A Study Of Risk And Regulation Of Algorithmic Decision-Making In Automated Administration

Compared with the traditional administrative act,the governance mode by the government and the way of exercising power have come to the era of digital intelligent algorithmic decision-making.Algorithmic decision-making in automated administration refers to the administrative act in which administrative organs make assisted or autonomous decisions by algorithms through automated devices in order to achieve the purpose of public management or service,under the logic and objectives of algorithmic operation.Algorithmic decision-making in automated administration differs from traditional administrative act in that it is widely applied in various fields and influences administrative act in an embedded manner with the characteristics of concealment,attachment and formal legitimacy.Algorithmic decision-making can be divided into auxiliary algorithmic decision-making and autonomous algorithmic decision-making depending on the degree of contribution to automated administration,and both types of algorithms are widely applied in traffic enforcement automation,environmental protection administrative punishment,credit regulation,public opinion management and other public fields.Under the influence of the aforementioned characteristics,the algorithmic decision-making widely applied in the field of public management and services may weaken the principle of equality,as follows: the screening and classification procedures erode the principle of equality,the reliance on algorithmic decision-making by administrative organs and administrative counterpart leads to the weakening of autonomy,the deviation of algorithmic operation will lead to digital discrimination and the abuse of algorithmic tools will violate the legitimate rights and interests of administrative counterpart.In addition,algorithmic decision-making challenges the administrative authorities’ enforcement methods,and the algorithmic black box hinders administrative disclosure and seriously impacts due process.Finally,algorithmic decision-making may weaken the legitimacy of administrative discretion due to imprecise translation of code and law into each other and overriding manual discretion.In addition to the above-mentioned reasons,the difficulty to identify the responsibility of administrative organs,the poor articulation of the rights and remedies for the administrative counterpart and the unknown standard of judicial review are also the reasons to induce or even intensify the above risks.Therefore,first of all,it is necessary to strengthen the main consciousness and final decision of administrative organs,and clarify the way of responsibility of administrative organs in algorithmic decision-making.Secondly,improve the right to withdraw from the algorithmic decision-making,protect the procedural rights of the administrative counterpart and improve the cohesion between regulation and remedy.Finally,the legal reservation principle and the algorithm list system can be used to clarify the application scenarios of algorithmic decision-making in automated administration,and the legal review standard of technology and other legal review standards can be used to clarify the standard of judicial review.