Anthropic says Alibaba must be punished for largest Claude cloning attack
Ars Technica AI · Ashley Belanger · 2026-06-25
Anthropic has accused Alibaba and its Qwen AI lab of conducting the largest known Claude capability-cloning campaign, generating over 28.8 million exchanges through 25,000 fraudulent accounts, and has asked US senators to impose penalties.
Appears in
Extraction
Topics: ai-ip-theftmodel-cloningus-china-ai-competitionai-policy
Claims
- Alibaba and Alibaba Qwen allegedly generated more than 28.8 million exchanges with Claude through nearly 25,000 fraudulent accounts between April 22 and June 5, 2026.
- The campaign specifically targeted Claude's most valuable capabilities, including agentic reasoning, software engineering, and long-horizon tasks.
- Anthropic sent a confidential letter to Senators Tim Scott and Elizabeth Warren on June 10, 2026, the day before a Senate hearing on AI and the American Dream.
- Anthropic describes this as the largest campaign to illicitly extract Claude's capabilities it has ever measured.
- The attacks occurred in the context of China racing to match US AI capabilities following Anthropic's Mythos model release and its subsequent restriction from foreign markets.
Key quotes
New, confidential evidence of the largest campaign to illicitly extract Claude's capabilities we have ever measured.
Operators affiliated with Alibaba and Alibaba Qwen...generated 'more than 28.8 million exchanges with Claude through almost 25,000 fraudulent accounts.'
This campaign 'targeted some of Claude's most valuable capabilities, such as agentic reasoning, software engineering, and long-horizon tasks.'