The Trump administration released a 28-page "AI Action Plan" on Wednesday, outlining three pillars for US AI policy: accelerating AI innovation, building American AI infrastructure, and leading international AI diplomacy and security. The plan, released ahead of President Trump's appearance at a Washington, DC event, is expected to be supported by related executive orders. Key proposals include removing "references to misinformation, Diversity, Equity, and Inclusion, and climate change" from federal risk management guidance and prohibiting federal contracts with large language model (LLM) developers unless they ensure "objective" systems.

The plan advocates for a "try-first" culture for AI adoption across American industries, particularly within the Armed Forces, and seeks to reduce regulatory hurdles at the state and federal levels, potentially denying AI-related funding to states with hindering regulations. It also emphasizes expanding the power grid to support energy-intensive AI industries and addressing gaps in semiconductor manufacturing export controls.

The administration aims to foster AI research, including in biosecurity, and create legal frameworks to combat deepfakes. The plan's introduction, penned by White House AI and crypto czar David Sacks, President's science and technology advisor Michael Kratsios, and Secretary of State Marco Rubio, asserts that implementing the plan and "Winning the AI race" is crucial for national security, competitiveness, and economic well-being.

Impact Statement: The plan's focus on deregulation and a "try-first" approach may significantly impact the US AI industry, while its stance on diversity, equity, and inclusion initiatives, and climate considerations, is likely to be highly controversial.