President Donald Trump has a new plan for artificial intelligence, and it starts with a war against woke ideology. On Wednesday, the Trump administration released its AI manifesto, a 28-page action plan focused on beating China in the AI race. The same day, Trump gave a speech in Washington promoting the plan and signed an executive order entitled “Preventing Woke AI in the Federal Government.”
The order is not subtle. It tells the Department of Commerce to strip references to misinformation, Diversity, Equity, and Inclusion, and climate change from Biden-era AI rules. The White House fact sheet accompanying the plan says LLMs “shall be truthful and prioritize historical accuracy, scientific inquiry, and objectivity.” Trump, in his speech, said: “The American people do not want woke Marxist lunacy in the AI models.”
The Order’s Targets
The executive order splits its attention between two goals. First, it directs the Federal Government to be “hesitant to regulate the functionality of AI models in the private marketplace.” Second, it says the government has an obligation “not to procure models that sacrifice truthfulness and accuracy to ideological agendas” in federal procurement.
The order spends several paragraphs demonizing AI that supports diversity, calls out racial bias, or values gender equality. It treats those features as ideological poison rather than technical design choices. The administration’s position is that AI should reflect mainstream American values, not dissenting ones.
Who Agrees With Trump
No Big Tech company has publicly objected to the Trump administration’s plan. That silence is notable. Instead, the companies have found ways to signal agreement.
Google celebrated the White House’s support of its pet issues, like boosting infrastructure. Anthropic published a positive blog post about the plan but complained about the White House’s sudden seeming abandonment of strong export controls earlier this month. OpenAI says it is already close to achieving objectivity.
An OpenAI engineer told the author the company already strives for neutrality and that meeting government standards like being anti-woke “shouldn’t be a huge hurdle” in a technical sense.
The Politics of Truth
The order’s stated goal is to protect the truth. But it defines truth in narrow terms. The White House fact sheet says LLMs “shall be truthful and prioritize historical accuracy, scientific inquiry, and objectivity.” Trump’s speech made the stakes personal. He framed the fight as a moral crusade against dissent.
The administration’s position is that AI should reflect mainstream American values, not dissenting ones. The order directs the Federal Government to be “hesitant to regulate the functionality of AI models in the private marketplace,” even as it says the government has an obligation “not to procure models that sacrifice truthfulness and accuracy to ideological agendas” in federal procurement.
What This Means for Companies
The order creates a strange alignment. The administration treats Big Tech as partners in the AI race against China, even as it demands they filter their products to match its political views. Companies that build AI systems will now face pressure to certify that their models meet government standards for objectivity.
The administration’s order is a warning to anyone building AI systems that dissenting ideas are not welcome.
Where the paper stands
The paper backs narrow rules against direct AI harm, such as forced disclosure of hidden safety failures, and is against broad rules that hand the market to the incumbents, including federal licensing and slowdowns that freeze today’s leaders in place. The Trump administration’s new AI order does neither of these things. It is a broad rulebook written to favor one set of beliefs over others, aimed at the federal marketplace while treating Big Tech as partners in the AI race against China. That combination locks out smaller companies that cannot certify their models to the government’s standards, and it hands the administration’s favored firms a moat they did not earn.
The paper’s position on regulation is plain: rules that raise the cost of entry for small businesses and protect the biggest players are the wrong kind of rules. The administration’s order does exactly that, dressing it up as a war against woke ideology.
The paper would prefer narrow rules aimed at actual harm, narrowly defined, applied only when a company’s product directly endangers people or the environment. It would also want any oversight regime kept free of the interests of the biggest firms, which often help write the rulebooks that lock out their competitors. The administration’s order points in the opposite direction. Its targets are not technical failures but dissenting ideas, and its partners are the very companies whose products it is regulating.
Key Facts Box
- Plan length: 28 pages
- Order title: “Preventing Woke AI in the Federal Government”
- Speech location: Washington
- Quote: “The American people do not want woke Marxist lunacy in the AI models.”
Timeline
| Date | Event |
|---|---|
| Wednesday | Trump signs executive order and delivers speech |
| Same day | Administration releases 28-page AI manifesto |
The administration’s plan is ambitious. It wants to reshape the global AI landscape by defining what counts as acceptable content. It wants to do that while treating the companies building that content as partners.
That is a hard balance to strike. The administration’s order is a warning to anyone building AI systems that dissenting ideas are not welcome.
Source material: “Trump’s Anti-Bias AI Order Is Just More Bias,” WIRED.
Get the Notebook.
The day's best stories and every fresh verdict, in plain English, in your inbox by seven. One email a day, no more.