Key Takeaways
- The Trump Administration’s AI Action Plan lays out a sweeping vision of America’s AI future, as well as the strategic and policy steps needed to enable the U.S. to “innovate faster and more comprehensively than [its] competitors.”[1] It emphasizes the central role of private industry in AI’s success and continues the Trump Administration’s de-regulatory push. The Plan sets forth more than 100 discrete recommendations across 30 policy goals, focused around three policy pillars: (1) Accelerating AI Innovation; (2) Building AI Infrastructure; and (3) Leading in International AI Diplomacy and Security.
- The AI Action Plan casts a shadow over state regulatory efforts, threatening to withdraw federal funds from states that adopt “burdensome AI regulations.” On the international level, the Plan envisions an “enduring global alliance” designed to prevent America’s “adversaries from free-riding on our innovation and investment.”
- The AI Action Plan was followed quickly by three Executive Orders on topics raised in the Plan: (1) data center infrastructure; (2) “woke” AI in federal systems; and (3) promoting AI exports. The Plan has been met with broad industry support, but has raised concerns from other observers that it does not provide sufficient safeguards to prevent bias, theft of intellectual property, and other potential harms, and many key details on implementation and execution of the Plan remain unresolved.
The AI Action Plan
Background
On July 23, 2025, the White House released Winning the Race: America’s AI Action Plan (the “AI Action Plan”),[2] accompanied by three Executive Orders directing various federal agencies to implement elements of the AI Action Plan.[3] President Trump directed the creation of the Plan in Executive Order 14179, which he signed just a few days after his second inauguration, to “clear[] a path for the United States to act decisively to retain global leadership in artificial intelligence.”[4]
The AI Action Plan carries forward many of the same strategic objectives identified by the Biden Administration in its work on AI—for example, harnessing AI in the interest of national security (including by maintaining the U.S. lead over China with respect to frontier model development), promoting the adoption of AI throughout the U.S. economy, and building out AI infrastructure over the long term. But the policy choices are markedly different. In particular, the AI Action Plan—along with the concurrently issued AI Executive Orders—calls for streamlining of federal and state regulation in order to promote a more widespread adoption of U.S. AI technologies. It also embraces open-source and open-weight AI models and directs federal procurement of AI systems that are “objective and free from top-down ideological bias.”[5]
Objectives of the AI Action Plan
Using the language of “retaining dominance” and “winning” the global AI race, the AI Action Plan establishes a policy road map organized around three pillars: (1) accelerating AI innovation, (2) building AI infrastructure, and (3) leading in AI diplomacy and security.
- Accelerate AI Innovation—The plan details policy recommendations for the federal government to “create the conditions where private-sector-led innovation can flourish.”[6] Specifically:
- removing federal regulations that hamper AI development and using the government’s levers (e.g., procurement) to enable AI adoption across government and the broader market;
- reviewing, repealing and/or revising Federal Trade Commission (“FTC”) final orders and consent decrees based on theories of liability deemed unduly burdensome to AI innovation;
- advancing AI science, including the creation of world-class AI-ready scientific datasets, and promoting the use of open-source and open-weight AI models;
- influencing state-level policy by withholding AI-related funding to states whose “AI regulatory regimes may hinder the effectiveness of that funding or award,”[7] mirroring a failed amendment to the One Big Beautiful Bill Act that would have placed a 10-year enforcement moratorium on state laws or regulations related to AI models, systems, or automated decision-making;
- directing the Federal Communications Commission (“FCC”) to evaluate whether state AI regulations may interfere with federal law; and
- building an ecosystem to evaluate and assess the performance and reliability of AI systems, including through public-private partnerships to invest in “AI testbeds for piloting AI systems in secure, real-world settings,” and other approaches such as coordinating an “AI hackathon initiative,” in partnership with academia and others, to test for system effectiveness, transparency, and security vulnerabilities.[8]
- Build American AI Infrastructure—The plan identifies measures to address the “stagnation” of American energy capacity, in particular by comparison to the rapidly built Chinese energy grid. Specifically:
- streamlining the permitting process to build data centers, semiconductor manufacturing facilities, and energy infrastructure, and otherwise developing a comprehensive strategy to expand the U.S. power grid;
- promoting worker (re-)training to address shortages in critical jobs that underpin AI infrastructure; and
- bolstering the cybersecurity of critical infrastructure, including public-private collaboration to share known vulnerabilities, and incorporating AI-related issues into incident response plans as a matter of industry-standard best practice.
