Typical Court Decisions in the Field of Copyright Infringement by Generative Intelligence: the Hangzhou Internet Court Practice

Objective: to study the approach of the Chinese regulator to model court decisions in cases of copyright violations committed using generative artificial intelligence technologies, based on the practice of the Hangzhou Internet Court.

Methods: the research used a set of complementary methods of scientific cognition. Analysis helps to thoroughly study the factual circumstances of two model cases solved by the Hangzhou Internet Court in 2024 and 2025. Synthesis and generalization made it possible to identify the system-forming factors that guided the Chinese court in classifying offenses and determining the guilt of generative artificial intelligence 

service providers. The comparative-legal approach was used to compare two opposite court decisions and to establish the sets of circumstances that led to a fundamental difference in the results of the proceedings. 

Results: the Hangzhou Internet Court used by a multifactorial assessment model when determining the responsibility of a generative artificial intelligence service provider for copyright infringement. The key criteria were the nature and specifics of the services provided; the current level of the applied artificial intelligence technologies development; the copyright infringement evident for a reasonable observer; and the scale and nature of the potential consequences of the offense. An analysis of two typical cases with diametrically opposite outcomes demonstrates that the described approach ensures flexibility in legal regulation: responsibility is not automatically assigned to the language model developer, but is differentiated depending on the specific set of factual circumstances.

Scientific novelty: it is due to referring to the relevant material, practically undeveloped in the Russian legal doctrine – the judicial decisions made by the Hangzhou Internet Court in 2024 and 2025, which received the official status of model precedents in the People’s Republic of China. The study allows introducing the practice of Chinese Internet courts into scientific circulation as an independent regulatory model worthy of comparative-legal study.

Practical significance: the factors for determining guilt and degree of responsibility formulated by the Hangzhou Internet Court represent readymade regulatory tools applicable in improving the mechanisms of legal regulation in the field of artificial intelligence technologies. The study results can be used by legislators and law enforcement officials to develop and interpret the norms governing the activities of generative artificial intelligence platforms, as well as to shape uniform judicial practice in similar categories of cases.