人工智能辅助司法裁判的伦理风险及治理  

Ethical Risks and Governance of AI-Assisted Judicial Adjudication

作　　者：陈淑婷 Chen Shuting(Kenneth Wang School of Law,Soochow University Suzhou 215000,Jiangsu,China)

机构地区：[1]苏州大学王健法学院

出　　处：《科学．经济．社会》2024年第5期99-113,共15页Science Economy Society

基　　金：国家社会科学基金后期资助一般项目“多数人之债法律问题研究”(22FFXB041);司法部重点项目“《民法典》物权编中的设权合同问题研究”(18SFB1008)。

摘　　要：人工智能司法应用具有缓解“案多人少”矛盾的积极作用,但同时也引发了伦理风险的担忧。在人工智能辅助司法裁判的应用场景,产生的最大伦理风险就是对人的主体性的挑战,人工智能把法官带进算法决策的技术规程之中,消解了法官审判独立的地位,产生司法不公、侵害人的尊严等问题。应从司法裁判的可接受性和融贯性考量,论证人工智能辅助司法裁判的正当性边界。基于目前人工智能的发展阶段,人工智能尚不能替代法官裁判,人工智能只能是辅助司法裁判的工具。面对人工智能司法应用存在的伦理风险,需制定人工智能伦理风险准则,根据技术产生的伦理风险严重程度,采取分类分级规制。明确人工智能提供者的责任,要求人工智能提供者承担解释说明、披露、合规等义务。规范人工智能提供者的行为,对人工智能使用者进行必要的提示和培训,增加问责机制,设置伦理风险预见和评估机制。实现人工智能与智慧司法的有效融合,相互促进。The judicial application of artificial intelligence has a positive effect on alleviating the contradiction of“many cases and few cases”,but it also raises concerns about ethical risks.In the application scenario of artificial intelligence-assisted judicial adjudication,the biggest ethical risk is the challenge to human subjectivity,and artificial intelligence brings judges into the technical regulations of algorithmic decision-making,which dissolves the independent status of judges’adjudication,and produces problems such as judicial injustice and infringement of human dignity.From the perspective of the acceptability and coherence of judicial adjudication,the legitimacy boundary of AI-assisted judicial adjudication should be demonstrated.Based on the current stage of development of artificial intelligence,artificial intelligence cannot replace judges’adjudication,and artificial intelligence can only be a tool to assist judicial adjudication.In the face of the ethical risks existing in the judicial application of AI,it is necessary to formulate guidelines for ethical risks of AI,and adopt classification and hierarchical regulation according to the severity of the ethical risks generated by the technology.Clarify the responsibilities of AI providers,and require AI providers to bear obligations such as explanation,disclosure,and compliance.Regulate the behavior of AI providers,provide necessary tips and training to AI users,increase accountability mechanisms,and set up ethical risk prediction and assessment mechanisms.Realize the effective integration of artificial intelligence and intelligent justice to promote each other.

关 键 词：人工智能伦理 人工智能 司法应用 伦理风险 伦理治理 

分 类 号：D923[政治法律—民商法学]