Govind Tekale
Karmactive Team
In a move that signals significant shifts in federal AI policy, President Donald Trump signed an executive order titled “Preventing Woke AI in the Federal Government” that requires federal agencies to procure only AI systems deemed “truthful” and “ideologically neutral.” The order specifically targets what it calls “diversity, equity, and inclusion” (DEI) influences in AI development, while also drawing rare criticism from Rep. Marjorie Taylor Greene over environmental concerns.
The Neutrality Mandate
The executive order establishes two core principles for large language models (LLMs) used by federal agencies:
“Truth-seeking” requires LLMs to “be truthful in responding to user prompts seeking factual information or analysis” and “prioritize historical accuracy, scientific inquiry, and objectivity.”
“Ideological Neutrality” mandates that LLMs function as “neutral, nonpartisan tools that do not manipulate responses in favor of ideological dogmas such as DEI.”
The order claims that DEI in AI development has led to “suppression or distortion of factual information about race or sex” and “manipulation of racial or sexual representation in model outputs.” It cites examples where AI systems allegedly changed the race of historical figures or refused certain requests based on race.
Federal agencies must include these requirements in new contracts for LLMs and, “to the extent practicable,” revise existing contracts within 90 days after the Office of Management and Budget (OMB) issues implementation guidance.
Building on Previous AI Frameworks
This order builds upon Executive Order 13960 from December 2020, which established initial guidelines for “Promoting the Use of Trustworthy Artificial Intelligence in the Federal Government.” The new directive narrows focus specifically to procurement requirements for large language models.
The OMB has 120 days to issue detailed guidance on implementation, including technical limitations, vendor transparency requirements, and exceptions for national security systems.
Technical Challenges
Achieving truly “neutral” AI presents significant technical hurdles. AI systems learn from vast datasets that inherently contain societal biases. Removing these biases without introducing others remains one of the most complex challenges in machine learning.
The order acknowledges these difficulties by instructing OMB to “account for technical limitations” and “avoid over-prescription” while affording “latitude for vendors to comply with the Unbiased AI Principles and take different approaches to innovation.”
Similar Posts
Greene’s Environmental Concerns
In an unexpected development, Rep. Marjorie Taylor Greene (R-GA) voiced opposition to aspects of the administration’s broader AI Action Plan, specifically regarding water usage by AI data centers.
“Data centers are like black holes when it comes to water requirements. They consume massive amounts of water for cooling, literally millions,” Greene posted on X (formerly Twitter). She expressed concerns about federal overreach into state water rights and potential threats to federalism through funding penalties.
This marks a rare public break between Greene and the Trump administration, alongside her previous dissent on Epstein files and Middle East policy.
Water Usage Reality
Greene’s concerns about water consumption have some basis in fact. Data centers, particularly those powering AI operations, require significant water resources for cooling systems. While specific figures vary by facility size and location, some large data centers can use millions of gallons daily, particularly in warmer regions where water-based cooling is necessary to prevent equipment overheating.
This water usage creates potential strain in drought-prone areas, raising questions about the environmental impact of expanded AI infrastructure—a concern that crosses political lines but has received limited attention in federal AI policy discussions.
Implementation Timeline
The executive order sets a 120-day deadline for OMB to issue implementation guidance, which would be approximately November 2025 based on the July 2025 issuance. After that, agencies have 90 days to adopt procedures ensuring compliance with the “Unbiased AI Principles.”
The guidance will need to address several complex issues, including:
- How to measure and verify “truth-seeking” and “neutrality” in AI systems
- Transparency requirements for vendors without requiring disclosure of sensitive technical data
- Exceptions for national security applications
- Whether these principles should apply to agency-developed LLMs and other types of AI models
Potential Legal Tensions
The order’s focus on eliminating DEI considerations from AI systems may create tensions with existing civil rights laws and procurement regulations that require consideration of equity and non-discrimination principles.
Federal agencies will need to navigate these potentially competing directives as they implement the new requirements in their procurement processes.
What’s Next
The full impact of this executive order will become clearer once OMB releases its implementation guidance, expected in November 2025 (approximately 120 days after the July issuance). Industry responses, potential legal challenges, and the practical feasibility of enforcing “ideological neutrality” in complex AI systems will determine the order’s lasting influence on federal AI deployment.
Meanwhile, Congress may exercise oversight of implementation through committee hearings, while civil rights groups are expected to scrutinize how the neutrality requirements align with non-discrimination mandates.
