In a letter to the U.S. Senate Committee on Banking, Housing, and Urban Affairs, Anthropic claimed that Alibaba executed the largest known distillation attack against it, involving 28.8 million exchanges through approximately 25,000 fraudulent accounts over a short period.
Distillation, a method used to create a less capable AI model from a more advanced one, poses significant risks to intellectual property and competitive advantage in the AI industry.
Anthropic emphasized the need for coordinated action between government and industry to combat such threats, especially following a recent memorandum from the White House aimed at helping AI companies address industrial-scale distillation.
The situation is further complicated for Anthropic as it faces an export control directive from the Trump administration, which has restricted access to its latest AI models for foreign nationals, citing national security concerns. This ongoing conflict highlights the increasing tensions in the AI sector and the importance of safeguarding technological advancements against foreign exploitation