1. Introduction
On July 23, 2025, the Trump administration debuted its long-awaited AI Action Plan (the Plan), identifying a range of policies to be implemented across the federal government to "usher in a new golden age of human flourishing, economic competitiveness, and national security for the American people." Alongside the Plan, President Trump also released three executive orders (EOs) that (1) promote the export of "full-stack American AI technology packages," (2) facilitate the build-out of AI data centers and related infrastructure, and (3) direct federal agencies to procure "truth-seeking" and "ideologically neutral" large language models (LLMs).
Building off Executive Order 14179, "Removing Barriers to American Leadership in Artificial Intelligence," the Plan and accompanying executive orders signal the administration's more permissive approach to AI development and diffusion while still seeking to shape developer behavior through government contracting and other mechanisms. At the same time, they maintain continuity with certain Biden administration AI policies, including with respect to accelerating AI adoption across the federal government, streamlining permitting, and revitalizing domestic semiconductor manufacturing.
It is not yet clear how–or how quickly–the Plan and EOs will be implemented in practice, but the documents envision a variety of regulatory processes that will take time to manifest despite some rapid deadlines. This means there will likely be additional opportunities for strategic engagement from the private sector as these policies develop in areas such as government contracting, energy policy, federal funding, export controls, regulatory reform (including preemption of state law), safety and security standards, and environmental permitting, and organizations can begin to evaluate those opportunities now.
2. AI Action Plan
The Plan contains three pillars: accelerating innovation, building American AI infrastructure, and leading in international diplomacy and security.
Acceleration
The Plan's first pillar focuses on addressing regulatory barriers to domestic AI innovation. It proposes a range of policies to remove red tape and regulation, including a public Request for Information about federal regulations that set back AI innovation and adoption, and an effort led by the Office of Management and Budget to revise or repeal regulations, rules, and other policy documents that "unnecessarily hinder AI development or deployment." The Plan also targets states with AI regulations that are perceived as restrictive, for example by recommending that federal agencies consider a state's "AI regulatory climate" when making funding decisions.
This pillar also includes some recommendations to shape private sector development and use of AI, including proposed update to federal procurement guidelines to ensure the government contracts only with AI developers "who ensure that their systems are objective and free from top-down ideological bias" (a policy also enshrined in one of the three executive orders) and a suggested update to NIST's AI Risk Management Framework to remove references to DEI, misinformation, and climate change.
In addition, this pillar seeks to enable AI adoption within critical sectors, such as healthcare and agriculture, by promoting a "try-first culture." The Plan recommends establishing regulatory sandboxes where AI tools can be tested, launching industry specific initiatives to develop standards for AI, and measuring AI productivity in industry-specific tasks. Within government, the Plan recommends increased strategic coordination and collaboration, echoing similar ambitions from the Biden administration.
Infrastructure
The "infrastructure" pillar recommends steps to speed the build-out of AI infrastructure, continuing the Biden administration's focus on permitting and power issues. The Trump administration proposes to streamline data-center construction permitting and expand and optimize existing electrical grids to support those data centers. The Plan recommends achieving these aims with new categorical exclusions to the National Environmental Policy Act (NEPA) to cover data-center-related activities that do not typically involve significant effects on the environment and prioritize the development of new energy sources at the "technological frontier" (like nuclear fission and fusion). It also seeks to establish a Department of Homeland Security-led AI Information Sharing and Analysis Center for private sector cybersecurity guidance.
Diplomacy and Security
The final pillar aims to embed American AI technology and ideals abroad, while countering foreign (especially Chinese) influence. This includes meeting global demand for AI by exporting US hardware, models, software, applications, and standards to countries willing to join in the United States' "AI alliance"; leveraging existing US diplomatic positions to set AI standards that reflect American values; imposing new export controls on component sub-systems and covering loopholes in existing controls on semiconductors; and expanding US national security measures through the hiring of AI experts into government and the investment in biosecurity tools. The Plan does not, however, include any details on planned or forthcoming export controls on AI chips or equipment.
