report from the nonprofit organization Public Citizen analyzed the more than $1 billion in influence spending by Big Tech intended to “distort federal policy to prioritize the industry’s profits over protecting the public.”
A national poll released earlier this year found Republicans and Democrats agree that states should be able to pass AI regulations. This poll was related to an earlier budget bill provision that would have enacted a moratorium on state AI policymaking for 10 years; the new EO does not specify how long state AI policymaking will be limited.
“Accordingly, states should call the Administration’s bluff and not be dissuaded from pursuing AI legislation as they see fit,” the article said.
