Treasury Threatens Sanctions After White House Claims Moonshot Distilled Anthropic’s Fable Model

The recent controversy between the White House and Chinese AI company Moonshot has highlighted a growing concern in Washington: the unauthorized use of advanced AI models by Chinese firms. At the center of this issue is U.S. Treasury Secretary Scott Bessent, who has threatened to impose sanctions on Chinese companies engaging in intellectual property theft.

Bessent’s warnings are part of a broader effort to rein in China’s influence in the global AI landscape. White House Science and Technology Policy Chief Michael Kratsios accused Moonshot of conducting large-scale distillation against U.S. models using Nvidia’s GB300-equipped servers, which are banned from being sold to Chinese companies. This raises questions about whether Moonshot has violated U.S. export-control rules.

Model distillation is a common AI training method in which a smaller model learns from the outputs of a larger one. While legitimate optimization methods, distillation can also infringe on intellectual property rights if used to replicate proprietary models without permission. The gray area surrounding this issue has sparked intense debate among experts and policymakers about what constitutes IP theft in the AI world.

Some argue that Chinese AI models like Kimi K3 cannot be solely attributed to distillation from Fable, given its relatively recent public availability. However, Moonshot’s release of K3 as an open-weight model has challenged the underlying business models of leading U.S. AI labs, forcing them to reevaluate their capital requirements.

The influx of Chinese open models has reignited a long-standing debate in Washington about preserving America’s technological advantage and mitigating potential national security risks. Some advocates, including former White House AI adviser Dean Ball, have suggested restricting or effectively banning the use of Chinese open-weight models. This proposal is not without merit; as the global AI landscape continues to shift, policymakers must reassess their approach to intellectual property protection.

Bessent’s warning shots across the bow of Chinese companies are a necessary step in this process. The U.S. Treasury secretary has made it clear that sanctions will be on the table if Chinese firms continue to engage in IP theft. However, this raises questions about the effectiveness of such measures and whether they will ultimately stem the tide of unauthorized AI model sharing.

The global AI landscape is increasingly complex, with multiple players vying for dominance. As the stakes grow higher, policymakers must develop a more nuanced understanding of intellectual property protection in AI. The debate over distillation and IP theft is merely one chapter in this ongoing saga.