Inside the Trump Administration’s Fragmented AI Policy Strategy Regarding China

The Trump administration currently faces significant internal friction regarding the regulation of artificial intelligence. Officials describe the situation not as a binary debate, but as a multi-faceted conflict involving numerous agencies and stakeholders with conflicting philosophies. Because there is currently a lack of cohesive interagency coordination, policy outcomes appear increasingly dependent on which individuals hold the most influence with the president at any given time. This power struggle spans several departments, with top officials attempting to balance the need for domestic AI innovation with the strategic objective of maintaining a technological lead over China. Observers within the administration have characterized the internal climate as highly complex, with various factions advocating for disparate approaches ranging from laissez-faire market freedom to aggressive sanctions and export restrictions.

Commerce Secretary Howard Lutnick has emerged as a central figure in the administration’s AI strategy, primarily due to his oversight of export controls via the Bureau of Industry and Security. Lutnick maintains a middle-ground position, having previously enforced restrictions on firms like Anthropic while simultaneously exploring ways to incentivize American labs to develop open-weight models to compete with Chinese counterparts. Supporting him is Arvind Raman, who recently took over as the acting chief of the Center for AI Standards and Innovation (CAISI). Raman assumed this role following the abrupt resignation of Chris Fall, who had spent considerable time negotiating safety guardrails for Anthropic’s Fable 5 model. These meetings were critical, as they involved intensive technical collaboration to ensure that powerful models could remain operational while meeting the government’s security standards for preventing jailbreaks.

National Cyber Director Sean Cairncross and Treasury Secretary Scott Bessent are driving the more aggressive, security-focused wing of the administration. Cairncross, who lacks direct technical experience, is nonetheless instrumental in coordinating White House responses to national security risks, specifically targeting the practice of distillation, where Chinese labs utilize US-developed model architectures for their own training. Bessent has escalated this stance further, labeling model distillation as intellectual property theft and threatening to utilize the US Entity List to impose sanctions on foreign labs. Additionally, the Treasury Department is expressing concerns over systemic economic risks, such as the potential for rogue AI to destabilize financial markets or compromise sensitive institutional data, with Luke Pettit serving as a key operational lead for these initiatives.

Despite his formal departure as the administration's AI czar in March, investor David Sacks remains a highly influential voice who consistently advocates for a deregulatory, hands-off approach to AI development. Sacks has maintained a direct line to the president, successfully lobbying to weaken specific regulatory provisions in the June 2 executive order. He frequently utilizes his significant social media presence to publicly validate the positions of officials like Howard Lutnick, specifically countering fears regarding the progress of models such as Moonshot AI’s Kimi K3. Meanwhile, White House Chief of Staff Susie Wiles acts as the final gatekeeper for these competing proposals. Wiles is responsible for vetting and scrutinizing the strategies presented by various cabinet secretaries before they reach the Oval Office, effectively holding the ultimate decision-making power in the administration's policy hierarchy.