Trump gets aggressive with new US AI action plan

The Trump administration on Wednesday outlined how it intends to win the global artificial intelligence race, with the launch of a new strategy that involves exporting US-developed AI, enabling innovation and adoption, and promoting a speedy buildout of data centers.

Those details and others are contained in a 28-page document, Winning the AI Race: America’s AI Action Plan, that, according to an accompanying press release, defines moves that “will usher in a new golden age of human flourishing, economic competitiveness, and national security for the American people.”

The plan, it states, identifies more than 90 federal policy actions across three pillars: Accelerating Innovation, Building American AI Infrastructure, and Leading in International Diplomacy and Security, that the US government plans to execute in the coming months.

Highlights of the plan include:

• Partnerships between the US Commerce and State department and industry to “deliver secure, full-stack AI export packages including hardware, models, software applications and standards to America’s friends and allies around the world.”
• Measures to expedite and modernize permits for data centers and semiconductor fabs and developing new initiatives to increase the number of people in “high-demand occupations” such as electricians and HVAC technicians.
• The removal of what it called “onerous federal regulations that hinder AI development and deployment”, as well as seeking private sector input on which rules to remove.
• Work with all federal agencies “to identify, revise or repeal regulations, rules, memoranda, administrative orders, guidance documents, policy statements, and interagency agreements that unnecessarily hinder AI development or deployment.”

DEI, inclusion and climate change are out

Bill Wong, research fellow at Info-Tech Research Group, said of the fourth highlight, “while the rest of  the world is looking to regulate AI to ensure safety, the US is taking an opposite approach.”

He also questioned two additional recommendations.

The first was the revision of the NIST AI Risk Management Framework (AI RMF) that will “eliminate references to misinformation, Diversity, Equity and Inclusion, and climate change”.

The claim, said Wong, is that “by removing all three, it will ensure that frontier AI ‘protects free speech and American values’. [But] they [the references] are usually part of responsible AI principles, which this administration has claimed introduces ideological bias or engineered social agendas. When did ‘misinformation’ become something that you want to enable?”

The second recommendation involves the update of federal procurement guidelines to specify that the government-only contracts with large language model (LLM) developers ensure that their systems are objective and free from top-down ideological bias. “Who decides what is objective and free?” asked Wong. “Does that mean all the LLM vendors have to remove the Responsible AI principles initiative?”

‘No reason to weaken the AI RMF’

Samir Jain, vice president of policy for the Center for Democracy & Technology, described the AI Action Plan as “a missed opportunity.”

He said, “while promoting innovation is important, [it] whiffs by failing to include measures to help ensure that AI development and deployment happens responsibly and addresses potential harms. Instead, the Plan includes actively detrimental provisions.”

For example, said Jain, “the administration should not hinder state officials in their efforts to respond to the real and documented harms created by AI. The government should not be acting as a Ministry of AI Truth,” nor should it insist that AI models adhere to its preferred interpretation of reality.

There is, he said, no reason to weaken the AI RMF by eliminating references to some of the real risks that AI poses. “To be sure, the Plan has some positive elements,” he said, “including promotion of open-source and open-weight systems, support for building an AI evaluations ecosystem, and an increased focus on the security of AI systems.” But ultimately, he said, it is “highly unbalanced, focusing too much on promoting the technology while largely failing to address the ways in which it could potentially harm people.”

‘An important step’

Matthew Eastwood, senior vice president of several technology research groups at IDC, including enterprise infrastructure and cloud, described Trump’s Plan as “a significant departure from what the [Biden] administration was leaning on specifically, trying to remove a number of friction points for enterprises. But having said that, it does open up some big questions around things like oversight, equity, long term security and whatnot.”

Despite those concerns,  he said, “based on where we are with the development of AI as a forcing function for our economy and for the need for a healthy ecosystem around that, it’s an important step.”

However, Eastwood said that he anticipates that “we will see pushback from civil groups like the American Civil Liberties Union (ACLU) and potentially legal challenges to some of the federal efforts that are being put forth here, and by states as well.”

Falk Gottlob, chief product officer at AI platform vendor Smartcat, thinks the plan is a positive move. He said, “[the plan] sets a clear direction for US leadership in artificial intelligence, but lasting impact hinges on effectively addressing the operational realities enterprises face during adoption. Success for most organizations requires more than advanced models or technical resources. It demands platforms that ensure secure deployment, enforce strong governance, and uphold data integrity.”

He added that the plan is just part of the overall AI equation. “AI has immense potential to democratize innovation. It broadens access to knowledge, dismantling barriers historically limiting participation and progress,” he said. “To fully capture this potential, ongoing investment in responsible deployment practices must complement national policy initiatives. The AI Action Plan is an important step forward, but its ultimate success will depend heavily on translating these strategic priorities into tangible, operational reality.”