US artificial intelligence company Anthropic has accused Chinese technology and e-commerce giant Alibaba of conducting what it describes as the largest known campaign to illicitly extract capabilities from its Claude AI model. In a letter dated 10 June and addressed to US Senators Tim Scott and Elizabeth Warren — the chair and ranking member of the Senate Banking Committee — Anthropic said operators affiliated with Alibaba and Alibaba Qwen, the company's AI research lab, carried out nearly 28.8 million exchanges with Claude through almost 25,000 fraudulent accounts between 22 April and 5 June 2026. Alibaba has not responded publicly to the allegations.

The technique at the heart of the accusation is known as a "distillation attack" — a method in which a weaker AI model is trained on the outputs of a more powerful one, effectively absorbing its capabilities at a fraction of the original development cost. Anthropic said the campaign specifically targeted its most advanced features, including those related to complex, long-form tasks and decision-making, with the goal of accelerating access to capabilities found in its cutting-edge Mythos Preview model. The company warned that such attacks, carried out on an "industrial scale", convert "hundreds of billions of dollars in American investment and research and development into a massive subsidy for geopolitical competitors."

The accusations form part of a broader and escalating pattern. Anthropic disclosed in February 2026 that three other Chinese AI firms — DeepSeek, Moonshot AI, and MiniMax — had conducted similar extraction campaigns, involving a combined total of more than 16 million exchanges. DeepSeek, a startup whose low-cost AI model caused significant disruption in global technology markets in early 2025, was the subject of the smallest of those campaigns. Anthropic described the attacks as growing in "intensity and sophistication" and called for coordinated action from industry, policymakers, and the wider AI community.

The letter was sent ahead of a Senate Banking Committee hearing on AI and urges Congress to penalise companies behind such attacks and strengthen protections for US technology. The broader political context is significant: Alibaba was added to the US Pentagon's list of companies alleged to have ties to the Chinese military earlier this month — a designation it is actively contesting in court. The White House in April accused China of stealing US AI intellectual property on an industrial scale, though the Commerce Department has so far refrained from placing DeepSeek on a trade blacklist, reportedly to avoid further straining relations with Beijing.

Complicating the picture further, two days after Anthropic sent its letter, the US Commerce Department imposed export restrictions on Anthropic's own Mythos and Fable models, citing concerns they could be accessed by military intelligence users in China and other countries of concern. The move forced Anthropic to disable global access to those models. The sequence of events underscores how AI technology has become a central arena of US-China strategic competition — with American companies simultaneously lobbying for stronger protections against Chinese rivals while themselves becoming subject to tightening government controls.