Trump's "Anti-Awakening AI" Order Could Reshape US Tech Firms' Model Training Approaches

When Chinese firms such as DeepSeek and Alibaba unveiled their latest AI systems, Western researchers quickly observed these models sidestepped questions critical of the Communist Party. U.S. officials later confirmed these tools were intentionally designed to reflect Beijing's stance, raising concerns about censorship and algorithmic bias.

American AI pioneers like OpenAI argue this validates the need for rapid technological advancement with minimal regulatory oversight. As OpenAI's Chief Global Affairs Officer Chris Lehane posted on LinkedIn last month, it represents a "contest between democratic AI led by America and authoritarian AI controlled by the Chinese Communist Party."

President Donald Trump's executive order issued Wednesday banning "woke AI" and non-"ideologically neutral" models from federal contracts could disrupt this balance. The directive labels diversity, equity, and inclusion (DEI) initiatives as "pervasive and destructive" ideologies that "distort output quality and accuracy," specifically targeting racial/gender data manipulation, critical race theory, transgenderism, implicit bias, intersectionality, and systemic racism.

Experts warn this may create a chilling effect on developers who might feel pressured to align model outputs with White House rhetoric to secure federal funding for their businesses. The same day as the order, the White House released Trump's "AI Action Plan" shifting national priorities from societal risks to building AI infrastructure, reducing regulatory burdens for tech firms, enhancing national security, and competing with China.

The order instructs the Office of Management and Budget Director, Federal Procurement Policy Administrator, General Services Administration Administrator, and Office of Science and Technology Policy Director to issue guidance to agencies on compliance. "We're going woke-free," Trump declared at an AI event hosted by All-In Podcast and Hill & Valley forum. "I will sign an order banning federal purchases of AI technology infused with partisan bias or ideological agendas like critical race theory, which is nonsense. From now on, the U.S. government will only deal with AI that pursues truth, fairness, and strict impartiality."

Defining what constitutes objectivity presents one of the command's major challenges. Philip Seargeant, senior lecturer in applied linguistics at the Open University, told TechCrunch that true neutrality is impossible to achieve. "A core principle of sociolinguistics is that language is never neutral," Seargeant explained. "Believing you can attain pure objectivity is an illusion."

Moreover, Trump's administration ideology doesn't reflect all American values. The president has repeatedly attempted to eliminate funding for climate projects, education, public broadcasting, research, social service subsidies, community programs, agricultural support, and gender-affirming care, often labeling these as "woke" government expenditures.

As former U.S. AI Science Envoy Ruma Chowdhury, CEO of tech non-profit Humane Intelligence, noted: "Anything the Trump administration dislikes gets immediately tossed into this pejorative woke pile." The recently issued order's definitions of "seeking truth" and "ideological neutrality" contain both vague and specific elements. While "seeking truth" is defined as LLMs prioritizing historical accuracy, scientific inquiry, and objectivity, "ideological neutrality" is described as "neutral, nonpartisan tools that don't manipulate responses to support ideological doctrines like DEI."

These definitions allow significant interpretive leeway while creating pressure. AI companies have consistently pushed for fewer operational restrictions. Though the executive order lacks legislative force, cutting-edge AI firms may still find themselves influenced by shifting political agendas. Last week, OpenAI, Anthropic, Google, and xAI each secured up to $200 million in DoD contracts to develop AI workflows addressing national security challenges.

It remains unclear which companies would benefit most from the woke AI ban or whether they'll comply. xAI might be the most aligned with the order initially - at least early on - despite its own biases. Elon Musk markets xAI's Grok chatbot as the ultimate anti-woke, "less biased" truth seeker. Grok's system prompts direct it to avoid mainstream authorities and media, seek politically inconvenient dissenting information, and even reference Musk's personal views on controversial topics. Recently, Grok published antisemitic content and praised Hitler on X, alongside other hate speech and misogynistic posts.

Stanford Law Professor Mark Lemley told TechCrunch the order "clearly constitutes viewpoint discrimination since [the government] just signed a contract with Grok, alias 'Mechanical Hitler.'" Beyond xAI's DoD funding, the company announced "Government Grok" was added to GSA's plans, meaning xAI products can now be purchased by every government office.

"The right question is: Will they ban Grok, the AI they just signed a big contract for because it's deliberately designed to provide politically controversial answers?" Lemley said in an email interview. "If not, that clearly discriminates against specific viewpoints."