算法嵌入政府环境治理的正向效应、潜在风险与规制路径OA北大核心CSSCICSTPCD

Compared with the traditional government environmental governance model,the government's use of algorithms to carry out environmental governance has obvious positive effects:it can improve the accuracy,predictability,and efficiency of environmental gov-ernance.However,embedding algorithms into government environmental governance can easily lead to risks in three dimensions:eth-ics,tools,and the rule of law.The ethical dimension faces the risk of value imbalance;the tool dimension includes the risk of insuffi-cient environmental data structure and insufficient transparency of environmental algorithm models;and the legal dimension includes the risk of abuse of power and uneven distribution of responsibilities by those who control the environmental algorithm models.There-fore,it is necessary to address the risks induced by embedding algorithms in government environmental governance and to build a sys-tematic legal regulation system,thereby fully leveraging the advantages of algorithmic automated decision-making in environmental gov-ernance.In the selection of regulatory paths,the potential risks of algorithms embedded in government environmental governance should be systematically regulated based on the guidance of basic macro principles and the design of specific regulatory plans.At a macro level,algorithms embedded in government environmental governance should follow the basic principles of legal governance,fair-ness and justice,floating proportions,and due process.In terms of specific regulatory plans,it is necessary to establish a support struc-ture for algorithms embedded in government environmental governance to take advantage of these algorithms.It is also necessary to make appropriate rule changes based on the loopholes of the algorithms themselves,using the'supportive structures and adaptive flow phase balance'method to improve the feasibility of specific regulatory plans.The structure of the specific regulatory plan should adhere to the perspective of whole-process governance and achieve the unity of prevention beforehand,supervision during the incident,and ac-countability after the incident.The pre-prevention part requires that algorithms embedded in government environmental governance ef-fectively meet ethical requirements,and a complete environmental data approval,feedback,and protection mechanism,along with an explanation mechanism for environmental algorithm models,should be established.The interim supervision part should establish de-partments and social organizations where algorithms are embedded in government environmental governance as well as a public supervi-sion system.The ex post facto attribution part needs to embed the algorithm into the responsibility allocation system of government envi-ronmental governance.