Trump’s AI Strategy Focuses on Growth Over Regulation in the Competition with China

The Trump administration unveiled its eagerly awaited AI Action Plan on Wednesday, representing a notable shift from former President Biden’s cautious perspective on AI risks. This new initiative keenly emphasizes upgrading AI infrastructure, streamlining regulations for tech companies, bolstering national security, and outpacing China.

This transformational approach is poised to affect multiple industries and may resonate with everyday American consumers. Significantly, the AI Action Plan places less focus on addressing potential AI harms, prioritizing instead the development of data centers to nurture the AI industry, even if this entails using federal lands or managing energy supplies during peak grid periods.

The success of the plan largely depends on how well it is executed, with many details yet to be finalized. More than just an exhaustive blueprint, the AI Action Plan is envisioned as a framework for action, with a clear intent: progress is crucial.

The Trump administration claims this is the sole approach to “usher in a new golden age of human flourishing.” It aims to convince the American public that funneling billions of taxpayer dollars into data centers is in their best interest. Furthermore, the plan includes initiatives for enhancing worker skills and partnering with local governments to create job opportunities in data center roles.

“To secure our future, we must harness the full power of American innovation,” Trump declared. “We will continue to dismantle radical climate dogma and bureaucratic red tape, as the administration has done since Inauguration Day. In short, we need to ‘Build, Baby, Build!’”

Crafted by the Trump administration’s team of tech and AI experts, many of whom come from Silicon Valley, the AI Action Plan includes Office of Science and Technology Policy director Michael Kratsios; AI and crypto specialist David Sacks; and Marco Rubio, assistant to the president for national security affairs. The plan was shaped by input from over 10,000 interest groups.

Deregulation and reinstating the AI moratorium

Earlier this month, the Senate removed a divisive provision from the budget bill that would have barred states from regulating AI for a decade. If passed, this provision would have linked federal broadband funding for states to compliance with the moratorium.

The situation remains unsettled as the AI Action Plan seeks fresh avenues to limit states’ powers in AI regulation. As part of a wider agenda aimed at “unleashing prosperity through deregulation,” the administration implies that it could reduce federal funding to states based on their AI regulatory frameworks.

The plan also instructs the Federal Communications Commission to “evaluate whether state AI regulations impede the agency’s duties and responsibilities.” Essentially, this means that if state AI regulations overlap with radio, TV, and internet regulations — which many do — the FCC could intervene.

At the federal level, the action plan calls for the Office of Science and Technology Policy to gather opinions from businesses and the public on current federal regulations that may restrict AI innovation, allowing for appropriate adjustments to be made.

Streamlining regulations for data centers

Trump’s deregulation agenda extends to plans for expediting the development of AI-related infrastructure such as data centers, semiconductor manufacturing plants, and energy sources. The administration contends that current environmental laws — including NEPA, the Clean Air Act, and the Clean Water Act — hinder America’s ability to keep pace with the urgent demands of the AI arms race.

As a result, the AI Action Plan stresses the stabilization of America’s energy grid. At the same time, it encourages the federal government to explore new strategies that enable major energy consumers — especially AI firms — to manage their energy consumption during peak grid times.

Some companies, including xAI and Meta, have faced criticism for contributing to pollution in vulnerable communities. For instance, xAI has been accused of bypassing environmental safeguards and exposing local neighborhoods to harmful emissions from gas-powered turbines at its Memphis data center.

The action plan promotes establishing categorical exclusions, fast-tracking permitting steps, and enhancing expedited programs like FAST-41 to simplify the construction of crucial AI infrastructure, particularly on federal lands, which include national parks, protected wilderness areas, and military bases.

Returning to Trump’s consistent focus on countering China, this strategy aims to exclude foreign technology and reinforce security measures to keep “adversarial technology” — such as Chinese-made chips and hardware — out of the U.S. supply chain.

Trump’s campaign against “biased AI”

A significant aspect of Trump’s AI Action Plan is its focus on protecting free speech and “American values”, partially by eliminating references to misinformation, diversity, equity, and inclusion (DEI), as well as climate change from federal risk assessment guidelines.

“It is essential that these systems are designed around the principles of free speech and expression, and that U.S. government policy does not hinder that objective,” the plan asserts. “We must ensure that free speech thrives in the age of AI and that AI acquired by the Federal government accurately reflects truth rather than sociopolitical agendas.”

