AI Meets DEI: Navigating the "Woke AI" Executive Order

Published: February 24, 2026

In July 2025, the Trump administration issued an executive order targeting “woke” artificial intelligence (AI). The order specifically takes aim at DEI in the AI context, which it describes as leading to “the suppression or distortion of factual information about race or sex; manipulation of racial or sexual representation in model outputs; incorporation of concepts like critical race theory, transgenderism, unconscious bias, intersectionality, and systemic racism; and discrimination on the basis of race or sex.” The order suggests that DEI “poses an existential threat to reliable AI.” Below is an explainer of the executive order, including its implications for private sector organizations.

Understanding Executive Order 14319, Preventing Woke AI in the Federal Government

The order establishes standards for federal procurement of large language models (LLMs), which are AI systems trained to understand and generate natural language in response to user prompts, such as ChatGPT and Microsoft Copilot. These federal procurement standards instruct agencies to “withhold contracts from companies that don’t align their technology with the [order’s] definition[s] of what it calls “unbiased AI principles,” which include “truth-seeking” and “ideological neutrality.”

• The truth-seeking principle is that LLMs should “be truthful,” prioritize “historical accuracy, scientific inquiry, and objectivity”, and acknowledge when the information provided is “incomplete or contradictory.” 

• The ideological neutrality principle is that LLMs should be “neutral, nonpartisan tools that do not manipulate responses in favor of ideological dogmas such as DEI.” 

The order instructs the Director of the Office of Management and Budget (“OMB”) to issue guidance to agencies on how to implement the above two “Unbiased AI Principles” within 120 days of the date of the order (i.e., by November 20, 2025). On December 11, 2025, the Director, Russell Vought, released a memo in accordance with this mandate. Key takeaways to note: 

• Vendor certification: The memo requires agencies to make “Unbiased AI Principles” essential contract terms, enabling contract termination for non-compliance;

• Transparency: The memo requires vendors to provide detailed documentation about how their AI was built and trained, including what data was used, how bias was tested, and what technical controls limit AI outputs;

• Monitoring: The memo requires ongoing monitoring, treating it as a lifecycle obligation rather than a one-time review, to address model updates and ideological drift. 

Implications for Private Sector Organizations

The order applies to federal agencies and does not impose direct restrictions on LLMs marketed to the private sector. However, federal procurement standards have historically influenced the broader market and could affect the quality of commercially available LLMs.

In our view, the principles of truth-seeking and ideological neutrality are not, on their face, inconsistent with the values of DEI. To the contrary, properly crafted DEI efforts reduce bias and discrimination, and promote fair and accurate information on topics such as race and gender. In particular, as Meltzer Center leaders have argued before, there is an important distinction between “lifting” DEI (which confers preferences on marginalized groups) and “leveling” DEI (which adopts identity-neutral methods to remove bias and create a level playing field for all). At the very least, leveling DEI efforts promote exactly the kind of “unbiased AI” that this executive order purports to encourage.