The U.S. Treasury Secretary Scott Bessent has doubled down on his warnings to Chinese AI companies, stating that sanctions remain on the table after a White House official accused Moonshot of improperly distilling Anthropic’s Fable model.
Model distillation is a common AI training technique in which a smaller model learns from the outputs of a larger one. While this process can infringe on intellectual property rights, it’s also widely used as a legitimate optimization method.
“Open source is not open season on American IP,” Bessent posted on X. “When [Chinese] firms conduct covert, industrial-scale distillation attacks that cross the line into IP theft, sanctions and Entity List designations will be on the table.” Earlier this week, Bessent stated that the U.S. government would examine open source models from China for signs of intellectual property theft and impose sanctions if found.
The controversy surrounds Moonshot’s release of K3, an open-weight model, which has raised questions about whether it was developed primarily through distillation from Fable. Fable has only been publicly available since July 1, and some experts dispute the idea that K3 could have been developed primarily through distillation.
The White House’s science and technology policy chief Michael Kratsios accused Moonshot of conducting large-scale distillation against U.S. models, alleging that the firm had acquired Nvidia’s “GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.” This raises questions about whether Moonshot violated U.S. export-control rules.
The GB300 servers are part of Nvidia’s Blackwell generation, which are banned from being sold to Chinese companies. The episode has also intensified a broader debate in Washington over the influx of Chinese open models.
Some experts have argued that the U.S. should restrict or effectively ban the use of Chinese open-weight models to preserve America’s technological advantage and mitigate potential national security risks. TechCrunch has reached out to Moonshot and the Treasury for comment.
### Background on Model Distillation ###
Model distillation is a common AI training technique in which a smaller model learns from the outputs of a larger one. While this process can infringe on intellectual property rights, it’s also widely used as a legitimate optimization method.
“Open source is not open season on American IP,” Bessent posted on X. “When [Chinese] firms conduct covert, industrial-scale distillation attacks that cross the line into IP theft, sanctions and Entity List designations will be on the table.” Earlier this week, Bessent stated that the U.S. government would examine open source models from China for signs of intellectual property theft and impose sanctions if found.
### Expert Reaction ###
Some experts have argued that the U.S. should restrict or effectively ban the use of Chinese open-weight models to preserve America’s technological advantage and mitigate potential national security risks.
### Conclusion ###
The controversy surrounding Moonshot’s release of K3 has raised questions about whether it was developed primarily through distillation from Fable. The White House’s science and technology policy chief Michael Kratsios accused Moonshot of conducting large-scale distillation against U.S. models, alleging that the firm had acquired Nvidia’s “GB300-equipped servers and has accessed GB300s in Thailand, likely to train its AI models.” This raises questions about whether Moonshot violated U.S. export-control rules.
Source: Original article