Anthropic is urging Congress to strengthen measures against unauthorized AI model distillation by Chinese entities, following allegations that Alibaba-affiliated operators conducted a significant campaign against its Claude chatbot, generating over 28.8 million exchanges through fraudulently created accounts. The company emphasizes that this “distillation attack” poses not only an intellectual property threat but also a national security concern, potentially enhancing China’s military and cyber capabilities while undermining U.S. technological leadership. This appeal comes as President Donald Trump recently signed an executive order to bolster AI cybersecurity, highlighting the administration’s focus on maintaining America’s competitive edge in AI technology amidst rising concerns about illicit practices from foreign competitors.
Alibaba: Alibaba is a Chinese multinational conglomerate listed on the NYSE that operates in e-commerce, cloud services, and AI through its Qwen lab. The news reports that operators affiliated with Alibaba and Qwen conducted what Anthropic described as the largest known AI model distillation campaign targeting Claude. The company maintains US business operations and is held accountable to American investors and regulators.
Anthropic: Anthropic is an AI research company that develops frontier large language models, including the Claude chatbot. In the news, the company is urging Congress to strengthen protections against unauthorized extraction of its model’s advanced capabilities by Chinese rivals. It sent a June 10 letter detailing alleged distillation attacks and calling for expanded intelligence sharing and export controls.
Elon Musk: Elon Musk leads xAI and other ventures in the technology sector. In April, he testified in federal court that xAI had partly used OpenAI models while training Grok, illustrating the established use of distillation techniques across the AI industry.
Tim Scott: Tim Scott is the Chairman of the Senate Banking, Housing, and Urban Affairs Committee. He is a primary recipient of Anthropic’s June 10 letter outlining the alleged distillation attacks and policy recommendations to protect US AI leadership.
Elizabeth Warren: Elizabeth Warren serves as the Ranking Member of the Senate Banking Committee. She received Anthropic’s letter alongside Chairman Scott, which urges stronger export controls, intelligence sharing, and penalties related to large-scale AI model extraction.
Executive Action: President Donald Trump signed an executive order expanding AI-powered cybersecurity initiatives earlier this month after initially delaying the measure over competitiveness concerns with China.
Industry Practice: Distillation remains a contested yet established method in AI development, with companies disputing the boundary between legitimate training and unauthorized extraction of frontier model capabilities.
National Security: Anthropic frames unauthorized distillation of frontier AI models by Chinese labs as a national security issue that could accelerate adversaries’ military and cyber capabilities.