Enterprise technology buyers evaluating AI models face a new risk dimension after a White House adviser publicly accused a Chinese AI firm of systematically stealing intellectual property from a US rival. The allegations raise questions about the provenance and training methods of models being deployed in global supply chains and logistics operations.
The Allegation
Michael Kratsios, President Donald Trump's Science and Technology adviser, said on Wednesday via a post on X that the US government has information that Moonshot AI "distilled" capabilities from Anthropic's Fable AI for the development of its K3 model, according to a BBC report. Distillation, Kratsios explained, is when a weaker AI model extracts answers from a stronger one. He also claimed Moonshot gained access to restricted cutting-edge Nvidia servers to train its models.
Moonshot AI is a Beijing-based developer of the Kimi K3 model, which was unveiled last week and drew attention globally. The company has stated its K3 model is able to rival top US technology, according to the BBC. The BBC has contacted Moonshot, Anthropic, the White House, and Nvidia for comment but no responses were reported in the source.
US Government Response
A day before Kratsios' accusation, Treasury Secretary Scott Bessent told Fox Business that the US would examine whether Chinese AI models have stolen capabilities from American rivals. "We've seen a lot of talk about open-source models coming and threatening the large language models in the US," Bessent said. "If we see, especially, that overseas models are stealing from our great companies, we have the ability to sanction them."
The allegations come amid escalating US-China tech tensions and ahead of an expected meeting between President Trump and Chinese President Xi Jinping in September, according to the BBC.
Implications for Enterprise AI Buyers
The incident underscores the necessity for enterprise technology decision-makers to conduct rigorous due diligence on the origins and training methods of AI models they integrate into their operations. For supply chain and logistics firms relying on AI for forecasting, automation, or trade documentation, using a model tainted by IP theft could expose them to compliance risks and potential sanctions.
| Key Entity | Role in Story |
|---|---|
| Moonshot AI | Accused of distilling from Anthropic; developer of K3 model |
| Anthropic | Victim; its Fable AI model was allegedly distilled |
| Michael Kratsios | White House Science and Technology adviser; made the accusation |
| Scott Bessent | Treasury Secretary; warned of possible sanctions |
| Nvidia | Allegedly provided restricted servers to Moonshot |
| Kimi K3 | Moonshot's AI model that gained attention |
The BBC report did not include any response from Moonshot AI or Anthropic to the allegations. As trade tensions rise, enterprise buyers should monitor regulatory developments and consider how third-party AI models are trained before deployment in sensitive logistics or trade finance systems.
Sources:
