Trump's AI Brain Trust Divided on China Strategy

The Trump administration is developing AI policy without a unified interagency framework, instead relying on several officials scattered across Commerce, Treasury, the National Security Council, and the White House, each with distinct views on how to regulate AI and compete with China. One senior administration official described the current situation to WIRED as "an argument with ten sides," underlining the degree of internal disagreement.

Howard Lutnick, Commerce Secretary, has emerged as an influential figure partly because his department oversees export controls through the Bureau of Industry and Security. According to people familiar with his thinking, Lutnick has explored ways to create incentives for top US AI labs to develop their own open-weight models to counterbalance China, and he has spoken with leaders at multiple AI companies. On Thursday, Lutnick played down the capabilities of Moonshot AI's Kimi K3 model, writing on X that his team found it performed worse than leading US frontier models in key benchmark tests. Lutnick appears to occupy a middle position on regulation: he imposed export controls on Anthropic but has been more permissive than other White House officials. His acting deputy at the Center for AI Standards and Innovation (CAISI), Arvind Raman, assumed the role last week after his predecessor Chris Fall abruptly resigned following weeks of negotiations with Anthropic over safeguards for its Fable 5 model.

Sean Cairncross, the National Cyber Director, has taken a harder line on Chinese AI and is empowered at the White House to develop policy countering potential national security risks from AI. He helped assemble President Trump's June 2 executive order outlining a framework to assess the most powerful AI models. Although Cairncross is a former political campaign lawyer without tech or AI background, staff credit him with engaging seriously with AI's hard national security challenges and working without excessive micromanagement. He and his chief of staff, Lara Smith, have focused particularly on preventing Chinese labs from training models through distillation—the practice of copying US innovations.

David Sacks, a tech investor and Trump's former AI czar who departed in March, remains one of the administration's most influential AI advisers and maintains direct access to the president. Sacks has consistently advocated a hands-off approach, successfully intervening to water down some regulatory provisions in the June 2 executive order. Using his X account, which has 1.6 million followers, he has shaped the administration's stance from outside his formal role. On July 23, Sacks wrote: "Secretary Howard Lutnick is right. The Kimi Panic needs to stop. American frontier models are still ahead. When you factor in what's in the lab, the gap is even larger. As long as we keep releasing, we will stay ahead." That post generated nearly 500,000 views.

White House Chief of Staff Susie Wiles serves as Trump's top political aide in the policy process, vetting proposals from all sides before they reach the Oval Office. Treasury Secretary Scott Bessent and other senior officials take their positions directly to her, and her decisions on administration issues are typically final. Bessent has adopted perhaps the most aggressive stance toward Chinese AI. Last week, he posted on X that he repudiated distillation as "IP theft" and threatened Chinese labs with sanctions or inclusion on the Entity List, a US government trade-restriction list. After playing little role in restricting Anthropic last month, Bessent has stepped up his AI policy involvement ahead of Xi Jinping's planned White House visit in September and an APEC summit in November. He has told Wiles and others that the US, Europe, Japan, and other Western allies must accelerate development of their own open-weight models to avoid losing the AI race to China. Luke Pettit, a Treasury deputy who previously served as a senior policy adviser in the US Senate, has become the point person on AI at Treasury under Bessent, working alongside the Treasury's chief information officer and former DOGE member Sam Corcos. The Treasury Department overall has focused on the systemic risk of AI—specifically, the danger that a model inadvertently closing trading positions or accessing sensitive information at banks could crash the economy.

The White House spokesperson Elizabeth Huston, responding to requests for comment, stated that the administration's deliberations on Chinese AI are centered on bolstering US AI: "The United States leads the world in AI innovation, and President Trump will keep it that way. The Trump Administration is doubling down on innovation to widen the gap between America and the rest of the world."

The Trump administration's approach to AI regulation is being shaped not by a unified interagency process but by a collection of officials with competing philosophies and access to the president. As one senior White House official told WIRED, the debate is "an argument with ten sides," reflecting the absence of coordinated policy-making. The fracture reveals two broad camps: those favoring a lighter regulatory touch (Sacks, who maintains a direct line to Trump despite leaving his formal AI czar role) and those pushing stricter controls on China and distillation practices (Bessent at Treasury, Cairncross at the National Cyber Directorate). Lutnick occupies the middle, having imposed export controls on Anthropic while exploring incentives for US companies to build competing open-weight models.

The stakes are high because China's open-weight AI models are growing more capable, and US officials see model distillation—the practice of training Chinese models on US innovations—as a core threat to American technological supremacy. Bessent has escalated his involvement ahead of Xi Jinping's planned White House visit in September and the APEC summit in November, suggesting that China policy will be a major diplomatic focus. The absence of a single decision-maker or coordinated process means that policy outcomes may depend less on strategic coherence than on which official has the strongest relationship with the president on any given issue.

No comments yet. Be the first to share your thoughts!