Anthropic told U.S. lawmakers that it detected what it described as the largest model-extraction campaign it has faced, attributing the activity to operators affiliated with Alibaba and Alibaba’s Qwen lab, according to a June 10 letter reviewed by Reuters.
Anthropic said the activity ran from April 22 to June 5 and generated more than 28.8 million exchanges with its Claude model through nearly 25,000 fraudulent accounts. Reuters said distillation in this context refers to training a less capable model on the outputs of a stronger model.
The letter was addressed to Senate Banking Committee Chair Tim Scott and Ranking Member Elizabeth Warren ahead of a scheduled hearing on artificial intelligence. Anthropic had previously disclosed in February that it identified similar campaigns linked to DeepSeek, Moonshot AI and MiniMax, which it said produced more than 16 million exchanges through about 24,000 fraudulent accounts.
The allegation comes amid wider U.S. scrutiny of unauthorized extraction of AI model capabilities. In April, the White House said it had information indicating foreign entities, principally based in China, were conducting industrial-scale campaigns against U.S. frontier AI systems. Separately, Anthropic said on June 12 that it disabled access to its Fable 5 and Mythos 5 models globally after a U.S. government export-control directive suspended access for foreign nationals, while saying its other models were unaffected.