By Key Points
- Trump’s Anti-Woke AI Order Could Reshape $200M Tech Deals
- Trump bans “woke” and non-neutral AI from federal contracts
- AI must align with “truth-seeking” and “ideological neutrality”
- xAI’s “Grok for Government” may benefit despite bias claimsExperts warn
- of pressure on AI firms to align with political views
President Donald Trump signed a sweeping executive order on Wednesday that could drastically change how U.S. tech companies build and train their artificial intelligence models.
The order bans the use of so-called “woke AI” in federal contracts, specifically targeting systems that promote diversity, equity, and inclusion (DEI) or address topics like race, gender, and systemic bias.
This marks the most aggressive step yet in aligning U.S. government AI procurement with Trump’s broader political ideology.
The new policy demands that all AI models used by federal agencies must be “ideologically neutral” and “truth-seeking.” That includes avoiding any influence from DEI frameworks, critical race theory, transgender rights, or other concepts deemed “partisan.”
President Trump issued an executive order today meant to remove wokeness from AI and the order is fantastic. Here’s some key points:
To get federal contracts, AI companies must tell the truth, be historically accurate, be objective, be totally non-partisan and NEVER deploy DEI… pic.twitter.com/w3zYq7CntL
— Robby Starbuck (@robbystarbuck) July 24, 2025
The administration’s official “AI Action Plan,” released the same day, makes it clear: future federal AI spending will prioritize national security, economic growth, and competition with China — not ethical considerations or social equity.
According to the plan, this is a move to combat the rise of “autocratic” Chinese AI, which the U.S. sees as biased toward the Chinese Communist Party.
But the big question now is: who decides what’s “neutral”?
Experts say the vagueness of the definitions in Trump’s order could lead to massive pressure on tech companies. AI developers may feel compelled to align their model behavior and training data with political preferences to win lucrative federal deals — some of which are already in motion.
Last month, we saw OpenAI making strategic infrastructure moves in its multi-year cloud partnership with Oracle, a sign that AI companies are anticipating long-term shifts in policy and scaling needs.
BREAKING: President Donald Trump signed new executive
ENDING WOKE AI
AI must be ideologically neutral
No more wokeism.
No more Critical Race Theory.
No more DEI code baked into the algorithm.
Just facts, logic, and accuracy.AI should inform, not indoctrinate. pic.twitter.com/aLL0M9kPEF
— Marc Nixon (@MarcNixon24) July 23, 2025
Big Tech Faces a Tough Choice Under New AI Rules
Last week, major AI players — including OpenAI, Google, Anthropic, and Elon Musk’s xAI — each secured contracts worth up to $200 million from the Department of Defense to develop AI tools for national security.
However, this new executive order could upend those agreements. If a company’s models are perceived as “woke” or ideologically biased, they could lose access to federal funding and contracts.
That puts intense pressure on these firms to evaluate — or potentially modify — their systems to match the Trump administration’s narrative.
Interestingly, xAI’s Grok chatbot, which openly promotes itself as “anti-woke,” could be the biggest beneficiary. Grok has a history of offering politically charged answers, rejecting mainstream media sources, and even referencing Elon Musk’s personal views.
While this behavior might disqualify other models under neutrality rules, Grok seems to be getting a pass — even being added to the federal procurement system through the General Services Administration.
Stanford law professor Mark Lemley didn’t hold back, saying the executive order represents “viewpoint discrimination”. He questioned whether the same rules would be applied to Grok, despite its track record of spreading offensive and politically loaded content.
Meanwhile, Rumman Chowdhury, CEO of tech nonprofit Humane Intelligence, warned of a dangerous precedent. She believes AI firms might soon tailor their training data to suit the administration’s ideology.
That could mean editing out content related to racial or gender equity, rewriting historical narratives, or removing scientific consensus in favor of political convenience.
Musk recently said that xAI’s latest version of Grok will “rewrite the entire corpus of human knowledge”, deleting what he sees as errors and retraining the model on “correct” information.
We will use Grok 3.5 (maybe we should call it 4), which has advanced reasoning, to rewrite the entire corpus of human knowledge, adding missing information and deleting errors.
Then retrain on that.
Far too much garbage in any foundation model trained on uncorrected data.
— Elon Musk (@elonmusk) June 21, 2025
Experts are alarmed. If AI models begin reflecting one political agenda — rather than a balanced range of viewpoints — public trust in these technologies could erode.
Critics also argue that “neutrality” is nearly impossible. As Philip Seargeant, a linguistics expert from the Open University, puts it, “language is never neutral.” Whether it’s intentional or not, every AI reflects the data it’s trained on and the choices made by its creators.
Trump’s order appears to ignore that reality. Instead, it calls for strict impartiality — but only when that impartiality aligns with conservative values.
AI systems like ChatGPT Agent, which offer dynamic workflows and learning capability, may face added scrutiny under such a policy if their decision-making appears influenced by social or cultural context.
Hill & Valley x All-In are hosting President @realDonaldTrump next week in DC for a day-long AI policy event.
Hill & Valley Forum x All-In will host Winning the AI Race, a summit in Washington D.C. featuring a Keynote Address by President Donald J. Trump.
This will be… pic.twitter.com/HSqGvOebnk
— Jacob Helberg (@jacobhelberg) July 15, 2025
The Risk of Political Control Over AI Narratives
This executive order not only impacts how AI tools are built, but also raises deep questions about who controls the narrative in a digital age increasingly shaped by artificial intelligence.
By labeling entire areas of academic and social discussion as “woke,” the Trump administration is setting boundaries on what is considered valid information — and, by extension, what an AI model is allowed to learn, process, or produce.
Conservative voices like David Sacks, recently appointed as the U.S. AI czar, argue that this move defends free speech and protects against ideological capture. He and other allies claim that most AI tools today lean too far left, and this policy levels the playing field.
But critics warn that the outcome could be worse: the creation of AI systems that echo only one side of the political spectrum. Chowdhury and other experts say this creates a dangerous precedent — where truth is defined by who’s in power, and objectivity becomes a moving target.
If AI becomes a tool of political enforcement rather than open inquiry, then its role as a neutral information processor is lost. Training models to reflect only the approved worldview of one administration could influence everything from education to journalism to law enforcement.
And the effects could last far beyond a single presidency. AI models built and deployed now could influence thinking and decision-making for decades to come.
Recent breakthroughs like Google and OpenAI’s win at the 2025 AI Math Olympiad show the immense capability of these tools, but under the new order, their creative and reasoning power may be curtailed to meet political expectations.
The fear is that even promising innovations — from voice-first interfaces like Le Chat’s voice recognition system to fully AI-powered browsers like Comet AI Browser — could be forced to conform or risk losing access to federal partnerships and contracts.
Whether this order succeeds or not, one thing is clear: the battle over AI’s future is no longer just technical, it’s deeply political.
What's your reaction?
