On March 20, 2026, the White House published a four-page document with “Legislative Recommendations” in its National Policy Framework for Artificial Intelligence (the “AI Framework”). The AI Framework does not include specific draft legislation or an executive order, but instead contains recommendations for Congress, setting out the administration’s vision for a comprehensive federal AI legislative package. The AI Framework is not legally binding either for on Congress or on private sector companies. The AI Framework, building on Executive Order 14365, outlines eight key policy areas for federal AI legislation aimed at preempting restrictive state laws and bolstering AI innovation.
Background
The AI Framework represents the latest significant step in the Trump administration’s technology agenda and is consistent with, and builds on, its past actions regarding the national AI strategy going back to the very first days of President Trump’s second term. Within the first week of returning to the presidency, President Trump revoked the Biden-era Executive Order 14110 on “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence”, which he swiftly replaced with Executive Order 14179 on “Removing Barriers to American Leadership in Artificial Intelligence”. Executive Order 14179 established the national AI policy to “sustain and enhance America’s global AI dominance in order to promote human flourishing, economic competitiveness, and national security,” but provided few specifics.
The Trump administration’s AI regulatory vision came into sharper focus in the months following Executive Order 14179. The notion of pulling on federal levers to mitigate state AI regulation first emerged in negotiations over the President’s “One Big Beautiful Bill Act”, but the proposed 10-year moratorium was removed before the law ultimately passed.
Then, following consultation with stakeholders, the White House unveiled its AI Action Plan in July, offering a more comprehensive roadmap for a federal AI strategy. The AI Action Plan notably called for the removal of regulatory barriers to AI innovation, chiding “states with burdensome AI regulations” while cautioning that the federal government “should also not interfere with states’ rights to pass prudent laws that are not unduly restrictive to innovation.”
The AI Action Plan was followed in December by Executive Order 14365 on “Ensuring A National Policy Framework For Artificial Intelligence”, which expands on the AI Action Plan’s roadmap with specific actions and directed federal agencies, including the Department of Commerce and the Federal Communications Commission, to take actions against states with “onerous AI laws” and to prepare an evaluation of particular state laws. Several states have enacted laws regulating artificial intelligence or let them go into effect before and after Executive Order 14365 was issued, including California, Colorado, Texas, New York, and Utah.
Executive Order 14365 also directed the Special Advisor for AI and Crypto and the Assistant to the President for Science and Technology to prepare a uniform legislative proposal for the regulation of AI that would specifically preempt state AI laws deemed burdensome by the administration. It further called on the Department of Justice to establish an AI Litigation Taskforce to challenge state laws that are inconsistent with national AI priorities; the Justice Department announced in January that it had in fact convened this Taskforce.
The AI Framework
The AI Framework presents legislative recommendations across eight distinct areas, advancing themes present in the administration’s earlier pronouncements on AI. These recurrent themes include preserving American dominance in innovation, balancing AI regulation with First Amendment concerns, avoiding burdensome regulatory patchworks, examining the impacts of AI on the American workforce, and moving away from the previous administration’s focus on AI safety to one that prioritizes innovation. In addition to these established themes, the AI Framework expands on the children’s safety aspects of the federal AI strategy, provides new detail on the administration’s positions on intellectual property aspects of AI, and addresses the impact of AI on energy affordability.
Protecting Children and Empowering Parents. The AI Framework calls on Congress to strengthen protections for children using AI services while equipping parents with effective control tools. Recommended measures include age‑assurance mechanisms, safeguards against sexual exploitation and self‑harm, application of existing child privacy laws to AI (including limits on data use for training and advertising), and the avoidance of vague content standards that could spur excessive litigation. It also urges preservation of states’ ability to enforce generally applicable child‑protection laws, which has been an area of intensive statehouse legislative activity over the past couple years.
Strengthening Communities and Infrastructure. The AI Framework emphasizes that AI development and infrastructure should promote and enhance energy grid reliability while protecting communities from harm. The AI Framework urges Congress to ensure that AI data center construction and operation does not result in increased residential electricity costs from AI data centers and to streamline permitting to accelerate AI infrastructure buildout so that on-site and behind-the-meter power generation options are more accessible and attractive to AI developers. On AI safety, the AI Framework directs Congress to support law enforcement efforts to combat AI‑enabled fraud and impersonation scams and bolster national security agencies’ technical understanding of frontier AI. The AI Framework also advocates for the expansion of grants, technical assistance programs, and tax incentives to facilitate small business adoption of AI tools.
Respecting Intellectual Property Rights and Supporting Creators. The framework expresses the administration’s policy position that the training of AI models on copyrighted material does not constitute copyright infringement. While acknowledging the debate between AI developers and content owners on this very issue — which has been the focal point of a recent pre-publication report by the Copyright Office and remains the subject of ongoing multi-district litigation — the AI Framework recommends allowing courts to resolve the ultimate question of whether such training constitutes fair use. Consequently, it advises against legislative interference on this issue. Yet, to address potential compensation models for content owners without resolving the underlying liability question, the AI Framework suggests exploration of a collective licensing or rights‑management framework that avoids antitrust liability. The framework does not propose a specific mechanism for such licensing, nor does it address the threshold question of whether or when such licensing would be required if training is ultimately deemed non-infringing. Additionally, the framework calls for federal protections against unauthorized AI‑generated digital replicas of individuals that are consistent with the First Amendment.
Preventing Censorship and Protecting Free Speech. The AI Framework argues that AI policy must uphold First Amendment values by preventing government coercion of AI or technology providers to suppress or manipulate lawful expression. It recommends creating effective avenues for individuals to seek redress when federal agencies attempt to censor speech or dictate AI‑generated information.
Enabling Innovation and Ensuring American AI Dominance. The AI Framework urges Congress to remove barriers to AI innovation by establishing regulatory sandboxes, expanding access to federal datasets in AI‑ready formats, and relying on existing sector‑specific regulators and industry‑led standards rather than creating a new federal AI regulator. The approach endorsed by the AI Framework focuses on accelerating deployment of AI applications across the economy while preserving flexibility for rapid technological advancement.
Educating Americans and Developing an AI‑Ready Workforce. The AI Framework stresses that workers must share in AI‑driven growth through education and skills development. It recommends legislative action to incorporate AI training into existing educational and vocational programs, expanding federal research on AI‑driven workforce shifts, and strengthening land‑grant institutions’ capacity to provide technical assistance, demonstration projects, and youth AI development initiatives.
Preempting State AI Laws. Finally, the AI Framework calls for a consistent nationwide AI policy that preempts state laws imposing undue burdens, while respecting core principles of federalism. States would retain authority over certain domains, like general consumer protection, fraud, child protection, zoning, and their own governmental use of AI, but would be precluded from regulating AI development itself, penalizing developers for third‑party misuse, or restricting lawful AI use in ways that undermine national competitiveness and U.S. leadership in AI.
Takeaways
Building on their prior actions in the AI regulatory space, the AI Framework provides new insight into the administration’s AI legislative vision. Nevertheless, the success of any legislation remains uncertain in an election year.
Businesses that develop, contract for, or deploy AI tools should continue to build structured, flexible AI governance programs that enable their organizations to adapt quickly to a globally changing legal and regulatory environment and the corresponding compliance burdens. The flexibility of this governance should not only seek to be adaptable to the changing legal and regulatory environment but also empower companies deploying governance to realize the benefits of these transformative new technologies. It is important to monitor both legislative developments as well as caselaw updates (for example, the AI Framework’s deferring to the courts to resolve active IP litigation).
