Anthropic alleges moonshot AI misused its models to steal user data

Anthropic PBC has accused Moonshot AI, a leading Chinese artificial intelligence company, of secretly redirecting thousands of user requests to its Claude models and presenting the responses as its own. This alleged practice, according to Anthropic, was aimed at gaining an unfair advantage in the global AI competition. The claim comes as part of a broader context where the U.S. has been urging AI firms to identify and redirect Chinese users to less capable models. The incident highlights ongoing tensions in the AI sector, with companies accused of intellectual property theft and unfair competition. Moonshot AI has not yet responded to these allegations. The situation underscores the complexities of international AI development and the challenges of maintaining data integrity and model security in a rapidly evolving technological landscape. The accusations follow a previous report that detailed how six Chinese AI firms were accused of copying U.S. frontier models. While the current allegations focus on Moonshot's alleged misuse of Anthropic's technology, they add another layer to the ongoing debate about ethical practices and competitive strategies in the global AI race.