Anthropic has sent a letter to the US Congress accusing Alibaba of using a large-scale "distillation attack" to extract intellectual property from Claude, according to Forbes. The allegation centers on claims that Alibaba registered about 25,000 fake accounts and used them over a three-month period to generate 28.8 million prompt-response interactions. Anthropic said those interactions pulled roughly 57.6 billion tokens of digital knowledge from its large language model.
The reported extraction data was allegedly used to train Alibaba’s own large language model, with a focus on advanced capabilities including agentic reasoning. The operation was also described as intentionally structured to resemble normal user behavior in order to avoid detection. The claims were presented in Anthropic’s communication to lawmakers and were cited in coverage by Forbes. No additional response from Alibaba was included in the source material.
Anthropic has sent a letter to the US Congress accusing Chinese tech giant Alibaba of carrying out a large-scale "distillation attack" to illegally extract AI intellectual property from its Claude large language model, according to Forbes.
Earlier reports said Alibaba allegedly registered about 25,000 fake accounts and used them to conduct 28.8 million prompt-response interactions over three months, extracting roughly 57.6 billion tokens of digital knowledge in the process. The data was suspected of being used to train Alibaba’s own large language model, with particular attention on advanced AI capabilities including agentic reasoning.
The activity was also described as deliberately designed to mimic normal user behavior to evade detection.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
