A US artificial intelligence company has accused the Chinese e-commerce giant Alibaba of orchestrating what it called the largest illicit extraction of its AI model’s capabilities.
In a letter sent to two members of the US Congress, San Francisco-based Anthropic said operators linked to Alibaba carried out almost 29 million exchanges with its Claude AI model using thousands of fraudulent accounts. The company described it as “the largest campaign to illicitly extract Claude’s capabilities”.
“Anthropic accuses Alibaba of carrying out the largest known campaign to illicitly extract its Claude AI model's capabilities.”
The attack, known as a “distillation attack”, involves extracting answers from a stronger AI model to train a weaker one. Anthropic said Alibaba-linked operators targeted Claude’s most valuable abilities, including handling longer and more complex tasks and its approach to decision-making.
“Distillation attacks turn hundreds of billions of dollars in American investment and research and development into a massive subsidy for our geopolitical competitors,” Anthropic said in the letter, dated 10 June and addressed to US Senators Tim Scott and Elizabeth Warren.
The company urged Congress to penalise companies behind such attacks and to ramp up measures to prevent US technology from being stolen.
The allegations come amid growing tension between the US and China over AI dominance. Anthropic cited US Department of Defense claims that Alibaba and several major firms, including car maker BYD and tech company Baidu, are tied to the Chinese military. The companies have denied any such allegations, while Alibaba this week sued the US government in a bid to get its name removed from the Pentagon blacklist.
US developers have previously accused Chinese competitors of using distillation attacks. OpenAI has also accused Chinese groups of employing the same practice.
Anthropic is a leading AI developer, alongside ChatGPT-maker OpenAI, and is gearing up for a blockbuster stock market debut that could make it one of the most valuable companies in the world. However, some of its more advanced models, such as Mythos, have raised cybersecurity concerns over their ability to target weaknesses in computer systems.
The BBC has contacted Alibaba for comment.