The Supreme People's Court has successively issued the people's courts' litigation rules, mediation rules, and operational rules, as well as opinions on strengthening the judicial application of blockchain and regulating and strengthening the judicial application of artificial intelligence, and its theoretical innovation and practical exploration have attracted global attention. While fully affirming the important and positive role of smart justice, it is also the proper meaning of smart justice to pay attention to and prevent the potential risks posed by digital technology to the realization of judicial fairness, and to integrate risk governance with the implementation of judges' judicial responsibilities.
At present, the smart judicial model of using digital technology tools combined with network, big data, cloud computing, 5G, artificial intelligence, and blockchain to assist the judiciary is developing in depth. The Supreme People's Court has successively issued the people's courts' litigation rules, mediation rules, and operational rules, as well as opinions on strengthening the judicial application of blockchain and regulating and strengthening the judicial application of artificial intelligence, and its theoretical innovation and practical exploration have attracted global attention. This is an important content and embodiment of the modernization of the people's court's adjudication system and adjudication capacity, and it is also a major issue related to the realization of digital justice in the new era. While fully affirming the important and positive role of smart justice, it is also the proper meaning of smart justice to pay attention to and prevent the potential risks posed by digital technology to the realization of judicial fairness, and to integrate risk governance with the implementation of judges' judicial responsibilities.
I. The profound impact of smart justice on the judicial responsibility system of judges
The application scenarios of smart justice mainly include three aspects: first, case management, including the collection, storage and sorting of case-handling information, judicial services and management supervision, etc., which provide data support for data analysis and judicial decision-making required for follow-up case handling, and also provide more convenient services for judges and party litigation; the second is litigation, including remote synchronous litigation and asynchronous litigation on the virtual trial platform, as well as evidence exchange and mediation; The third is judicial decision-making, including the review and identification of blockchain evidence, precedent retrieval, push of similar cases, legal and regulatory inquiries and research, intelligent generation of legal documents, etc., as well as data analysis through historical data and case-related factors, so as to conduct trial results and post-judgment risk assessment, and substantially assist judges in judicial decision-making. As a new judicial model, smart justice can effectively improve the quality and efficiency of justice, promote the unification of adjudication standards, automatically warn and prevent non-standard judicial behaviors, strengthen the efficiency of trial supervision and management, and further highlight the judicial service functions of the people's courts.
At the same time, smart justice increasingly covers many fields and links of trial enforcement work, which will inevitably have a significant and profound impact on the judicial responsibility system of judges. First, the continuous expansion of smart justice to the traditional judicial field has expanded the scope of judges' judicial responsibility. In particular, the judicial application of information technology, the field of litigation, and the judicial application of artificial intelligence have become important and extensive new fields for the judicial responsibility of judges. Second, the increasingly extensive and in-depth application of digital technology, especially artificial intelligence, in smart justice has increased the risk of judicial responsibility of judges. For example, whether the network information technology is correctly used, whether there is a negligence in protecting the litigation rights of the parties in the litigation, whether the review and determination of the blockchain evidence is correct, whether there are errors in the process of intelligent generation of judgment documents, and whether there will be wrong judgments due to technical dependence, etc. Third, digital technology and its rapid development in smart justice have put forward higher requirements for judges to assume judicial responsibility. For example, the risk of leaking state secrets, commercial secrets, and personal privacy in smart justice increases, and judges' responsibility to ensure the security of case data and information intensifies; The unified application of law has increased judges' duty of care to "make the same judgment in similar cases", and digital technology has brought convenience to judges in handling cases, but also brought higher responsibility requirements. With the increasing application of digital technology to more judicial fields, judges' judicial behavior is supervised by traces throughout the judicial process, enhancing the traceability of judicial responsibility. Fourth, the human-machine mixed work and decision-making mechanism of smart justice have increased the difficulty of identifying and identifying judges' responsibility for illegal trials. There are uncertainties in digital technology itself, such as data is constantly changing, technology is constantly developing, artificial intelligence is constantly learning, algorithms are diverse, etc., all of which will affect the certainty and correctness of smart justice. Smart justice also puts forward new issues and new requirements for the judicial accountability of judges, such as whether it is necessary to establish a special smart judicial supervision and management mechanism; How to analyze and determine judges' subjective faults and objective illegal trial behaviors in the current judicial and technical fault responsibilities of judges in smart justice is no longer a hypothetical problem.
