The Trump administration's AI Action Plan marks a clear change in how the federal government wants to shape artificial intelligence. The plan, published Wednesday, moves away from former President Biden's more cautious posture on AI risk and puts growth, infrastructure, national security and competition with China at the center of federal strategy.
The document is less a detailed operating manual than a policy blueprint. Many of its effects will depend on how agencies interpret and carry out its recommendations. Still, its direction is explicit: the administration wants the United States to build faster, regulate less and treat AI as a strategic contest.
A plan built around speed
The AI Action Plan presents AI development as a national priority tied to economic growth, defense and global competition. It emphasizes data centers, semiconductor fabs, power sources and the energy grid needed to support them.
The administration argues that expanding AI infrastructure may require federal lands and faster permitting. The plan also raises the issue of keeping large power consumers running during critical energy grid periods, while asking the federal government to find ways for those consumers to manage power use when the grid is under stress.
President Trump framed the strategy as a rejection of limits he sees as barriers to American innovation. In a statement, he said, "To secure our future, we must harness the full power of American innovation." He added, "To do that, we will continue to reject radical climate dogma and bureaucratic red tape, as the Administration has done since Inauguration Day. Simply put, we need to 'Build, Baby, Build!'"
The plan also seeks to make the public case for spending billions of taxpayer dollars on data centers. It includes ideas for upskilling workers and partnering with local governments to create jobs connected to data center work.
Deregulation is a central lever
A major theme of the AI Action Plan is removing rules the administration views as slowing AI innovation and adoption. The plan directs the Office of Science and Technology Policy to ask businesses and the public which federal regulations stand in the way, so agencies can respond.
It also revisits the fight over whether states should be able to regulate AI. At the start of this month, the Senate removed a controversial budget bill provision that would have barred states from regulating AI for 10 years. That provision would have tied states' federal broadband funding to compliance with the moratorium.
The new plan explores another route. As part of a broader push to "unleash prosperity through deregulation," it threatens to limit states' federal funding based on their AI regulations. It also directs the Federal Communications Commission to evaluate whether state AI rules interfere with the agency's obligations and authorities.
That matters because many AI regulations can touch radio, TV and internet issues. Under the plan's framing, that could give the FCC a role in disputes over state AI rules.
Data centers, energy and environmental rules
The infrastructure section of the plan targets permitting and environmental review. The administration argues that rules such as NEPA, the Clean Air Act and the Clean Water Act are slowing America's ability to meet the demands of the AI arms race.
To accelerate construction, the plan calls for creating categorical exclusions, streamlining permitting processes and expanding fast-track programs like FAST-41. The goal is to make it easier to build critical AI infrastructure, especially on federal lands.
The source article notes that federal lands include national parks, federally protected wilderness areas and military bases. That makes the infrastructure push one of the most consequential parts of the plan, because it links AI growth to land use, energy supply and environmental oversight.
The plan also connects infrastructure to supply chain security. It focuses on keeping "adversarial technology" out of the U.S. supply chain, including Chinese-made chips and hardware.
The fight over biased AI
Another prominent section focuses on free speech and "American values." The plan calls for removing references to misinformation, DEI and climate change from federal risk-assessment frameworks.
It says, "It is essential that these systems be built from the ground up with freedom of speech and expression in mind, and that U.S. government policy does not interfere with that objective." The plan also says, "We must ensure that free speech flourishes in the era of AI and that AI procured by the Federal government objectively reflects truth rather than social engineering agendas."
One policy recommendation would update federal procurement guidelines so the government contracts only with frontier large language model developers that "ensure their systems are objective and free from top-down ideological bias." The unresolved issue is how the government would define and measure neutrality.
Rumman Chowdhury, a data scientist, CEO of the tech nonprofit Humane Intelligence, and former U.S. science envoy for AI, told TechCrunch, "The only way to be neutral would be literal non-engagement." Eugene Volokh, an American legal scholar who specializes in First Amendment and Second Amendment issues, said an order refusing to do business with any company that produces a non-neutral AI model would likely violate the First Amendment. He said a narrower approach focused on buying models that are sufficiently neutral would be more constitutionally defensible, though hard to implement.
The stakes are not theoretical. Anthropic, xAI, Google and OpenAI have all secured government contracts worth up to $200 million each to help integrate AI applications into the Department of Defense.
Open models, safety and China
The plan encourages development and adoption of open AI models that are free to download online and created with American values in mind. This appears to respond in part to open AI models from Chinese AI labs, including DeepSeek and Alibaba's Qwen.
To support open AI work, the plan calls for startups and researchers to have access to large computing clusters. It also says Trump wants to partner with leading AI model developers to increase research access to private AI models and data. American organizations that have taken an open approach, including Meta, AI2 and Hugging Face, could benefit from this emphasis.
The plan does include provisions aimed at AI safety and security. It calls for a federal technological development program to research AI interpretability, AI control systems and adversarial robustness. It also instructs agencies including the Department of Defense and Department of Energy to host hackathons to test AI systems for security vulnerabilities.
At the same time, the plan gives less emphasis than Biden's AI executive order to requiring leading AI model developers to report safety and security standards. The document acknowledges risks tied to cyberattacks and chemical and biological weapons, and asks frontier AI model developers to work with federal agencies to evaluate those risks.
China remains a defining concern. Federal agencies are directed to collect intelligence on foreign frontier AI projects that could threaten American national security. The Department of Commerce is tasked with evaluating Chinese AI models for alignment with Chinese Communist Party talking points and censorship.
The phrase "National security" appears 23 times in the plan. Its strategy includes integrating AI into the U.S. defense and intelligence apparatus, building AI data centers for the DoD, assessing how U.S. AI adoption compares with rivals like China, and securing preferential access to compute resources during national emergencies.