Research On The Judicial Application Of Artificial Intelligence Algorithms

With the iterative upgrade of artificial intelligence technology,the problem of judicial application of artificial intelligence algorithms has become more and more prominent.The connotation of judicial application of artificial intelligence algorithms should adopt a broad category,which refers to both the use of artificial intelligence algorithms to assist judicial activities and the replacement of human judges by artificial intelligence algorithms to carry out judicial activities autonomously,such as artificial intelligence algorithms can replace judges to carry out judicial decisions,judicial reasoning,judicial decision-making and judicial sentencing.Its theoretical foundation mainly includes Karl’s contextualism,Holmes’ social induction,Hart’s open structure and Hayek’s view of justice,which has laid a solid theoretical foundation for the judicial application of AI algorithms.At present,the judicial application of artificial intelligence algorithm highlights three aspects of the value of the times,it can meet the urgent needs of a large number of cases,but also significantly improve judicial quality and efficiency,and to a certain extent is also conducive to the correction of unjust and false cases.The judicial application of artificial intelligence algorithm is specifically manifested in four aspects,namely,judicial adjudication level,judicial decision-making level,judicial reasoning level,and judicial sentencing level.Among them,the early artificial intelligence algorithm adjudication technology mainly uses the algorithm of closed rules to help judicial personnel to conduct legal search or key information retrieval.However,with the rapid development of machine deep learning algorithm technology,what could only be made by judges in the past can now be partially done by machines independently and make judicial decisions independently.At the level of judicial reasoning,judicial reasoning software is a kind of software based on knowledge mapping,which presents the case-related subjects and concepts in a visual and precise way through artificial intelligence algorithms such as natural language processing and graphical analysis.At the level of judicial sentencing prediction,artificial intelligence algorithms actually become an auxiliary tool for judicial sentencing and provide sentencing reference opinions.Dialectically,there are two sides to everything.As the judicial application of artificial intelligence algorithms become more widespread,the potential problems are also more prominent.Such as artificial intelligence algorithm judges affect judicial autonomy,artificial intelligence algorithm errors can affect judicial injustice,artificial intelligence algorithm reasoning affects substantive justice,artificial intelligence algorithm sentencing affects discretionary power.In order to solve the above problems,it is necessary to establish the auxiliary positioning of AI algorithm adjudication,clarify the auxiliary nature of AI algorithm adjudication,establish the auxiliary AI algorithm adjudication mechanism and the monitoring and evaluation mechanism of AI algorithm adjudication;establish the AI algorithm decision-making transparency mechanism,which requires the realization of data transparency of AI algorithm decision-making and the establishment of the content of AI algorithm decision-making transparency principles;Continuously improve the procedural specifications of AI algorithm reasoning improvement,establish the procedural specifications of data processing for AI algorithm reasoning correction and the procedural specifications of traceability of AI algorithm reasoning process,and improve the online consultation and supervision platform of AI algorithm reasoning process;reduce the influence of AI algorithm sentencing on discretionary power,clarify the boundary of AI algorithm intervention in the judicial field,and build a human-machine collaborative judicial The new model of sentencing,as a whole,promotes the standardization and rule of law development of judicial application of AI algorithms.