Chinese AI Model Kimi Stokes Internal Conflict Within Trump’s AI Strategy Circle

President Donald Trump's AI advisors are openly clashing over how to respond to Kimi, a free open-source AI model from Chinese startup Moonshot that matches the capabilities of paid offerings from OpenAI and Anthropic.

The dispute erupted publicly over the weekend when David Sacks, Trump's former AI and crypto czar (who left the role in March), branded Anthropic's models as "lobotomized" and "woke." Emil Michael, a top Pentagon official, responded by calling Dean Ball, a former Trump AI advisor now working for OpenAI, the AI industry's "supreme village idiot."

The underlying tension reflects a genuine policy dilemma. Each time a capable free model from China reaches the market, it undermines the economic case for paying for American alternatives. This matters because investment enthusiasm in US AI companies has been driving an outsized share of broader economic growth, according to observers in Trump's orbit. Anton Leicht, a fellow at the Carnegie Endowment, noted that Chinese models create "a threat for an administration that really doesn't want more economic bad news." The releases have already rattled US stock markets.

The administration's leadership is split on solutions. Michael and Secretary of Defense Pete Hegseth have favored a government oversight approach, supporting a new White House review process that vets AI models' security before release. Ball, representing OpenAI's position, suggested Trump could handle the issue through informal pressure on US companies to avoid Chinese models, framing direct government intervention as a "de facto licensing regime for frontier AI." Michael dismissed this as insufficient, insisting on "the democratic process" rather than what he called a "Deep State scheme."

Sacks, by contrast, has argued that Chinese models succeed partly because they carry fewer restrictions on use (though he glossed over built-in state censorship). He previously criticized top AI companies for wanting "the government to eliminate their open source competition." His view that more open AI benefits society represents the administration's earlier thinking, but that position has been largely sidelined in favor of greater government control.

One unexamined question is how Kimi achieved its capabilities despite China's constrained access to top semiconductors. Export controls have loosened under Trump, who made the controversial decision to allow Nvidia to sell more chips to China in exchange for US government revenue. There are also allegations of chip smuggling. Kimi's creators may have used model distillation, a technique where new models train on outputs from existing ones. OpenAI and Anthropic have long complained about this practice and requested government intervention. The Trump administration obliged in April with efforts to restrict it.

Kimi's emergence marks a shift in how China's AI development is perceived within the Trump administration. The model's quality—nearly matching an Anthropic system the government once deemed so powerful it was briefly shut down for national security concerns—has forced a reckoning among senior officials about what competitive pressure from open-source Chinese AI actually means for US policy and economic interests. The lack of consensus on a response underscores the complexity of the challenge.