White House Announces National Plan to Prioritize AI Growth and "Made in America" Exports, Facing Criticism over Safety Trade-offs.
The White House released the AI Action Plan on Wednesday, aiming to strengthen America's AI dominance by accelerating infrastructure deployment and technology exports. The plan follows the $500 billion Stargate Project led by OpenAI, Oracle, and SoftBank, but has faced opposition from civil rights organizations and policy experts who argue the strategy sacrifices safety for economic power.
Michael Kratsios, Director of the White House Office of Science and Technology Policy, emphasized that the plan represents a comprehensive federal effort to accelerate innovation capacity, build advanced infrastructure, and provide global leadership. He affirmed that the administration is acting with urgency to ensure American workers and families thrive in the AI era.
The plan includes four main pillars. First, exporting US AI, with the Commerce and State Departments collaborating with the private sector to provide safe "full-stack" AI systems, including hardware, software, and standards for allied nations. Second, accelerating infrastructure deployment by expediting licensing for data centers and chip manufacturing plants, and launching workforce training programs. Third, reducing regulations to eliminate federal rules that slow AI development. Finally, ensuring free speech in AI by requiring federal contracts to be free from government-imposed ideological bias.
Controversy over AI Politicization
Matthew Mittelsteadt, a technology policy research expert at the Cato Institute, called the plan a "confusing policy package". He welcomed the emphasis on rapid innovation compared to Biden-era orders but warned about the risk of political abuse. According to Mittelsteadt, linking federal contracts to government-oriented models could politicize the entire AI field, causing developers to build systems to please the administration.
He warned that if US AI models begin reflecting a political agenda, international users might view them as American influence tools rather than neutral technology. Mittelsteadt emphasized that anyone wanting to use US AI models abroad would perceive them as influenced by the US government, similar to how the world views Chinese models as Beijing government tools.
The plan also threatens to halt federal AI funding for states with regulations deemed "obstructive", which critics argue would stifle local innovation and spark new constitutional disputes. Mittelsteadt expressed skepticism about enforcing this threat, especially for states with large tech industries like California, where most funding goes to federal entities like national laboratories.
Eric Null, Co-Director of the Privacy & Data Project at the Center for Democracy & Technology, argued that the AI Plan could undermine effective AI regulation. He noted that completely eliminating ideological bias from AI is extremely difficult, and regulation enforcement would be complex. Null warned that under a partisan administration, this could lead to unfair implementation, treating some perspectives as biased while overlooking others.
