Abstract:

At present， Artificial Intelligence （AI） is involved in judicial evidence reasoning and has aroused widespread concern. AI is still “top-down” in that its logical reasoning is good， but it cannot completely simulate human cognitive thinking. It operates in an auxiliary position in judicial evidence reasoning where its advantages include massive data screening and positioning， efficient and consistent operation of algorithms， but its include evidence information extraction， evidence language transformation， legal thinking simulation， judicial experience learning， etc. As an auxiliary to the judicial process， AI helps with sorting evidence materials， provides early warnings for evidence information loss， and creates both pre-guidance and post-correction for evidentiary reasoning. At the same time， AI should be principled so that the judiciary can provide intelligent trials； the AI algorithms should be equally available and public； and， there are limited restrictions on machine-aided judicial evidence reasoning. Such an AI auxiliary system must also let AI participate in the evidence reasoning of different types of cases and different stages of litigation in a reasonable manner such that each matter has the proper AI tools. Through the complementary advantages of artificial intelligence and human judges， the fairness and efficiency dividends of judicial work will open a new chapter in judicial trials in the new era.

Key words: artificial intelligence, fact finding, auxiliary evidential reasoning, Thinking simulation, mathematical operation