On July 23, 2025, the White House released its artificial intelligence action plan, “Winning the AI Race: America’s AI Action Plan,” which follows from President Donald Trump’s AI Executive Order 14179, signed in January. The Plan sets out the Administration’s policy recommendations for the U.S. to achieve “global AI dominance.” To this end, it focuses on three key pillars: Accelerate AI Innovation, Build American AI Infrastructure and Lead in International AI Diplomacy and Security. The Plan mainly lays out high-level policy recommendations for each pillar and provides some insight into the administration’s AI priorities.

Pillar I: Accelerate AI Innovation 

The first pillar of the Plan aims to create an environment that favors AI innovation. The Plan encourages the withholding of AI-related federal funds from states with onerous AI regulations, working with federal agencies to revise or repeal laws and guidance that might hinder AI development or adoption, and requires a review of past agency investigations and orders that may be inconsistent with unhindered innovation. This pillar also stresses the importance of protecting free speech, the development of open-source AI models, rapid adoption of AI in higher-risk sectors (such as health care) and in government, empowering workers to embrace AI in the workplace and capitalizing on manufacturing and scientific opportunities arising from AI innovation. This section of the Plan outlines some key priorities for AI development, including a focus on interoperability and transparency of AI and developing guidelines and processes for assessing and evaluating AI systems. There is also an acknowledgment that AI can pose certain risks to the broader public, such as enabling malicious deepfakes or AI-generated fake evidence. The Plan suggests arming courts and law enforcement with new tools and guidance for overcoming these specific challenges.

Pillar II: Build American AI Infrastructure

The second pillar of the Plan focuses on energy and other infrastructure needed to support large-scale AI development. It outlines recommendations for streamlining environmental permitting processes and other regulatory hurdles, suggests upgrading the electrical grid to support increased energy demands, and seeks to ramp up semiconductor manufacturing and data center construction (including by training workers for the skills necessary to support these projects). The Plan also notes the importance of emphasizing infrastructure security and secure development practices for AI technologies.

Pillar III: Lead in International AI Diplomacy and Security

Finally, the Plan’s third pillar highlights the Administration’s focus on positioning the U.S. as a dominant leader in the AI space, as well as protecting U.S. innovation from our adversaries (and specifically, China). The Plan recommends policies aimed at making the U.S. a major exporter of AI technologies to our allies and promoting U.S. influence on AI policy and adoption globally. At the same time, the Plan also aims to strengthen export controls for AI and related technologies, close loopholes that might be exploited by our adversaries and prepare the U.S. government to evaluate and counter novel security risks posed by AI, such as sophisticated cyber threats or other novel types of attacks. 

Key Takeaways

• Pillar I of the Plan focuses heavily on eliminating barriers and regulatory red tape to the development and adoption of AI. The Office of Management and Budget (OMB) is given a specific directive to eliminate, repeal or revise regulations, rules and the like that may unnecessarily hinder AI development or deployment. We already saw OMB revise its policies on federal agencies’ use and procurement of AI earlier this year with M-25-21 (“Accelerating Federal Use of AI through Innovation, Governance, and Public Trust”) and M-25-22 (“Driving Efficient Acquisition of Artificial Intelligence in Government”). With this new AI Action Plan, we should expect to see more (and potentially more drastic) changes from OMB.
   
• The Plan asks the Federal Communications Commission (FCC) to evaluate whether state AI regulations interfere with the agency’s ability to carry out its obligations and authorities under the Communications Act of 1934.  This means that the FCC may become a more prominent player in AI (de)regulation and could potentially affect how (or to what extent) states are able to pass or enforce their own AI-related laws. 
   
• The Plan also instructs the Federal Trade Commission (FTC) to review previously settled investigations to ensure they do not “advance theories of liability that unduly burden AI innovation.” The Plan quite literally directs the FTC to reopen or relitigate previously settled matters to the extent they advance theories of liability that may “unduly burden AI innovation.” This directive is particularly interesting since just this year the FTC pursued enforcement against Cleo AI and Workado. It’s also possible that the Commission will decide to review previous AI cases brought under the last administration’s “Operation AI Comply” initiative.  
   
• While the Plan mainly focuses on three pillars (summarized above), it’s also important to recognize what it does not focus on. Unlike the Biden AI Executive Order, the Plan does not include much detail on how AI’s use and adoption could affect everyday consumers. Unfortunately, this means that there is not much more clarity on what the federal government believes businesses should be doing to combat common concerns like bias, discrimination or unfairness. Instead, the Plan focuses on combatting “ideological bias” in AI models and designing AI systems to “pursue objective truth rather than social engineering agendas.”