Trump's AI Policy Divided Among Commerce, Cybersecurity Officials With Conflicting China Strategy
Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross lead competing factions shaping US AI export controls and regulation, with little coordination across agencies.
Last verified:
The Trump administration’s approach to AI regulation remains deeply fractured, with competing officials pursuing conflicting strategies on Chinese AI containment and domestic open-weights incentives. According to Wired AI, Commerce Secretary Howard Lutnick and National Cyber Director Sean Cairncross lead opposing camps, and without formal interagency coordination, the final policy direction will likely depend on which officials maintain the president’s ear.
Lutnick’s Middle-Ground Export Control Strategy
Commerce Secretary Howard Lutnick has positioned himself as a pragmatist on AI regulation, overseeing export controls through the Bureau of Industry and Security while simultaneously exploring market-based alternatives. According to Wired AI, Lutnick has privately discussed ways to incentivize leading US AI labs to develop their own open-weights models as a counterweight to Chinese advances. His office has engaged directly with AI lab leadership on this proposal.
However, Lutnick’s flexibility has limits. He imposed export controls on Anthropic earlier in 2026 to enforce compliance with government safeguarding requirements. On July 24, he went public with skepticism about Chinese capability, posting on X that his Commerce team evaluated Moonshot AI’s Kimi K3 model and found it underperformed US frontier models on key benchmarks—a claim difficult to independently verify given the benchmark specifics were not disclosed.
Cairncross’s Hardline Approach to Chinese AI Threats
In contrast, National Cyber Director Sean Cairncross has advocated for stricter restrictions on Chinese AI systems. According to Wired AI, Cairncross holds an outsized White House role in developing policy to counter what he frames as national security risks from overseas AI labs. His authority over cybersecurity and emerging technology policy gives him direct influence on potential restrictions.
Internal Friction Over Anthropic’s Fable 5 Model
The administration’s lack of coordination surfaced in recent negotiations over Anthropic’s flagship Fable 5 model. Wired AI reports that Arvind Raman, who assumed the role of acting chief at the Center for AI Standards and Innovation (CAISI) last week, took over negotiations after his predecessor Chris Fall resigned following weeks of contentious back-and-forth with Anthropic over jailbreak safeguards. Both CAISI and Anthropic’s technical teams eventually reached agreement on safeguards deemed sufficient to return Fable 5 to online availability.
Why This Matters
The absence of unified interagency coordination creates both risk and opportunity for AI companies operating in the US. Organizations like Anthropic must navigate conflicting signals from Commerce and cybersecurity officials—complying with one faction’s demands may not satisfy another’s. Meanwhile, the administration’s interest in incentivizing open-weights development (per Lutnick’s discussions) could reshape funding and competitive dynamics if formalized into policy. The final direction depends less on bureaucratic consensus than on which officials retain President Trump’s confidence on AI matters.
Frequently Asked Questions
Who are the key figures shaping Trump's AI policy in 2026?
Commerce Secretary Howard Lutnick, Acting Director of the Center for AI Standards and Innovation Arvind Raman, and National Cyber Director Sean Cairncross are among the primary power brokers. According to Wired AI, they represent competing factions with differing views on export controls and Chinese AI competition.
What is the main disagreement within the Trump administration on AI?
The core tension is between incentivizing US labs to build open-weights models (Lutnick's approach) and imposing stricter export controls on Chinese AI (Cairncross's hardline position). A senior White House official told Wired AI the policy debate has 'ten sides,' not two.
How has Commerce Secretary Lutnick responded to Chinese AI models?
On July 24, Lutnick downplayed the capabilities of Moonshot AI's Kimi K3 model on social media, claiming his team found it underperformed leading US frontier models on key benchmarks.