Trump Administration Faces Internal Conflict Over AI Vetting
Trump administration is sharply split over whether to give spy agencies or Commerce Department control of AI model evaluation in what one official described as a "knife fight."
Objective Facts
President Trump's administration is sharply split over a plan to give U.S. intelligence agencies a bigger role in evaluating AI models, with the debate described as a "knife fight" between Commerce Department officials and national security aides over regulatory authority. The Office of the National Cyber Director has proposed developing a large center within the Office of the Director of National Intelligence to evaluate new AI models, giving intelligence agencies significant new power over AI policy. The Commerce Department's CAISI has built documentation showing it has a documented record of vetting large language models and has hired AI experts, with some in the agency arguing this infrastructure makes it better positioned to lead on testing. National Economic Council Director Kevin Hassett's suggestion of FDA-style AI vetting was quickly tamped down by White House Chief of Staff Susie Wiles and former AI czar David Sacks, who argued the real issue is hardening U.S. systems against Chinese capabilities within six months. Disagreement among administration officials and Trump's timing for a China summit with President Xi Jinping are holding up any immediate federal response, with early talk of federal safety reviews slowing as it becomes clear Trump administration officials aren't on the same page.
Left-Leaning Perspective
Left-leaning outlets like The Washington Post and Axios reported on the internal conflict with emphasis on the administration's chaotic reversal from its deregulatory stance. Helen Toner, interim executive director at Georgetown Center for Security and Emerging Technology, told CBS News that David Sacks' deregulation approach "was not a…well-considered whole of government, really thoroughly endorsed position on how to do AI regulation, but just he was the person in the room who had a lot of thoughts." Progressive analysts argue the fundamental problem is that any government vetting system built now risks weaponization by future administrations regardless of partisan affiliation. As noted by Cornell professor Kristen Kreps in an analysis, "The challenge is doing the coordination without building an approach that is either quickly obsolete because of the fast-moving technology or that gets weaponized by the next administration." Critics like those cited in the Slopagandist newsletter point out the administration "spent a year calling AI safety regulation unnecessary government overreach" and now "quietly pivoted to pre-deployment testing" after Mythos forced a reassessment. Left-leaning coverage has also highlighted the absence of clear democratic oversight or public input in these internal debates, emphasizing that whoever controls AI vetting—whether Commerce or intelligence agencies—will effectively control which companies succeed or fail in AI development without proper congressional authorization or transparency.
Right-Leaning Perspective
Right-wing commentary, while sparse on this specific internal conflict, emphasizes through figures like David Sacks and Kevin Hassett that any vetting must prioritize U.S. competitiveness against China over safety concerns. The Daily Signal reported the internal differences as reflecting officials who "prefer a light touch to regulation while others want to aggressively vet new models," framing the light-touch advocates' position more favorably. Conservative policy analysts at the Abundance Institute and R Street Institute argued that an FDA-style approval process would represent a dangerous reversal of Trump's stated deregulatory goals. They emphasized that the administration correctly "rejected the Biden-Harris approach to AI" and should not abandon that principle under pressure from national security officials. Right-leaning coverage has characterized the vetting debate as an unnecessary distraction from the real priority: preventing China from gaining AI dominance. David Sacks' Fox Business appearance emphasized that U.S. officials should focus on "how future AIs that also potentially create vulnerabilities should go through a process" but explicitly framed this as defensive (hardening U.S. systems) rather than as approval-based regulation. Right-wing outlets largely avoided covering the Commerce vs. intelligence agency turf war, instead focusing on the policy outcome—avoiding bureaucratic approval processes—rather than on which agency should hold vetting authority.
Deep Dive
The internal Trump administration conflict over AI vetting authority reflects a deeper tension between national security absolutism and tech-industry accommodation. The intelligence community's proposal to centralize AI evaluation in the Office of the Director of National Intelligence reflects Cold War-era thinking: specialized agencies with threat expertise should control critical technologies. Commerce Department officials counter that their established CAISI infrastructure, existing relationships with tech companies, and mandate to promote U.S. competitiveness make them better stewards of AI policy. Neither side is wrong on the merits—intelligence agencies do possess unique threat modeling expertise, while Commerce has built functional evaluation systems and maintains industry trust essential for voluntary cooperation. What both the Washington Post and Axios reporting reveal is that this turf war exists because the Trump administration failed to establish clear leadership on AI policy when David Sacks departed in March 2026. Sacks embodied a deregulatory vision, and when he left the formal czar role (though remained on PCAST), power fragmented across the National Security Council, the National Economic Council, Commerce, and various intelligence agencies. Kevin Hassett's FDA comparison comment and the subsequent walk-backs from Wiles and Sacks reveal officials improvising responses to Mythos without a coordinated framework. The intelligence community saw an opening to expand its authority; Commerce fought back; deregulation-focused officials like Sacks used national security language (defending against China) to argue against approval authority without admitting policy reversal. What comes next likely depends on Trump's personal preference, which has not been clearly stated. His past behavior suggests he could go either direction—expanding executive power through intelligence agencies (consistent with his authoritarian instincts) or maintaining lighter regulation to reward tech allies who support him. The delayed announcement as he prepared for China talks in May suggests the issue remains genuinely unresolved. One tech industry source told Axios the administration may wait on the outcome of the China summit before making final decisions on AI. This delay suggests neither side has won the internal debate, and resolution may depend on external pressure—either from China's AI advances or from public/congressional reaction to Mythos capabilities.