It was like watching the Dallas Mavericks trade away five-time All-Star Luka Dončić while declaring their NBA championship aspirations. That’s precisely the cognitive dissonance the nation witnessed during the February 25th confirmation hearing for Michael Kratsios to become Office of Science and Technology Policy (OSTP) Director.
The nominee (in a “double date” confirmation hearing with Mark Meador to become FTC Commissioner) passionately claimed to champion American AI dominance over China while steadfastly refusing to defend the scientific institutions being dismantled around him or articulate a vision for AI innovation, workforce development or the use of AI in government.
Having had the honor to lead government modernization initiatives at OSTP under John Holdren’s leadership, I watched with both interest and dismay as senators failed to challenge a strategy as self-defeating as trading away a generational talent while claiming to pursue greatness.
Rhetoric vs. Reality: Competing with China While Cutting America's Advantages
China loomed large throughout the hearing, with both nominees and senators expressing alarm about Chinese technological advancement. Kratsios emphasized in his opening remarks that "China has emerged as both our preeminent geopolitical rival and our most formidable technological and scientific competitor," a sentiment echoed by multiple senators concerned about models like DeepSeek.
Kratsios presented his strategy as "promote and protect"—enforcing export controls while spurring innovation. But when pressed on specifics, his answers revealed a remarkably narrow vision limited to restricting access to chips while relaxing regulation for American firms.
When Senator Jacky Rosen (D-NV) asked about collaborating with the private sector on AI security requirements for models like DeepSeek, Kratsios spoke about measurement standards at the National Institute of Standards and Technology (NIST)—yet refused to acknowledge, let alone criticize, the massive layoffs that agency is facing.
During the same week as the hearing, the White House Office of Management and Budget directed agencies to submit plans for "maximum elimination of functions that are not statutorily mandated"—a directive that could result in the dismantling of crucial research programs that don't have explicit legislative protection. "We actually fired people who just got promoted because of their outstanding work," noted Senator Gary Peters (D-MI).
The National Science Foundation (NSF) is anticipating 66 percent budget cuts, yet Kratsios responded with vague assurances but no commitments to protect scientific talent.
This approach amounts to competing against China with one hand tied behind our back: restricting technology transfers while simultaneously dismantling America's research capabilities.
Deregulation as Industrial Policy: The Poverty of the Approach
The administration's vision of AI leadership appears primarily centered on deregulation, not X Senator Ted Cruz (R-TX) criticized the Biden administration for focusing on AI risks, including misinformation and model bias. Senator Schmitt (R-MO) characterized concerns about AI bias as "overregulation" designed to foster AI models that would be considered more progressive.
Such framing transforms technical questions about AI system performance into political battlegrounds—a shift that will hamper America's ability to build trustworthy, widely deployable AI systems. Kratsios avoided defending these important research areas.
Conspicuously missing was any substantive vision for how the government might actively foster an AI ecosystem—beyond simply removing guardrails. There was no discussion of explainable or transparent AI, or strategies to ensure American AI innovation serves broader societal needs. There was no discussion--and shockingly no questions--about investing in building public and accountable AI as they are doing in Switzerland or Sweden. The approach represents industrial policy at its most minimal: get out of the way and hope for the best.
The Elon in the Room
Remarkably, no senator questioned Kratsios about Elon Musk's expanding role in government technology policy through the Department of Government Efficiency (DOGE), which is seeking to use AI developed by Musk’s company xAI for government functions. These potential conflicts of interest and influence over government AI policy went unexamined in the hearing.
This oversight is particularly troubling given OSTP's traditional role in coordinating technology policy across government. If OSTP is sidelined while private sector figures with commercial interests shape federal AI policy, the results could further skew toward deregulation without consideration of the public interest.
The Missing Modernization Agenda
Perhaps the most telling absence in the hearing was any substantive discussion of using AI to improve government itself. During the Obama administration, OSTP played a central role in initiatives like open data, digital services, and creating the Presidential Innovation Fellows program—all of which both modernized government operations and fueled innovation in the private sector.
Under President Biden, the administration championed an ambitious AI agenda that included responsible innovation, guardrails against harmful applications, and efforts to use AI for public benefit. Executive Order 14110 created a comprehensive framework for safe, secure, and trustworthy AI development, while also directing agencies to use AI to improve government services.
By contrast, Kratsios offered no comparable vision for using technology to make government work better for citizens. He also had nothing to say about AI upskilling for the public sector. The silence on this front suggests that OSTP under his leadership would retreat from its traditional role as an advocate for digital government transformation.
A Rubber Stamp Hearing
The general tone of the hearing was notably unchallenging from Republican senators, who largely lobbed softball questions, allowing Kratsios to repeat talking points about American technological leadership without explaining how the administration’s gutting of scientific agencies would achieve this goal. Democratic senators pressed harder on specific contradictions, but the joint format with FTC nominee Meador diluted focus and accountability.
This lack of scrutiny suggests that Congress may provide little oversight as the administration reshapes federal science and technology policy. When Senator Cantwell (D-WA) asked pointedly about potential 66% cuts to NSF's budget, Kratsios would only commit to "relay[ing] the importance of research and development" in White House discussions—hardly a robust defense of American science.
Executive Control and the End of Independence
Beyond the direct AI policy questions, the hearing revealed troubling trends about the regulatory independence of both nominees. Senator Tammy Duckworth (D-IL) pressed Meador on the administration's recent position that FTC commissioners should be fireable at will by the president, contrary to nearly a century of precedent established in Humphrey's Executor.
When Meador repeatedly refused to take a position, Baldwin noted his "sucking up to the president" while avoiding a direct answer.
This consolidation of executive power over independent regulatory agencies has profound implications for AI governance. If agencies tasked with overseeing AI applications and impacts become mere extensions of White House policy preferences, America will lose crucial guardrails at precisely the moment when technological capabilities are accelerating.
The Cost of Abandoning Innovation Policy
The contrast with previous administrations' approaches to technology is stark. Under Obama, OSTP coordinated major initiatives including the National Nanotechnology Initiative, the Materials Genome Initiative, the BRAIN Initiative, and robust climate science programs. These efforts recognized that the government plays a crucial role in de-risking frontier research and creating the conditions for private sector innovation.
The Biden administration continued this tradition, with significant investments in semiconductor manufacturing, quantum information science, and developing regulatory frameworks for responsible AI innovation that would both protect Americans and give US companies a competitive advantage through trustworthiness.
Kratsios's testimony suggests a retreat from this proven approach—replacing active innovation policy with passive deregulation. History shows this approach is unlikely to maintain American leadership. The semiconductor industry wasn't built through deregulation alone but through decades of federal research funding, defense procurement, and strategic industrial policy.
A Self-Defeating Strategy
The most fundamental contradiction in the vision presented at the hearing is that it undermines its own stated goal. If American AI dominance is truly a "national security imperative," as Kratsios quoted President Trump saying, then gutting the scientific agencies that make that dominance possible is self-defeating.
You cannot lead the world in AI while firing your scientists, defunding your research institutions, abandoning efforts to ensure AI systems work reliably and fairly, and reducing technology policy to export controls and deregulation.
As the confirmation process proceeds, those concerned about America's AI future should look beyond the patriotic rhetoric to the actual policies being implemented. The stakes could not be higher: in a world where AI capabilities are increasingly determinative of economic and strategic advantage, the narrow vision presented at this hearing risks ceding American leadership at precisely the moment when it matters most.