Newsletters
The American artificial intelligence company Anthropic has accused Chinese technology giant Alibaba of carrying out its largest known effort to extract the capabilities of its Claude AI models. The incident occurred shortly before U.S. regulatory restrictions compelled Anthropic to disable international access to its newest models.
The allegations are laid out in a June 10 letter sent to the chair and ranking member of the U.S. Senate Banking Committee ahead of a scheduled hearing on artificial intelligence, Reuters reported.
Anthropic claims the campaign ran between April 22 and June 5, 2026, generating more than 28.8 million exchanges with Claude through nearly 25,000 fraudulent accounts.
According to Anthropic, the operation relied on "distillation," a technique that trains a less capable AI model using the outputs of a more advanced one, the report noted. The company alleges the campaign, which it links to individuals affiliated with Alibaba and its AI research unit Qwen, was intended to accelerate China's ability to match its Mythos Preview models.
Anthropic also voiced support for U.S. government efforts to counter AI-related intellectual property theft, including closer cooperation with private-sector developers through threat intelligence sharing.
Earlier this year, Anthropic disclosed similar attempts involving other Chinese AI firms. In a February post, it reported that DeepSeek generated more than 150,000 exchanges with Claude, while Moonshot AI exceeded 3.4 million and MiniMax topped 13 million.
The company warned at the time that these campaigns were becoming increasingly sophisticated and called for coordinated action across the AI industry and government.
Two days after Anthropic sent its letter to lawmakers, the U.S. Commerce Department imposed restrictions on Anthropic's latest Mythos and Fable AI models over concerns they could be used by military intelligence entities in China and other countries of concern.
Anthropic subsequently disabled global access to those models.
The allegations come as Washington ramps up efforts to curb China's access to advanced AI technology after accusing Beijing in April of stealing U.S. AI intellectual property on an industrial scale.
Alibaba, whose Qwen family of large language models has emerged as one of China's leading AI platforms, has accelerated investment in artificial intelligence in recent years to compete with U.S. developers such as Anthropic, OpenAI and Google.
DeepSeek, another prominent Chinese AI startup that gained global attention with its low-cost AI models, has also become a key player in China's push to narrow the gap with U.S. rivals.
Alibaba was added to the Pentagon's list of Chinese military companies this month, a designation it is challenging. Meanwhile, the Commerce Department has held off on placing DeepSeek on a U.S. trade blacklist despite an interagency committee having identified the company as a national security risk.
Set as preferred source
Set as preferred source
1 min read
1 min read
