White House science advisor Michael Kratsios accused Moonshot, the Chinese company behind Kimi K3, of copying Anthropic’s Fable LLM. However, experts doubt distillation played a significant role in Kimi K3's advanced capabilities.

‘Large-scale, covert industrial distillation aimed at stealing proprietary U.S. technology and undermining American research is unacceptable,’ Kratsios stated, amid discussions about banning Chinese models that have sparked controversy within the AI sector.

The claim was echoed by Treasury Secretary Scott Bessent, who noted finding watermarks of U.S. large language models on many Chinese ones. However, experts believe distillation alone cannot explain Kimi K3's performance and suggest more complex techniques like reinforcement learning might be involved.

Brad Hansen from the Laude Institute said: ‘It’s just not feasible to get this model quickly through distillation given Fable has only been available since July. The process is too time-consuming and resource-intensive.’

The debate highlights a broader concern about data security and the global race in AI development, where both copying and innovation are at play.