Politics & Policy

A Great Start for Trump on AI

President Donald Trump speaks at a podium during the “Winning the AI Race” summit in Washington, D.C.
President Donald Trump delivers remarks on artificial intelligence at the “Winning the AI Race” summit in Washington D.C., July 23, 2025. (Kent Nishimura/Reuters)

On Wednesday, during his appearance at the “Winning the AI Race” summit in Washington, D.C., Donald Trump launched the White House’s AI Action Plan. Per the executive orders that were signed during the event, the president’s project contains three major planks. First, it establishes a set of procurement rules that prohibit the federal government from purchasing any AI product that, in Trump’s words, “has been infused with partisan bias or ideological agendas.” Second, it seeks to “expedite construction of all major AI infrastructure projects,” by waiving or reversing Biden-era environmental obstacles and “easing Federal regulatory burdens.” Third, it encourages the export of American hardware and software, “to preserve and extend American leadership in AI and decrease international dependence on AI technologies developed by our adversaries.” All of these policies are within the federal government’s remit, and all three are welcome on the merits.


The White House acknowledged that “the Federal Government should be hesitant to regulate the functionality of AI models in the private marketplace,” but proposed that, “in the context of Federal procurement, it has the obligation not to procure models that sacrifice truthfulness and accuracy to ideological agendas.” The success of this policy will be entirely contingent upon the execution, but there is nothing in the order that ought to raise alarm. As an example of pernicious wokeness “in the AI context,” the president cited Google Gemini’s much-mocked post-launch habit of changing “the race or sex of historical figures — including the Pope, the Founding Fathers, and Vikings — when prompted for images because it was trained to prioritize DEI requirements at the cost of accuracy.” Among the categories that it singled out for scrutiny are “the suppression or distortion of factual information about race or sex; manipulation of racial or sexual representation in model outputs; incorporation of concepts like critical race theory, transgenderism, unconscious bias, intersectionality, and systemic racism; and discrimination on the basis of race or sex.” There is not — and there will never be — such a thing as a perfectly neutral artificial intelligence, but keeping some of the real world’s most destructive dogmas out of government-used software is a laudable aim.




So, too, is the facilitation of “the rapid and efficient buildout” of AI infrastructure. In essence, Trump wishes to make sure that the construction of the many data centers and transmission lines that will be necessary for the United States to dominate the global AI industry is not slowed down by red tape. Specifically, Trump has instructed the director of the EPA, the secretary of interior, and the secretary of energy to ensure that any environmental reviews are swift, to expedite the relevant permitting processes on both federal and private land, and to develop or modify regulations under the Clean Air Act and the Clean Water Act so that they cannot be used as a brake.


Finally, the president expressed a desire to secure America’s “continued technological dominance” in technology, by exporting American hardware, software, and — crucially — law. As well as the physical systems, Trump wishes to replicate the success that the United States has had with the internet more broadly, and to create a global framework within which “American AI technologies, standards, and governance models are adopted worldwide.” Naturally, the United States cannot guarantee that other countries will acquiesce to this desire. But, as the last 30 years have shown, America’s uniquely open culture is a worthy global template, and there is nothing at all wrong with the federal government recommending it at every available opportunity.

During the question-and-answer section of the event, Trump offered a couple of comments that have overshadowed the policies that he was there to promote. When asked whether the federal government ought to preempt state law in this area, as Congress had considered doing in the “big beautiful bill,” Trump submitted that in his view, “we need one commonsense federal standard that supersedes all states.” When asked about the problem of copyright infringement, Trump said that “you can’t be expected to have a successful AI program when every single article, book or anything else that you’ve read or studied, you’re supposed to pay for, you just can’t do it because it’s not doable.” “China,” he added, “is not doing it.”


Neither of these issues was addressed in Trump’s orders. But, at some point soon they will need to be resolved. When they are, Congress ought to split the baby. AI is, indeed, a federal concern — the internet is the very definition of an instrumentality of interstate commerce — and, if Washington’s preeminence in the area is not already covered by the 1934 Communications Act and the Telecommunications Act of 1996, Congress should pass a law guaranteeing it. Likewise, the application of federal environmental and copyright laws are federal concerns. At no point in the history of the internet has national preemption been controversial, and it should not be here, either. Copyright, by contrast, is a great deal trickier than Trump implies. It is true that China does not respect the concept. But this does not obviate the very real problems that will be caused in America if AI services are permitted to absorb without recompense — and thereby to render obsolete or superfluous — the work of millions of people on the existing internet. At present, the web works because those who create and maintain its content are rewarded with the monetized visits of those who consume that content. By bypassing those creators completely — while benefiting from their work — AI presents an existential threat to that model. (And, because it is reliant on those creators for its information, it presents a threat to itself.) President Trump may well believe that it is not “doable” for AI firms to consider copyright, but, practically, it is also not “doable” for them to ignore it. Balancing these two concerns is a job for Congress, not the White House, to undertake.

Those, though, are questions for another day. At the outset of the event, President Trump said that he had attended “today to declare that America is going to win” the global AI race. His three orders represent an excellent start to that endeavor.

The Editors comprise the senior editorial staff of the National Review magazine and website.
Exit mobile version