自动化行政对行政程序的挑战及其制度因应

The Challenge of Automated Administration to Administrative Procedure and Its System Response

在自动化行政中广泛使用的自动化决策系统主要包括基于规则的专家系统和基于机器学习的算法系统,上述两类系统目前已经应用于交警非现场执法、自动化公共信用评价、自动化行政审批等场景中。由于受到数据来源、算法及系统因素的影响,自动化决策系统在公共行政中的使用可能会对告知、陈述意见、信息公开、说明理由等程序保障造成破坏。对此,理论与实践提出了两种解决路径:一种是数据权利的程序化,另一种是行政程序的数字化。不过,上述两种解决方案主要是从普遍适用性的角度出发,未考虑实际执行中的不确定性因素。基于此,应当以现有方法为基础,对自动化行政进行分级,根据程度不同的自动化来匹配相应的行政程序要素,既回应行政程序的现实困境,也发挥自动化决策系统的优势,不断推进数字法治政府建设。

The automated decision-making systems widely used in automated administration mainly include rule-based expert systems and machine learning-based algorithm systems. The above two types of systems have been used in traffic police off-site law enforcement, automated public credit evaluation, automated administrative license, and other scenarios. Due to the influence of data sources, algorithms and system factors, the use of automated decision-making systems in public administration may damage procedural safeguards such as notification, statement of opinion, information disclosure, and giving reasons. In response, theory and practice have proposed two solutions: one is the proceduralization of data rights, and the other is the digitization of administrative procedures. However, the above two solutions are mainly from the perspective of general applicability, without considering the uncertainty factors in practical implementation. Based on this, the existing methods should be used as the basis for grading the automated administration and matching the corresponding procedural elements according to different degrees of automation, both to respond to the real dilemma of administrative procedures and to take advantage of the automated decision-making system.