Jacob Klein, Anthropic's head of threat intelligence, has highlighted a growing threat from foreign adversaries, particularly from Chinese AI labs like Moonshot AI, which he claims are illegally distilling Anthropic's Claude models to create cheaper, competing technologies.
This practice, while sometimes legal, is being conducted through fraudulent means, including the creation of thousands of fake accounts to bypass security measures. Klein noted that this issue is not isolated to China, as similar threats are emerging from countries like Iran, Russia, and North Korea, where access to leading AI technologies is restricted.
The U.S. government has expressed concern over these practices, with the Trump administration labeling such distillation as unacceptable and promising to explore measures to hold foreign actors accountable.
As Anthropic approaches a potential public offering with a valuation nearing $1 trillion, the implications of these threats could impact its market position and the broader AI sector, especially as competition intensifies and concerns about the misuse of AI technology grow