摘要
人工智能在刑事司法中的应用已经是不可逆转的大趋势,其正在经历由浅入深、由窄到宽的发展过程。从世界范围来看,刑事司法领域的人工智能应用主要集中在犯罪预测、犯罪侦查和犯罪预防,辅助审前羁押与保释,辅助量刑,辅助假释与矫正等领域。与其他国家和地区对于人工智能的有限应用相比,我国刑事司法领域的人工智能应用在发展规划上呈现无限扩张的态势。对于刑事司法中人工智能的应用,存在着积极主义和消极主义两种不同立场,对于政策制定者而言,前者产生了“鼓励创新”模式,后者产生了“预警原则模式”。人工智能对刑事司法制度带来的冲击和挑战主要在于合法性、公正性、民主性、有效性、伦理性等五个方面的争议。现阶段应当侧重于对刑事司法中人工智能的应用提供方向性指导,在数据治理层面保障数据的全面性、准确性以及合法性,在算法治理层面保障算法的科学性、透明性和公正性,在应用系统治理层面保障应用系统的安全性、有效性和诉讼主体使用应用系统的平等性,并且坚守司法人员在刑事司法中的主导地位,做到人工智能介入有限、运用有度、辅助有力。
The application of artificial intelligence(AI) in criminal justice has become an irreversible trend, and it is undergoing a development expanding its depth and width. From a worldwide perspective,the application of AI in criminal justice mainly focuses on crime prediction, crime investigation and crime prevention, assisting the decisions of pretrial detention and bail, sentencing decision and the decision of parole and correction. Compared with the limited application in other countries and regions, the application of AI in China’s criminal justice system has shown an infinite expanding trend. There are two different positions on the application of AI in criminal justice, i.e. positivism and negativism. For the policymakers,the former leads to the Encouraging Innovation Model and the latter leads to the Early Warning Principle Model. The impact and challenges brought by the application of AI in criminal justice mainly lie in disputes in five aspects: legality, fairness, democracy, effectiveness, and ethics. At this stage, we should focus on providing directive guidance for the application of AI in criminal justice. We should ensure the comprehensiveness, accuracy and legality of data at the level of data governance;ensure the scientificity,transparency and fairness of algorithms at the level of algorithm governance;ensure the security and effectiveness of the application system and the equality of participants of criminal proceeding in using the application system at the level of application system governance, and adhere to the dominant position of judicial officials in criminal justice so as to make AI’s intervention be limited, its use be moderate, and its assistance be powerful.
出处
《上海政法学院学报(法治论丛)》
2022年第6期99-116,共18页
Journal of Shanghai University of Political Science & Law(The Rule of Law Forum)
关键词
刑事司法
人工智能
合法性
公正性
可信赖性
criminal justice
artificial intelligence
legality
fairness
reliability