II. Clarify the subject and scope of judicial responsibility in smart justice
Undoubtedly, the intelligent justice of man-machine confusion objectively poses a certain problem to the implementation of the principle of "let the adjudicator make the judgment and the adjudicator is responsible", but it should not cause contradictions and conflicts in the implementation of the judicial responsibility system, but it is necessary to further clarify the subject and scope of judicial responsibility.
The first is to further clarify the main responsibility of judges in smart justice. At present, the legal characteristics of smart judicial application products are still weaker than their technical characteristics, the authenticity and correctness of the source data information used for learning and training are not completely reliable, the transparency and uniformity of algorithms are still insufficient, and the lack of sufficient judicial practice application foundation and the latest judicial knowledge graph supply, the persuasiveness of its "explainability" is not strong enough. Therefore, although many of the work and decision-making of smart justice are made by mixing man and machine, all the technical means of smart justice are still only auxiliary trials rather than leading trials, let alone replace judges' justice, and their formation is still technical achievements, which can only be used as a reference for trial work or trial supervision and management, and the organization of trial activities, the performance of judicial duties, the exercise of judicial power, and the decision of adjudication results must ensure that the trial organization and the judges therein make independent judgments and decisions in accordance with the law. It must be insisted that judges are always the leaders of judicial decision-making. Therefore, even in the judicial application of generative AI, the subject of judicial responsibility can still only be the judge who conducts the trial, so the proposition that the so-called AI technology will share judicial power and that "technology" should be allowed to share the judicial responsibility of judges is not valid.
The second is to further clarify the scope of judges' responsibilities in smart justice. First, strictly implement the responsibility of judges to follow litigation procedures. At present, the important value function of litigation is to effectively dissolve the inherent limitations of time and space on litigation, and asynchronous litigation expands this function. However, as a form of intelligent justice that focuses on instrumental functions, there are drawbacks of utilitarian use, coupled with the limitations and constraints of the litigation form itself, compared with the traditional litigation order of trial centrism, there may be many troubles and problems: such as how to comprehensively, fully and specifically protect the litigation rights of the parties; how to implement the principle of direct oral trial; how to ensure that judges have sufficient personal experience and sense of presence required for observation and judgment, so as to ensure that judges can effectively form and strengthen their free will; how to avoid fragmentation of the trial process; How to avoid accidental disruption of the order of the trial, technical intrusion to destroy information security, leakage of the content of the trial or random dissemination, etc. To this end, the application of smart justice must still adhere to the people-oriented ethics of judicial responsibility, and the most basic thing is to comply with the judicial principles of equal litigation, fairness and justice, transparency and credibility, pay more attention to the priority protection of the parties' procedural rights, and ensure the realization of procedural fairness. It is necessary to face up to the limitations of litigation, pay attention to eliminating the negative impact of technical power and technical omissions on litigation, such as the establishment of a non-intelligent disposal of case information involving state secrets, commercial secrets and the privacy of the parties, and handle the relationship between ensuring the security of case information and ensuring judicial openness, so as to maximize the realization of the trial function. Where a party raises objections or defenses to procedural issues in litigation, it shall be explained and its due rights protected.
