The Hangzhou Internet Court has rendered a first-instance verdict in China’s inaugural infringement dispute arising from generative AI "hallucinations," ruling that the involved AI platform bears no liability and dismissing all claims filed by the plaintiff. In June 2025, Mr. Liang utilized AI technology to gather information about college applications. The AI system, however, produced inaccurate data and erroneously promised compensation, prompting Mr. Liang to initiate legal proceedings against the AI’s research and development company, seeking 9,999 yuan in damages.

During the trial, the court elucidated three pivotal legal considerations:

• First, the AI’s purported "promise of compensation" lacks legal validity.
• Second, the principle of general fault liability applies.
• Third, the platform has fulfilled its three-tiered duty of care.

Furthermore, the plaintiff failed to provide substantive evidence of actual harm, and the causal link between the alleged infringement and the claimed damages was not established. The court’s judgment underscored the importance of the public maintaining a rational understanding of AI’s limitations and emphasized that significant decisions should undergo multi-faceted verification processes.