Anthropic slams Alibaba for 'illicitly extracting' AI capabilities

Thu Jun 25 2026 09:22:26 GMT+0000 (Coordinated Universal Time)

AI firm Anthropic accuses Alibaba of a large‑scale data siphoning operation targeting its Claude model, calling the move ‘brazen’ and demanding congressional penalties.

AI company Anthropic has accused the Chinese tech giant Alibaba of illegally harvesting the capabilities of its Claude model through a massive distillation attack, claiming the operation involved nearly 29 million exchanges via thousands of fraudulent accounts. The letter, sent to US Senators Tim Scott and Elizabeth Warren, urges federal action to curb such thefts and warns of broader security risks. Alibaba has denied the allegations and recently sued the US government to remove its name from the Pentagon blacklist. The incident underscores heightened tensions over AI IP and raises concerns about cross‑border technology espionage.

Anthropic slams Alibaba for 'illicitly extracting' AI capabilities

Image: Anthropic CEO Dario Amodei, Bloomberg via Getty Images

US artificial intelligence (AI) giant Anthropic has accused Chinese e‑commerce and technology firm Alibaba of “brazenly” and “illicitly” extracting its Claude AI model's capabilities.

In a letter sent to two members of the US Congress, the San Francisco‑based company said operators linked to Alibaba carried out almost 29 million exchanges with Claude using thousands of fraudulent accounts in what it called the largest extraction campaign of its kind.

Anthropic urged Congress to penalise the companies behind attacks like this and to ramp up measures to prevent US tech from being stolen.

The company’s letter, dated 10 June and addressed to Senators Tim Scott and Elizabeth Warren, claims Alibaba performed the largest campaign to illicitly extract Claude's capabilities.

According to Anthropic, the campaign was carried out through what are known as “distillation attacks”, in which answers from a stronger AI model are extracted to train a weaker one.

Alibaba‑linked operators targeted Claude's most valuable capabilities, such as its ability to tackle longer, more complex tasks and its approach to decision‑making.

These attacks are performed on an “industrial scale” to enable Chinese companies to harvest and repurpose US AI capabilities as their own.

The letter also cites other alleged attacks that Anthropic says pose a threat to the US military.

“Distillation attacks turn hundreds of billions of dollars in American investment and research and development into a massive subsidy for our geopolitical competitors,” the company said.

Anthropic pointed to US Department of Defense claims that Alibaba, BYD and Baidu are tied to the Chinese military.

The companies have denied any such allegations, while Alibaba recently sued the US government to have its name removed from the Pentagon blacklist.

US developers have previously accused Chinese competitors of taking a similar route to train their models to rival American AI at a fraction of the cost.

OpenAI has also accused Chinese groups of using the same practice.

Anthropic, a leading AI developer, is preparing for a blockbuster stock market debut that could make it one of the most valuable companies in the world.

Some of Anthropic's advanced models, such as Mythos, have raised cybersecurity concerns over their ability to target weaknesses in computer systems.