U.S. Treasury Warns of Sanctions Against Chinese AI Firm Moonshot Amid Distillation Allegations

U.S. Treasury Secretary Scott Besent issued a warning on July 22 about the potential imposition of sanctions on Chinese AI company Moonshot. This followed allegations from White House science and technology policy officials, accusing Moonshot of distilling Anthropic’s Fable model.

Besent posted on his X account: “Open-source does not equate to unrestricted hunting of U.S. intellectual property. If Chinese companies secretly conduct large-scale distillation attacks and cross the line of IP theft, sanctions and designation on the Entity List will be considered.”

This statement came just hours after Michael Kratsios, Director of the White House Office of Science and Technology Policy, publicly criticized Moonshot for conducting large-scale distillation on U.S. AI models.

Loopholes in Semiconductor Regulations

Kratsios alleged that Moonshot had acquired servers equipped with Nvidia’s GB300 chips and accessed them within Thailand. The GB300 servers belong to Nvidia’s Blackwell generation, whose export to China is prohibited.

These allegations suggest potential loopholes in U.S. semiconductor export regulations. If Chinese companies are accessing restricted products via third countries, it raises critical questions about the effectiveness of these controls.

Earlier this week, Secretary Besent had already declared that the U.S. government would impose sanctions if evidence of intellectual property infringement by Chinese open-source models was discovered during their investigations.

Technical Issues Surrounding Model Distillation

Model distillation is a common AI technique where smaller models are trained using the outputs of larger models. While it can improve computational efficiency and inference speed, it also poses risks of intellectual property infringement.

The core issue lies in whether Moonshot’s recently released open-weight model “Kimi K3” was distilled from Anthropic’s Fable model. Kimi K3 has demonstrated advanced performance, raising questions about the business models of major U.S. AI labs.

However, some experts remain skeptical about the possibility that Kimi K3 was solely developed by distilling the Fable model, which was only released on July 1.

The Expanding Policy Debate

This controversy further escalates the ongoing debate in Washington over the influx of Chinese open-weight models. Dean Ball, a former White House AI advisor and current Strategic Futures Director at OpenAI, has called for the U.S. to restrict or ban the use of Chinese open-weight models.

Ball argues this is essential to maintain U.S. technological superiority and mitigate risks to national security.

TechCrunch has reached out to both Moonshot and the U.S. Treasury for comments. As of now, no formal responses have been received from either party.

The Clash Between Distillation and Trade

Paradigms

This incident highlights a fundamental paradigm clash in AI development. The U.S. emphasizes safety and transparency, with major labs investing heavily in creating state-of-the-art closed models. In contrast, Chinese companies are advancing open-weight strategies, employing techniques like model distillation to catch up efficiently.

This trade-off has evolved beyond a technical debate into a geopolitical confrontation. By hinting at sanctions, the U.S. government appears intent on deterring China’s actions that threaten America’s AI dominance.

Moonshot’s Kimi K3, with its advanced capabilities, has cast doubt on the validity of the expensive investment model followed by U.S. AI labs. If similar performance can be achieved through distillation alone, it undermines the premise that cutting-edge AI development requires billions of dollars in investment, as advocated by OpenAI and Anthropic.

Editorial Opinion

In the short term, this incident could ignite new tensions in U.S.-China AI relations. Should the Treasury move forward with sanctions, Chinese AI firms may be forced to find even more circuitous routes to acquire Nvidia GPUs. This could lead to greater scrutiny and transparency in global semiconductor supply chains as efforts to unearth third-country smuggling networks intensify.

In the long term, this controversy will likely accelerate the establishment of international rules addressing the legality and ethics of model distillation. The inherent conflict between the open-source ethos of AI and national sovereignty seems unavoidable, with each country needing to strike a balance between protecting its AI industry and fostering technological exchange. Export controls alone may prove insufficient, necessitating a new framework tailored to decentralized AI development ecosystems.

From our editorial standpoint, framing this issue as a binary choice between distillation and independent development misses the mark. If Kimi K3 was developed solely through distillation, it could spell the downfall of the business models employed by major U.S. AI labs. Conversely, if it stems from proprietary innovation, it underscores the limitations of America’s closed system approach. Verifying the truth is imperative, as either outcome will have significant ramifications for the industry’s structure.

References

• “Treasury threatens sanctions after White House claims Moonshot distilled Anthropic’s Fable”, by Rebecca Bellan — TechCrunch AI, 2026-07-22T20:49:03.000Z (ARR)
• Source URL: https://techcrunch.com/2026/07/22/treasury-threatens-sanctions-after-white-house-claims-moonshot-distilled-anthropics-fable/