AI-assisted courts, trial efficiency, and judicial effectiveness : empirical evidence from China

Abstract

In recent years, a growing and notable concern within China's judicial institutions has been the evident paradox of “Case Overload and Judge Shortage”. Under the dual pressures of a surge in case volume and increasing dispute complexity, the traditional judicial mode has struggled to meet the demands of the modern era. The rapid advancement of AI technologies offers a crucial solution to these challenges, providing a pathway for judicial modernization. Despite the widespread adoption of AI technologies in courts in China, little is known about whether and how these AI-assisted reforms influence the performance of the justice system. This paper addresses this gap by integrating data from court judgment documents, trial livestreams, and internal records on AI deployment. We focus on two key judicial reforms—the Intelligent Voice Transcription (IVT) Reform and the subsequent “Trial Without Clerk” (TWC) Reform—and employ a difference-in-differences approach to estimate their causal effects. Our findings suggest that AI-assisted reforms, through human-machine collaboration, significantly enhance trial efficiency and judicial effectiveness. These positive effects appear to operate through several mechanisms, including optimizing trial documentation processes, reducing linguistic frictions in court proceedings, and enabling more effective time allocation for judicial personnel. Further analysis reveals that the efficiency gains achieved through these reforms do not come at the cost of increased judicial workload, but rather stem from process optimization and resource reallocation.