- Lead in International AI Diplomacy and Security—The plan stresses that dominance in global AI competition is contingent on the more widespread adoption of U.S. AI systems, computing hardware, and standards. Some of the policy recommendations here include:
- exporting the full AI technology stack “to all countries willing to join America’s AI alliance,”[9] strengthening export control enforcement to deny foreign adversaries access to AI resources, and plugging existing loopholes in export controls related to semiconductor manufacturing;
- countering Chinese influence in international governance bodies (e.g., the United Nations, Organization for Economic Co-operation and Development, G7), which the AI Action Plan states has resulted in burdensome regulations, vague codes of conduct, or other standards that have promoted Chinese companies’ interests; and
- ensuring U.S. preeminence in evaluating national security risks in frontier models, including by building national security-related AI evaluations and, in collaboration with AI frontier model developers, evaluating those systems for national security risks.
Others in the AI Race
The AI Action Plan enters an already crowded field of legislative and regulatory developments related to AI, both internationally and at the state level. For example, in May 2024, Colorado became the first U.S. state to enact a comprehensive AI law, and since then a number of other states—from California to Tennessee—have enacted or are considering general or specific laws regulating various aspects of AI governance and/or limited uses of AI (e.g., deep fakes, misinformation, election integrity).[10] Moreover, several state attorneys general have explained that, even where an AI law is not on the books, their offices still consider AI-related wrongdoers as subject to established consumer protection laws (e.g., regulating unfair or deceptive acts or practices).[11]
Outside the U.S., the EU has already made headlines for its enactment of the AI Act, which the Trump Administration has criticized for being unduly burdensome and an unfair trade barrier to U.S. tech companies.[12] More recently, just days after the AI Action Plan was unveiled, the Chinese government released its own global action plan for AI, calling for the establishment of a multinational World Artificial Intelligence Cooperation Organization. If created, this would pit China’s policy interests and technological solutions against those of the U.S.[13]
What’s Next?
The AI Action Plan and associated Executive Orders reposition federal AI policy to focus on the Trump Administration’s de-regulatory, pro-export, and anti-ideological bias goals. However, the details of exactly how the AI Action Plan will be implemented will be critical, and much work remains. Policy differences between the U.S., China, and the EU are also likely to drive regulatory uncertainty at the international level, and it remains to be seen how or whether the Trump Administration may challenge states like Colorado and California that have implemented AI-specific laws that increase the regulatory burden. As federal agencies move forward with implementing the AI Action Plan, companies will need to stay abreast of developments and should continue to monitor other reactions from U.S. state and international governments.
Click here to download this article.
[1] Executive Office of the President, “AI Action Plan” (July 2025), available here.
[2] Id.
[3] See Exec. Order 14318, “Accelerating Federal Permitting of Data Center Infrastructure,” 90 Fed. Reg. 35385 (Jul. 23, 2025), available here; Order 14319, “Preventing Woke AI in the Federal Government,” 90 Fed. Reg. 35389, available here; Order 14320, “Promoting the Export of the American AI Technology Stack,” 90 Fed. Reg. 35393, available here [hereinafter, collectively, the “AI Executive Orders”].
[4] Exec. Order 14179, “Removing Barriers to American Leadership in Artificial Intelligence,” 90 Fed. Reg. 8741, 8741 (Jan 23, 2025), available here.
[5] AI Action Plan, at 4; see also the AI Executive Orders, at footnote 2.
[6] Id. at 3.
[7] Id. at 3.
[8] Id. at 10.
[9] Id. at 20.
[10] See, e.g., Willkie Client Alert, “U.S. State AI Update” (May 8, 2024), available here; Willkie Client Alert, “California Enacts 17 AI Bills in 2024” (Oct. 10, 2024), available here; IAPP “U.S. State AI Governance Legislation Tracker” (last visited July 28, 2025), available here.
[11] See, e.g., Cal. Dep’t of Justice, “Attorney General Bonta Issues Legal Advisories on the Application of California Law to AI” (Jan. 13, 2025), available here; Mass. Office of the Att’y Gen., “Attorney General Advisory on the Application of the Commonwealth’s Consumer Protection, Civil Rights, and Data Privacy Laws to Artificial Intelligence” (April 16, 2024), available here.
[12] See J.D. Vance, “Remarks by the Vice President at the Artificial Intelligence Action Summit in Paris, France” (Feb. 11, 2024), available here.
[13] Action Plan on Global Governance of Artificial Intelligence, Xinhua News Agency (July 26, 2025), available here.
