- Anthropic has petitioned Congress to strengthen protections against AI model distillation, citing a massive alleged campaign by Alibaba-affiliated operators.
- The company claims operators linked to Alibaba and its Qwen AI lab generated over 28.8 million exchanges with its Claude chatbot using nearly 25,000 fraudulent accounts.
- Anthropic argues this large-scale, unauthorized model extraction threatens national security and inverts the economic logic underpinning U.S. AI leadership.
In a June 10 letter to Senators Tim Scott and Elizabeth Warren, AI company Anthropic alleged that operators affiliated with Alibaba executed the largest known AI model distillation attack against its Claude chatbot from late April to early June. The company framed this distillation attack as a national security threat that could narrow America’s technological lead. Consequently, Anthropic urged lawmakers to expand intelligence sharing, strengthen export controls, and penalize firms responsible for large-scale model extraction.
The campaign reportedly targeted Claude’s advanced capabilities in agentic reasoning and software engineering. Meanwhile, the letter’s timing aligns with increased Washington efforts to protect U.S. AI leadership following a recent presidential executive order on Cybersecurity. However, the debate over distillation’s legitimacy remains complicated, as the technique is an established industry practice according to reports.
Anthropic contends unauthorized extraction violates its terms of service, while observers have previously criticized similar allegations. The company stated, “We believe combating the threat of illicit distillation requires coordinated action between government and industry.” This incident follows Anthropic’s February claims that other Chinese AI developers generated millions of Claude exchanges using fraudulent accounts.
✅ Follow BITNEWSBOT on Telegram, Facebook, LinkedIn, X.com, and Google News for instant updates.
