White House unveils “America’s AI Action Plan”
A sweeping plan for government-backed AI may sweep aside state regulations.
The Trump administration wants to usher in “a new golden age of human flourishing” powered by AI that the government ensures will be folded into every part of our lives.
“AI will enable Americans to discover new materials, synthesize new chemicals, manufacture new drugs, and develop new methods to harness energy — an industrial revolution. It will enable radically new forms of education, media, and communication — an information revolution. And it will enable altogether new intellectual achievements: unraveling ancient scrolls once thought unreadable, making breakthroughs in scientific and mathematical theory, and creating new kinds of digital and physical art — a renaissance.”
The White House’s grand plan for the US to fully embrace AI, announced on Wednesday, calls for the nascent technology to be deeply integrated across all corners of government and society.
Based on three conceptual pillars — accelerating innovation, building AI infrastructure, and diplomacy and security — the 23-page document is a blindingly bright green light for America’s tech sector to create a “try-first” culture for AI across US industry.
The ambitious plan identifies many ways that federal agencies can partner with US tech companies to guarantee the country’s global dominance in AI (and ensure healthy streams of federal revenue along the way).
It even calls for government “priority access to computing resources” during national emergencies or significant conflicts.
But while the initiative signals full speed ahead for using AI in pretty much every facet of the US government, it also awkwardly acknowledges the current dim understanding of how AI works and what potential risks there might be.
“Today, the inner workings of frontier AI systems are poorly understood. Technologists know how LLMs work at a high level, but often cannot explain why a model produced a specific output. This can make it hard to predict the behavior of any specific AI system. This lack of predictability, in turn, can make it challenging to use advanced AI in defense, national security, or other applications where lives are at stake.”
Plans for thwarting state regulation
While an audacious proposal to impose a 10-year ban on state regulation of AI was killed in Congress via amendments to President Trump’s massive tax bill, the federal government is sketching out a strategy for how to deal with any pesky sub-federal regulations that may slow the roll of AI. (They’re looking at you, California.)
The document outlines that federal AI dollars should be withheld from “states with burdensome AI regulations,” while saying it should not interfere with states’ “prudent laws” that don’t restrict innovation.
The plan directs the Office of Management and Budget to “consider a state’s AI regulatory climate when making funding decisions and limit funding if the state’s AI regulatory regimes may hinder the effectiveness of that funding or award.”
The Federal Communications Commission is directed to “evaluate whether state AI regulations interfere with the agency’s ability to carry out its obligations and authorities under the Communications Act of 1934.’’
Don’t let our enemies get our AI tech, unless we sell it to them
“America currently is the global leader on data center construction, computing hardware performance, and models. It is imperative that the United States leverage this advantage into an enduring global alliance, while preventing our adversaries from free-riding on our innovation and investment.”
The plan confronts a tricky balancing act that has captured the attention of investors backing AI-adjacent companies.
The government considers America’s leading AI tech to be essentially national assets that should be kept from our enemies and shared with our friends. But the American companies making this tech (like market leader Nvidia) don’t want to miss out on a massive market like China.
The document equivocates on this point and, with words you could imagine Nvidia CEO Jensen Huang whispering to Trump at Mar-a-Lago, encourages “creative approaches to export control enforcement.”
Build, baby, build
Despite hundreds of billions of Big Tech dollars flowing into the ever-larger data center projects that are under construction around the US at breakneck speeds, it’s not fast enough for the AI advocates in the Trump administration.
Declaring that America’s regulatory permitting processes make it “almost impossible to build this infrastructure in the United States with the speed that is required,” the plan suggests removing any and all guardrails for these data centers, despite environmental concerns.
The plan notes the massive amounts of electricity that the aging US grid must accommodate for AI data centers, but while it calls for “new sources of energy to power it all,” there is not a single reference to “renewable energy” in the document, such as solar or wind (though it does reference geothermal power).
Many of the biggest data center projects — such as those being built by Meta— include the creation of new renewable energy plants for the communities where they are built. Trump’s distaste for solar and wind sources of power are reflected in the document.
The initiative also prioritizes domestic chip manufacturing, which was a key accomplishment of the Biden administration. The text mentions the “revamped” CHIPS Program Office, which it says should “continue focusing on delivering a strong return on investment for the American taxpayer.”
Rapid retraining
One of the biggest fears of the effect of widespread AI adoption is the potential for massive disruptions to the labor force, as some jobs are lost to automation.
The plan tasks the Department of Labor with addressing this serious challenge. It outlines a “worker-first” AI agenda, which incentivizes AI literacy and skills development, as well as a mandate to “fund rapid retraining for individuals impacted by AI-related job displacement.” The DOL will also have to pilot new approaches to “shifting skill requirements for entry-level roles.”
But while a robot may take your job, it won’t likely face much regulation. The text encourages clearing obstacles for America to lead in the manufacturing of “autonomous drones, self-driving cars, robotics, and other inventions for which terminology does not yet exist.”
NIST’s many roles
There are many calls for the National Institutes of Standards and Technology (NIST) to do a lot of the work in this plan. Among its responsibilities:
“Publish evaluations of frontier models from the People’s Republic of China for alignment with Chinese Communist Party talking points and censorship.”
“Revise the NIST AI Risk Management Framework to eliminate references to misinformation, Diversity, Equity, and Inclusion [DEI], and climate change.”
Measure AI productivity at “realistic tasks” in various industries.
Create “automated cloud-enabled” labs for AI-powered scientific testing.
Build an AI evaluations ecosystem, and “support development of the science of measuring and evaluating AI models.”
Assess deepfake evaluation systems to protect the courts and law enforcement from AI-generated evidence.
“A missed opportunity”
Samir Jain, the VP of policy at the nonprofit Center for Democracy and Technology, took issue with many of the initiatives described in the plan, such as the removal of references to climate change and DEI from NIST’s AI Risk Management Framework. Jain told Sherwood News in an emailed statement:
“The government should not be acting as a Ministry of AI Truth or insisting that AI models hew to its preferred interpretation of reality.”
Jain praised the plan’s calls for improved AI evaluation systems, open-source and open-weight AI models, and focus on AI security, but on the whole described the proposal as a missed opportunity.
“Ultimately, the Plan is highly unbalanced, focusing too much on promoting the technology while largely failing to address the ways in which it could potentially harm people.”
