US Treasury Secretary Scott Besent doubled down on his warnings to Chinese AI companies on Wednesday, saying sanctions were still on the table after a White House official accused Moonshot of incorrectly distilling Anthropic’s Fable model.
Typical distillation It is a common AI training technique where the smaller model learns from the output of the larger model. Although this process can infringe intellectual property rights, it is also widely used as a legitimate method of improvement.
“Open source is not open season on American intellectual property.” Bessent is published on X. “when [Chinese] Companies are carrying out industrial-scale secret distillation attacks that cross the line into intellectual property theft, and sanctions and Entity List designations will be on the table.
Earlier this week, Besant said the US government would examine open source models from China for signs of intellectual property theft and impose sanctions if found.
Besant’s latest comments come hours after those of Michael Kratsios, head of science and technology policy at the White House. accused China-based Moonshot to conduct large-scale distillation against US models. He claimed Moonshot took over “GB300-equipped Nvidia servers and accessed GB300s in Thailand, likely to train its own AI models,” raising questions about whether the company had violated US export control rules.
The GB300 servers are part of Nvidia’s Blackwell generation, which is banned from being sold to Chinese companies.
Some experts dispute the idea that Kimi K3 could have been developed primarily through distillation from Fable, which has only been available to the public since July 1. Moonshot launched the K3 last week as an open-weight model, and its advanced capabilities have raised questions about the fundamental business models of leading U.S. AI labs, casting doubt on whether they can continue to justify the massive capital requirements underpinning the frontier AI race.
This incident also intensified the broader debate in Washington about the influx of open Chinese models. Some, including former White House AI advisor and current head of strategic futures at OpenAI, Dean Paul, have argued that the United States should restrict or effectively ban the use of Chinese open-weight models to maintain America’s technological advantage and mitigate potential national security risks.
TechCrunch has reached out to Moonshot and the Treasury Department for comment.
When you buy through links in our articles, we may earn a small commission. This does not affect our editorial independence.