The Plan illustrates that the Trump administration's goal is not only to establish American dominance in AI capabilities and infrastructure, but also to define the values that govern how AI systems operate globally. The Plan also highlights the potential limits of the federal government's ability to codify these values through the levers of federal policy making, and the key opportunities for stakeholders to shape and contest the implementation of federal AI policy in the coming months.
3. Executive Orders
The following executive orders, also released July 23, 2025, seek to implement the roadmap described in the Plan.
Executive Order on "Accelerating Federal Permitting of Data Center Infrastructure"
This executive order directs the creation of new expedited review processes for certain large scale data center projects and the development of new incentives to encourage data center construction.
The EO directs various expedited review processes for data centers (1) with at least $500m in committed capital expenditures; (2) involving electric load addition of greater than 100MW; (3) that "protect national security;" and (4) any others designated as such by the Secretary of Defense, Secretary of Interior, Secretary of Commerce, or Secretary of Energy. For these "qualifying projects," the EO directs the relevant agencies to pursue various expedited permitting and environmental reviews, including by identifying categorical exclusions from National Environmental Policy Act review, designating them for expedited review under FAST-41 (the interagency process for improved federal agency coordination on infrastructure project environmental review), expediting permitting on federal and nonfederal lands through new regulations under federal environmental laws including the Clean Air Act, Clean Water Act, and Toxic substances Act, and accelerating agency consultation under the Endangered Species Act.
In addition, the EO seeks to provide expanded access to federal lands for construction of qualifying data centers and offer whole-of-government financial support. The Departments of Interior and Energy are directed to offer appropriate authorizations for construction of data centers on federal lands identified by the Secretary of the Interior and Secretary of the Energy. The Secretary of Defense is also tasked with identifying suitable sites on military installations and competitively lease those lands for qualifying projects. The Commerce Department and the Office of Science and Technology are directed to launch an initiative to provide financial support (including loans, grants, and tax incentives, among others) to qualifying data centers, and relevant agencies are further directed to submit proposals for existing financial support that can be granted to projects, "consistent with the protection of national security."
Executive Order on "Promoting the Export of the American AI Technology Stack"
This executive order launches a strategic effort to promote the export of comprehensive, US-developed AI "technology packages." Recognizing AI as a critical driver of future economic and national security outcomes, the order highlights the need for the United States to lead not only in developing frontier AI capabilities but also in shaping global standards and governance. As such, it directs the Secretary of Commerce—with input from other key federal agencies and offices—to create the American AI Exports Program. The Commerce Department will solicit proposals from an "industry-led consortia" for "full-stack AI technology packages" to develop and promote for export. Alongside their proposals for AI models and supporting technologies to export, applicants must identify target export regions, explain operational models, and ensure compliance with US export laws and regulations, among numerous other requirements. The order lays out a process for evaluating and supporting proposals that demonstrate both technological readiness and international alignment. Selected proposals will be designated as "priority AI export packages" and gain access to a range of federal resources, such as financial support and policy coordination.
To further support this initiative, the EO empowers the Economic Diplomacy Action Group to coordinate the mobilization of federal financing tools, including loans, investment guarantees, and technical assistance. The Secretary of State is directed to lead diplomatic efforts that promote favorable regulatory and investment environments in partner countries, while also working to remove trade and market access barriers to help increase "the export of American AI technologies and standards." The Small Business Administration and the State Department will actively support small businesses to encourage broader participation in AI development.
Executive Order on "Preventing Woke AI in the Federal Government"
This executive order establishes new procurement requirements for AI systems used by the federal government. Specifically, all large language models (LLMs) procured by the federal government must comply with two "Unbiased AI Principles": (1) "truth-seeking", which prioritizes "historical accuracy, scientific inquiry, and objectivity" and acknowledges any uncertainty when relevant; and (2) ideological neutrality, which requires that LLMs function as "nonpartisan tools that do not manipulate responses in favor of ideological dogmas" like DEI.
The Office of Management and Budget, in consultation with the Administrator for Federal Procurement Policy, the Administrator of General Services, and the Director of the Office of Science and Technology Policy, are directed to issue guidance to agencies to implement the EO. Agency heads shall, in turn, implement the Unbiased AI principles in their current and future contracts.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
