【摘要】 当前，人工智能辅助司法裁判展现出诸多优势，但同时在审判实践中引发了效率与公正的内在张力、法官主体性消隐危机、当事人诉讼权利侵蚀风险，理论上还面临着人机协作责任划分困境、价值判断机械化、伦理框架缺失等难题。为调和人工智能应用司法领域的“人—机”矛盾，宜划定人工智能司法辅助的正当性边界，包括坚守司法权不可让渡的工具性边界、以当事人权利保障为核心的程序可控性边界，以及坚持司法责任制人类中心主义的责任性边界。制度上应将算法治理融入诉讼法规范治理体系，创新人工智能辅助司法裁判应用场景下的程序权利保障机制，在实现技术与法律深度融合的基础上建构制度体系。未来，宜采取立法与司法解释双重规范进路，以充分发挥人工智能在司法裁判中的辅助效能。更多还原

【Abstract】 Currently, artificial intelligence-assisted judicial adjudication exhibits numerous advantages. However, it has simultaneously given rise to inherent tensions between efficiency and fairness in judicial practice, posed a crisis of diminishing judicial subjectivity among judges, and potential risks of infringement upon litigants’ rights. Theoretically, it also confronts such challenges as a dilemma in delineating responsibilities in human-machine collaboration,the mechanization of value judgments, and the lack of an ethical framework. To reconcile the "human-machine" contradictions in the application of AI to the judicial domain, it is advisable to delineate legitimacy boundaries for AI judicial assistance. These boundaries encompass adhering to the instrumental boundary of judicial power that cannot be transferred, ensuring the boundary of procedural controllability centered around safeguarding litigants’ rights, and maintaining a human-centric approach in judicial responsibility system. Systematically, algorithmic governance should be integrated into the regulatory framework of procedural law. Innovative mechanisms for safeguarding procedural rights in AI-assisted judicial adjudication scenarios should be established. This endeavor aims to construct a system based on the deep integration of technology and law. In the future, a dual regulatory approach encompassing legislation and judicial interpretations should be adopted to fully harness the auxiliary efficacy of AI in judicial adjudication.更多还原