Trump Issues 3 Executive Orders on Artificial Intelligence to Accelerate AI Exports, Data Centers, and “Unbiased” Federal Models

Publication Date: July 23, 2025

Overview

Key Takeaway: President Donald J. Trump signed three executive orders on July 23, 2025 to (1) promote export of a full-stack U.S. AI “technology package,” (2) accelerate federal permitting for large AI data centers and related infrastructure, and (3) bar federal agencies from procuring large language models (LLMs) that embed diversity, equity, and inclusion (DEI) concepts, labeling them “ideological.”
Together, the orders seek to consolidate U.S. technological leadership abroad, speed domestic build‑outs, and set procurement rules for “truth-seeking” and “ideologically neutral” AI in government.

Facts

The White House issued three executive orders dated July 23, 2025: “Promoting the Export of the American AI Technology Stack,” “Accelerating Federal Permitting of Data Center Infrastructure,” and “Preventing Woke AI in the Federal Government.”
The export order creates the American AI Exports Program and directs the Secretary of Commerce—within 90 days—to solicit industry-led, full-stack AI export proposals covering hardware, data pipelines, models, cybersecurity, and sector-specific applications. Proposals must identify target countries and request federal support while complying with U.S. export controls.
Selected “priority AI export packages” are to receive coordinated federal financing tools (e.g., loans, guarantees, insurance) through the Economic Diplomacy Action Group (EDAG) and agencies such as the Small Business Administration, State, and Commerce.
The permitting order defines “Data Center Projects” as facilities needing >100 MW of new load for AI uses and “Qualifying Projects” as those meeting thresholds like ≥$500M capital expenditure or national security relevance.
It revokes Executive Order 14141 and directs agencies to expand categorical exclusions and leverage processes to speed National Environmental Policy Act (NEPA) reviews. EPA must identify Superfund/Brownfield sites for reuse and consider rule changes to expedite permits.
The “Preventing Woke AI” order instructs agencies to procure only LLMs that meet two “Unbiased AI Principles”: Truth-seeking and Ideological Neutrality. OMB must issue implementing guidance within 120 days. Contracts for LLMs signed thereafter must include compliance terms and vendor-paid decommissioning costs if models violate the principles.
Each order states it creates no enforceable rights and must be implemented consistent with law and available appropriations. Publication costs are assigned respectively to Commerce, Energy, and GSA.

Perspectives

White House (via President Donald J. Trump): Emphasizes AI as foundational for economic growth, national security, and global leadership, arguing these orders preserve U.S. dominance by promoting unbiased, innovative technologies free from adversarial influence.
Christopher F. Rufo (Conservative Activist): Views the anti-bias order as crucial to counter DEI ideologies distorting AI outputs, proudly contributing to its drafting to ensure systems prioritize truth and accuracy over social agendas.
David Sacks (White House AI Czar): Supports the neutrality mandate, stating it prevents federal procurement of “woke AI,” crediting Rufo’s input in defining bias and ensuring government tools remain nonpartisan.
Sam Altman (OpenAI CEO): Endorses the initiatives as incredible for the country, highlighting their potential to create jobs and advance AGI in America, reflecting a changed positive perspective on Trump’s leadership.
Elon Musk (xAI CEO and Former Advisor): Criticizes the export and infrastructure pushes as underfunded and flawed, questioning partners’ financial commitments while favoring U.S. dominance but prioritizing his own neutral AI models.
Henna Virkkunen (EU Tech Sovereignty Chief): Represents Europe’s divergent stance, advancing the AI Act for safety and ethics despite U.S. deregulation, stressing no single region can lead alone and warning of risks in weaponizing tech dependencies.

Considerations

Exporting U.S. AI stacks could strengthen alliances and economic ties short-term but heighten long-term proliferation risks if not tightly controlled.
Streamlining data center permitting may spur rapid infrastructure growth and jobs, yet could exacerbate environmental strains without balanced oversight.
Mandating unbiased federal AI promotes objectivity and trust, potentially shifting paradigms toward merit-based systems while challenging DEI integrations.
Overall, these actions signal a pro-innovation U.S. surge, likely accelerating tech leadership but diverging from global regulatory trends, influencing future international standards.
Revoking prior orders removes immediate barriers, fostering quick wins in competitiveness, though sustained impacts depend on implementation amid workforce shifts.