Trump's AI Executive Order: Regulatory Uncertainty Ahead

On December 11, President Trump published an executive order titled Ensuring a National Policy Framework for Artificial Intelligence (the “Executive Order”). It forms some of the first concrete instruction regarding the Trump administration’s stance on AI. The order advances a "minimally burdensome" national framework while directly challenging state-level AI regulation, going a step further than July's AI Action Plan. Despite its ambitious scope, significant uncertainty remains about its practical impact on the AI landscape.

What the Executive Order Does

Creates an AI Litigation Task Force to challenge state AI laws inconsistent with federal policy and evaluate existing state laws for conflicts (implemented by the Attorney General on January 9, 2026)

• Restricts BEAD Program funding by making states with "onerous" AI laws ineligible for non-deployment funds under the Broadband Equity Access and Deployment Program
• Requires agencies to determine whether they may condition discretionary grants on states either not enacting conflicting AI laws or agreeing not to enforce existing ones
• Demands development of uniform Federal reporting and disclosure standards and policy frameworks
• Instructs the FTC to explain which state laws are preempted by the FTC’s prohibition on engaging in deceptive acts

Immediate Targets: State AI Laws

The Executive Order targets laws that involve “ideological bias,” an extension of the Administration’s actions on DEI. It specifically calls out the Colorado AI Act, C.R.S.A. § 6-1-1701, which bans algorithmic discrimination in high-risk AI systems [1] and requires AI developers to document training processes and discrimination mitigation efforts. Both regulators in other states and companies alike were looking for how this framework would function and how it might be adopted in other states. Instead, the Executive Order threatens to prevent such state-level frameworks from taking effect before they can be fully implemented and evaluated.

The Executive Order is also an extension of the Order President Trump released in July 2025, Preventing Woke AI in the Federal Government . That order required agencies to only procure LLMs that prioritized truth-seeking and ideological neutrality, and particularly those that do not encode DEI into their AI. However, AI developers aren’t the target of either order, as the Trump administration puts the onus on agencies and states to prioritize partisan AI models. However, for those developers Trump views as in conflict with the policy, there may be more push for agencies to prosecute.

It seems likely that the Colorado AI Act and other similar laws, like those in California and Texas, will see litigation challenges soon. In addition to litigation, the states themselves could lose parts of their BEAD funding, which is the program designed to fund better infrastructure for high-speed internet. [2]

What does this mean for AI use and development going forward?

It functionally keeps us in the same place we have been: the wild west of AI regulation. If the comprehensive AI laws already in place are put on hold for litigation or if states agree to not enforce those laws in exchange for access to funding, we will have to look to older laws not designed with AI in mind for governance. Consumers may look to data privacy laws for recovery, arguing that their personal data is being used against their wishes to train AI systems. They might look to copyright laws or established state consumer protection laws. Companies must prepare for enforcement under these alternative legal theories and recognize that compliance efforts cannot cease even if AI-specific statutes are blocked.

However, the Executive Order cannot overturn existing state law, only threaten it. Until Congress passes a law preempting state law in this field or the courts overturn a statute, companies still risk being held liable under the comprehensive AI statutes that have gone into effect. The Executive Order creates political pressure but does not immediately change legal obligations, making compliance under state law still the best option.

Practical Implications for Organizations

AI developers should be ready for continued ambiguity around disclosure requirements, particularly if their product is one that is designed to be used for the significant questions that arise in areas like employment, healthcare, financing, and legal services. In the short term, it is likely that AI-specific regulations will be slowed or stopped as states react to the Executive Order and litigation is initiated.

In the long term, companies should maintain comprehensive documentation regardless of regulatory direction. Whether future governance comes from state law, federal legislation, or litigation, thorough records will be essential. There’s no deadline on Trump’s demand for policy frameworks or uniform reporting standards, so it’s difficult to say when a federal framework might be enacted. Any eventual federal framework may lag behind rapid technological advancement.

Organizations cannot afford to wait for clarity. Compliance with existing state law remains the safest course despite political uncertainty. Keeping abreast of the changing landscape, as well as being flexible and ready to respond, will serve businesses well as we wait to see what litigation is brought, which states sign deals to not enforce any AI laws, or what Congress has to say on the issue.

Final Thoughts

The Executive Order promises a "national policy framework" but delivers regulatory fragmentation. Rather than resolving the tension between federal and state authority, it deepens it. Companies face a landscape where state laws remain enforceable but politically contested and federal frameworks remain undefined.

The timeline for regulatory resolution remains unpredictable, so the path forward demands strategic flexibility. Organizations must simultaneously comply with existing state regulations, prepare for potential federal frameworks, and defend against claims under general-purpose statutes. Documentation, adaptability, and continuous monitoring of legal developments are no longer best practices, but instead business necessities. Proactive preparation today will prove far more valuable than reactive scrambling when the regulatory dust finally settles.

[1] AI systems that make, or are a substantial factor in making, consequential decisions (employment, housing, financing, etc.).

[2] https://www.benton.org/blog/three-questions-about-bead-non-deployment-funding