Legal Regulation Of Algorithmic Decision In Automated Administration

In the era of big data,algorithmic decision-making in automated administration is widely used,relying on its own technical advantages to greatly improve the efficiency of administrative management activities and reduce the work pressure of administrative personnel.At the same time,for the disadvantages of algorithmic decision-making technology and the existence of many subjective and objective reasons in the social environment,the legal risks of algorithmic decision-making in automated administrative applications are gradually being exposed,threatening the protection of citizens’ legitimate rights and interests.At present,preventing and regulating the legal risks of the application of algorithmic decision-making in automated administration has become the primary task of balancing the development of science and technology and protecting the legitimate rights and interests of citizens.First of all,the connotation definitions of algorithmic decision-making and automated administration are basically sorted out,and according to the type of algorithmic decision.Divideing the application status of algorithmic decision-making into two application categories,reactive algorithmic decision-making application and predictive algorithmic decision-making application,and a brief introduction to typical examples of each category.Secondly,analyzing the application risks of algorithmic decision-making in automated administration.These risks include: reducing the autonomy of administrative subjects,weakening the privacy protection of administrative counterparts,conflicting with the principles of administrative due process,administrative openness and administrative fairness,and focusing on the subjective and objective causes of the risk.Afterwards,inspecting the legal and regulatory experience of algorithmic decision-making in the United States,the European Union and Germany,and providing experience reference for our country to prevent and avoid the application risk of algorithmic decision-making in automated administration.Finally,on the basis of risk analysis,cause consideration and foreign experience investigation,expanding the prospect of algorithmic decision-making regulation in our country’s automated administration,and five regulatory measures are proposed.First,in terms of subjective consciousness,administrative agencies should be more active and clear algorithmic decision-making tools.Secondly,in the application of algorithmic decision-making,it is necessary to limit the application scenarios of algorithmic decision-making in automated administration.Thirdly,in terms of risk assessment,it is necessary to use the principle of proportionality to conduct a preliminary review of the decision-making system,and to learn from the German personal data protection consultant system to strengthen data protection and supervision.Fourthly,in terms of procedures,new developments and content of due process principles in traditional administrative law should be given to ensure that the principle of due process can effectively standardize the procedural issues of algorithmic decision-making assisting administrative decision-making.Fifth,in terms of responsibility,the responsible party should be clarified,the principle of responsibility should be confirmed,and the responsibility consciousness of the designers and developers of the decision-making system should be strengthened.