Trump Administration Unveils Sweeping AI Action Plan Centered on Deregulation, Data Center Buildout, and Global Tech Reach

The Trump administration on Wednesday released a 28-page “AI Action Plan” outlining more than 90 policy steps it wants federal agencies to execute in the near term to secure what it calls U.S. leadership in a global artificial intelligence race. The framework leans hard into accelerating domestic innovation, rapidly scaling data center and energy infrastructure, and promoting widespread adoption of American AI systems abroad.

David Sacks, the White House’s AI and crypto lead, framed the moment as a high-stakes competition with national security implications, saying the administration’s goal is to ensure the United States “wins” in AI. The plan’s opening thrust is deregulatory: federal agencies are to inventory and roll back rules judged to “injure” AI development across sectors including finance, agriculture, health care, and transportation.

The administration is also signaling pressure on states—potentially tying federal funds to curbing state-level AI restrictions—and asking the Federal Communications Commission to review whether state AI rules interfere with federal mandates.

In a similar vein, the framework calls for reexamining Federal Trade Commission investigations that originated under the prior administration, warning against liability theories that could chill AI innovation. Michael Kratsios, who heads the White House Office of Science and Technology Policy, cast the approach as a deliberate break from what he characterized as Europe’s overly restrictive regulatory model.

Policy rewrites extend to technical guidance. The Commerce Department is directed to revise its AI risk framework, stripping references to misinformation, climate change, and diversity, equity and inclusion (DEI). Updated federal procurement guidelines would steer government contracts toward systems deemed “objective” and insulated from “top-down ideological bias.”

A large portion of the plan addresses infrastructure. To meet anticipated AI computing demand, the administration proposes sweeping exemptions or expedited permits for data center projects under federal environmental laws.

It suggests fast-tracking “all” data center and related energy facilities within the national permitting system, exploring a nationwide permit that could allow projects to proceed despite potential water-quality impacts, and reducing environmental review for actions that “normally” lack significant effects—though the plan leaves that category loosely defined. The government would also open federally owned land for data centers and associated power generation, pursue grid-stabilization efforts, and continue pushing domestic semiconductor manufacturing.

International strategy rounds out the plan’s three main pillars. The administration wants to drive the adoption of U.S. AI technology overseas, counter Chinese influence in emerging standards bodies, and tighten export control levers in ways that reinforce American leadership. “Having the best models isn’t enough,” Kratsios said in briefing materials; true leadership comes when those systems are deployed at scale in U.S. and global markets.

The framework tracks closely with earlier administration moves. Vice President JD Vance has criticized “excessive regulation” in global forums, President Trump rescinded his predecessor’s AI executive order aimed at guardrails, and the White House has spotlighted massive infrastructure partnerships—including the $500 billion “Stargate” initiative announced in January with OpenAI, SoftBank, and Oracle. Last week, the president touted an additional $92 billion in private commitments for data centers and energy projects in Pennsylvania.

Further implementation details—and the regulatory rollbacks agencies may pursue—are expected to emerge as departments respond to the directives in the weeks ahead.