Second, strictly implement the responsibility of judges to conduct substantive trials. The judicial adjudication process is to apply legal norms to the facts of the case and form a judgment conclusion through reasoning, while artificial intelligence generates results based on correlation rather than causality, big data computing is an inductive method in the final analysis, and judges use a deductive method in adjudication. Therefore, although generative artificial intelligence can give clear references to the analysis of the rights and wrongs of cases and the determination of adjudication results, and this is also based on the in-depth application of big data such as judicial cases and legal research results, it is still qualitatively different from the "syllogism" logical reasoning model of judges' judgments, and qualitatively different from judicial existence in the real sense. Therefore, even the judicial application of generative AI cannot completely replace judges to automatically complete the judicial tasks of comprehensively reviewing and judging evidence, objectively and accurately determining facts, in-depth investigation of applicable laws, and making judgments fairly in accordance with the law. In order to ensure the fairness of the final judgment, judges must standardize the application of technology from the substantive requirements of the trial, and must bear the judicial responsibility for the substantive performance of the right to judge and adjudicate, and even the judge must fulfill the responsibility of reviewing and verifying the laws and regulations and similar cases provided by artificial intelligence, so as to prevent being misled by erroneous views and false content, otherwise it may cause dereliction of duty. In practice, some judges follow the nature of seeking advantages and avoiding disadvantages, in order to achieve "the same case and the same judgment" and thus realize the avoidance of the risk of judicial responsibility, they may form an excessive dependence on smart judicial technology, and thus produce a kind of inertia in thinking and work, and may also form the objective phenomenon of algorithms interfering in the judiciary, as well as the mechanical judicial consequences caused by dependence on technology and blind obedience, which will affect the realization of individual justice. Generative AI is embedded in smart justice, and can also assist judges in writing judgment documents, including being able to reason in judgment documents through learning and training, but judges must still prevent and resolve the ethical and moral risks that may arise therein, adhere to justice for the people, adhere to the core values of socialism, and strive to achieve the organic unity of political effects, social effects, and legal effects, which cannot be replaced and automatically generated by artificial intelligence, even if some generative artificial intelligence is input into moral tendencies, Its legitimacy and justice still need to be reviewed and measured by judges in accordance with the law, which requires the implementation of judges' substantive trial responsibilities in smart justice.
III. Establish supporting mechanisms for implementing judges' responsibilities in smart justice
Therefore, it is necessary to establish a supporting mechanism for the implementation of judges' judicial responsibilities in smart justice to promote the realization of a higher level of fairness and efficiency.
The first is to establish a smart judicial education and training mechanism. Increase the intensity of training, so that judges master the necessary interdisciplinary knowledge and skills, continuously improve the level of understanding and application of relevant smart judicial technology principles, including the recognition and prevention of their limitations, avoid judicial responsibility caused by operational errors, and strive to achieve the organic integration of technology and law, technology and ethics.
The second is to establish a mechanism for the supervision and management of smart judicial trials. In the construction of smart courts, establish a mechanism for the pre-evaluation of big data and algorithmic rules and a mechanism for post-decision-making review and inspection, to ensure that intelligent judicial products and services are universally inclusive and equal in opportunity for all types of users, and to prevent factors such as technological intervention, data or model bias, and algorithmic bias from affecting the fairness of the judicial process or adjudication results. Establish mechanisms for ethical review of the judicial application of AI to ensure that public order and good customs are not violated, and the public interest and public order are not harmed. Establish mechanisms for the supervision and management of smart justice and mechanisms for early warning and correction of deviations in case trials, establish mechanisms for higher courts to review the results of the application of smart justice in second-instance trials and retrials, and where parties raise objections or defenses to the technology, process, and results of the application of smart justice, they should be prudently reviewed and correctly explained, and errors made by lower courts should be promptly corrected.
The third is to clarify the chain of responsibility for smart justice. It is necessary not only to prevent judges from shirking and transferring their judicial responsibilities to abstract media under the pretext of technical problems, and to strengthen the technical security care obligations that judges should bear in addition to their judicial decision-making responsibilities, but also to study and implement the responsibilities of developers, designers, and maintainers of smart judicial technology systems in the assessment, review, inspection and maintenance, and ensuring safe operation of smart judicial technology systems. However, it must be made clear that this division of responsibility is aimed at pursuing the respective responsibilities of the relevant parties after the damage to the parties has been caused, and the judicial responsibility of the judge must not be assigned to other subjects for sharing. In determining the responsibility of judges in smart justice, it is still necessary to adhere to the principle of imputing responsibility that is subjective and objective. For the erroneous consequences of smart justice, if the judge uses it improperly, fails to fulfill the duty of due diligence, or violates the procedural obligations and substantive trial obligations that should be followed, the judge shall be held responsible; If the judge is not otherwise at fault, he or she should not be held responsible.
*: People's Court Daily.
Author: Kang Tianjun.