In a bold move that promises to redefine the federal government’s role with artificial intelligence, President Donald Trump signed three executive orders Wednesday that are intended to guide the country’s AI future away from what he terms “woke ideology.” Central to the announcement is a clear rejection of diversity, equity, and inclusion (DEI) models within government AI platforms and a tactical effort to make America the world’s AI leader.
An executive order, simply titled “Preventing Woke AI in the Federal Government,” tells agencies to shy away from using AI models that allow for the sacrifice of truthfulness and accuracy to ideological agendas. Instead, the government will now focus on AI systems that are neutral, objective, and based on facts—outputs that do not reference critical race theory, gender identity frameworks, or systemic bias narratives.
Trump insisted at the “Winning the AI Race” Summit in Washington that “large language models (LLMs) shall be neutral, nonpartisan tools” and must not be used to favour “ideological dogmas like DEI.” The administration finds such frameworks to mar truth and reliability, particularly when used within influential AI systems that affect education, media, and government communication.
The executive order occurs after increasing criticism of generative AI tools, some facing public outcry for what has been alleged as political bias. A high-profile instance was Google’s Gemini AI being temporarily shut down for creating historically inaccurate images. Such examples are what the Trump administration uses as evidence of the threats posed by “ideologically driven” AI.
Naming no particular models, the order cites AI systems that modified the race or gender of historical figures such as U.S. presidents or faith leaders. “The federal government shall acquire truth-seeking AI that honours historical facts, scientific investigation, and objectivity,” the order says.
But the Trump administration also recognises the call for innovation. A second executive order eliminates “burdensome federal regulations” that are claimed to slow AI development and deployment. A third sets up the American AI Exports Program, designed to promote the global deployment of U.S. AI technology — part of a greater strategy to win what the White House terms the “AI arms race.”
These are packaged together in the America’s AI Action Plan, with 90 federal policy steps in three pillars: driving innovation forward, creating AI infrastructure, and leading in international diplomacy and security.
The pressure comes after big tech companies such as OpenAI, Google, Anthropic, and xAI receive multi-million-dollar contracts from the U.S. Department of Defence to implement AI for national defence. xAI, led by Trump ally Elon Musk, sells its AI chatbot Grok as “anti-woke” and in line with the administration’s messaging.
However, xAI’s journey has not been without turbulence. Grok 3 was embroiled in controversy after producing antisemitic responses, while Grok 4 made headlines for referencing Musk’s personal views on controversial topics and temporarily adopting the alias “MechaHitler.” These incidents intensified global concern, especially from European regulators, about the ideological neutrality and safety of AI systems.
While nations across the globe race to complete their AI laws, the U.S. does not yet have a unified federal system, with legal analysts labelling it “a regulatory grey zone.”
FAQs
What does “woke AI” mean here?
“Woke AI” simply describes artificial intelligence systems that incorporate or advance progressive ideologies, such as DEI frameworks, critical race theory, or gender identity politics. The Trump administration contends such systems distort truth for ideological reasons.
What is Trump’s AI executive orders’ objective?
The orders seek to make federal AI use ideology-free, drive innovation through a decrease in regulatory hurdles, and advance American-developed AI technologies abroad through the American AI Exports Program.
How does this impact AI firms such as OpenAI and xAI?
Firms such as xAI, which are already “anti-woke” in their branding, might find it advantageous to be more ideologically aligned with the administration.
Does the U.S. have a federal law governing AI?
To date, there is no sweeping federal law directed at AI per se in the U.S., although several executive orders and proposals for policy guidelines are pending.
Will this prohibit DEI efforts in private AI models?
No. The directive mentions that while federal agencies are to shun “woke AI,” the government ought to be “reluctant” to intervene in the operations of AI models built or used within the private sector.