Despite intentions to safeguard free speech, the AI Action Plan may inadvertently achieve the opposite result.

One of its proposed policy actions involves revising federal procurement standards to ensure that the government engages solely with large language model developers that guarantee their systems are impartial and free from ideological bias imposed top-down.

This language is consistent with what The Wall Street Journal reported would feature in Trump’s anticipated executive order, expected to be issued today.

The challenge remains in defining objectivity, as the government has not clarified how it plans to evaluate models for their neutrality.

“The only way to ensure neutrality would be complete disengagement,” stated Rumman Chowdhury, a data scientist and CEO of the tech non-profit Humane Intelligence, and a former U.S. science envoy for AI, in remarks to TechCrunch.

Companies like Anthropic, xAI, Google, and OpenAI have all secured government contracts valued at up to $200 million each to assist with integrating AI applications within the Department of Defense. The potential impacts of Trump’s proposed policies and upcoming executive order could be profound.

“For example, an order stating, ‘We won’t engage with any company producing a non-neutral AI model’ would likely violate the First Amendment,” observed Eugene Volokh, a legal scholar focused on First Amendment and Second Amendment issues, via email. “On the other hand, an order that says, ‘We will only engage in contracts for models that are sufficiently neutral’ would be more constitutionally valid, although implementing it could prove quite challenging (mainly due to the complexities in defining ‘neutral’).”

Furthermore, he noted, “If the order instructs agencies to select AI systems based on accuracy and neutrality while allowing some interpretative flexibility, that may represent a more practical approach.”

Promoting an open AI approach

Trump’s AI Action Plan seeks to enhance the proliferation and use of open AI models, which are freely available online and created with American values in consideration. This initiative seems largely a response to the rise of competitive open AI models from Chinese organizations like DeepSeek and Alibaba’s Qwen.

To facilitate this strategy, Trump plans to ensure that startups and researchers working on open models have access to significant computing resources. These resources can be expensive and have customarily been accessible only to tech companies able to negotiate substantial contracts with cloud providers.

Additionally, Trump has expressed a wish to partner with leading AI model developers to enhance the research community’s access to private AI models and datasets.

American AI companies and organizations that have adopted an open approach — including Meta, AI2, and Hugging Face — may benefit from Trump’s push for open AI.

AI safety and security

The AI Action Plan includes various initiatives designed to address concerns raised by the AI safety community. One major component is launching a federal technology development initiative focused on researching AI interpretability, control systems, and resilience against adversarial attacks.

Moreover, the plan directs federal agencies, including the Department of Defense and the Department of Energy, to organize hackathons aimed at evaluating their AI systems for security vulnerabilities.

It recognizes the potential risks posed by AI systems in enabling cyberattacks, as well as the creation of chemical and biological weapons. The plan calls upon leading AI model developers to work with federal agencies to assess these dangers and their consequences for national security.

In contrast to Biden’s AI executive order, Trump’s plan places less emphasis on requiring leading AI model developers to disclose safety and security standards. Many tech companies view such reporting obligations as burdensome, a sentiment that Trump seems inclined to alleviate.

Countering China’s influence

Unsurprisingly, Trump weaves his rivalry with China into the AI competition outlined in his action plan. A sizable portion of Trump’s AI Action Plan revolves around safeguarding against “national security” threats arising from advanced AI technologies.

Through this initiative, federal agencies are expected to collaborate in gathering intelligence on foreign AI efforts that pose threats to American national security. One specific task involves the Department of Commerce scrutinizing Chinese AI models to evaluate their compatibility with the Chinese Communist Party’s narratives and censorship efforts.

These teams will also conduct assessments regarding the level of AI adoption among rival nations.

National security

The term “national security” appears 23 times within the AI Action Plan — more often than terms such as “data centers,” “jobs,” “science,” and other essential concepts. The national security strategy laid out in the plan emphasizes integrating AI into the U.S. defense and intelligence frameworks while proactively countering foreign threats.

Among numerous directives, the plan requires that the Department of Defense and the intelligence community continually assess how U.S. AI adoption compares with competitors like China and adjust strategies accordingly, while also evaluating risks from domestic and foreign AI systems.

Within the Department of Defense, the strategy reiterates the importance of upskilling military personnel, automating operational procedures, and ensuring prioritized access to computational resources in times of national emergencies.