Should the Common Law System Be Intelligentized?: A Case Study of China’s Same Type Case Reference System

Abstract

China is building a Same Type Case Reference system (STCR) which makes legal precedents binding by creating a national database of judgments and requiring judges to follow prior similar cases. This makes STCR having the same central characteristic, i.e. stare decisis, as the common law. However, instead of upholding democracy and separation of powers like the common law, STCR functions to strengthen supervision over ordinary judges and restrict judicial discretion. Thus, STCR provides an opposite case to show that the stare decisis characteristic of the common law is not intrinsically associated with the democratic function of the common law. After studying STCR’s artificial intelligent (AI) programs, this article argues that currently these machines need human supervision and intervention and cannot be deemed to replace human judiciary. Nonetheless, it is possible to replace humans in some judicial decision-making tasks. The article provides a machine-centered taxonomy, dividing judicial work into readily machine-resolvable (RMR), possibly machine-resolvable (PMR) and impossibly machine-resolvable (IMR) tasks. It argues that substituting machines in place of a human judiciary in RMR and PMR tasks should be encouraged because these cases involve simple and repetitive fact patterns; the substitution occurs only when machines provide suggestions constantly accepted by experienced human judges; and machine judges can follow precedents and respond to changes. The specification characteristic of algorithmic logic is not exclusive of legal vagueness, judicial discretion or incremental changes. AI machines may be used to help the judiciary match other political powers in the information era and should not be perceived as a threat to the human judiciary. The law should facilitate the judiciary’s participation in the process of developing AI machines and should be creative in utilizing AI to resolve legal issues, prevent and correct machine errors, and provide fair remedies.