Research On Legal Protection Of Personal Information In Automated Decision-making

In the era of big data,data has become the core asset,and personal information is the most important part of it.Personal information contributes to government management,transaction cost reduction and the development of new business models.Automated decision-making is one of the main ways to use personal information.Through the collection and analysis of personal information,more optimized transaction pricing and more effective personalized recommendation can be achieved,which has generated huge economic benefits.Baidu,Douyin,Taobao and other large platforms all carry out automated decision-making activities.The circulation and use of personal information not only brings convenience to individuals,but also brings the risk of disclosure and abuse of personal information and other serious harmful consequences.The main problem facing personal information is the harm brought by automated decisions,such as big data killing and concept domination generated by personalized recommendation.In order to give full play to the value of personal information and promote the improvement and development of automatic decision making,we should analyze how to protect personal information in automatic decision making.To discuss the protection of personal information,we should first clarify the general principles of personal information,including the definition of personal information,the interests of personal information and the legal attributes of personal information.In the definition of personal information,when defining personal information beyond ordinary people’s ability to identify but can be identified by powerful information processors,it is necessary to investigate the realistic identification possibility of personal information.By interpreting the identification target in the "identification theory",namely the identity of a natural person,as the identity that can form a stable and continuous contact object,it can replace the function of the "association theory" which is contrary to the personality nature of personal information,so as to solve the problem that important information such as cookie and device identifier cannot be included in the scope of personal information.Personal personality interests in personal information can be clearly summarized into two aspects:information subjects build their own self-image and use personal information to help them develop their own personality in communication.Personal property interests in individual personal information are often very thin and have no protection value and should be denied.The empowerment of personal information is not a factor hindering the circulation of information.Only by establishing the independent personality right status of personal information right,and clearly protecting the two aspects of information subjects’ self-image building and self-use of personal information to help communication,can the development of information circulation and data economy be better assisted.Automatic decision-making refers to the activity of making decisions by automatically analyzing and evaluating an individual’s behavior,interests,or economic,health,or credit status through computer programs.Automatic decision-making has the characteristics of pure computer processing,wide range of collection,persistence of processing behavior and depth of information analysis.The personal information processing behavior in automatic decision can be divided into collection,control and analysis and application of three kinds of processing behavior,and specific personal information processing behavior,such as stealing,forced collection,ineffective protection and abuse of personal information interests.The personal information processing behaviors in these automated decisions may bring risks to the information subject,such as sense of surveillance,downstream criminal threat,data discrimination and concept domination,etc.The protection of personal information should be fully prevented from these problems.The current personal information protection norms in automatic decision-making in China have made detailed provisions in three aspects: informed consent,algorithm interpretation right and refusal right,but they can not meet the needs of information subject.It should be noted that the protection of personal information in automatic decision-making has three particularities: the irrationality of information subject,the regularability of information processing tool and the accuracy of user profile.The information subject has two disadvantages in terms of information and ability in Automatic decision-making.The "rational man hypothesis",which is the basis of the informed consent principle,has collapsed and cannot realize the function of protecting personal information.The introduction of the concept of "context" coincides with the practice of automatic decision making,which can effectively promote the realization of information subject autonomy and eliminate the sense of monitoring.Automatic decisions are made purely by using computer programs.The tool for analyzing and applying personal information is algorithm.Compared with the personal information processing tool in other activities--natural person’s thinking,algorithms without subjectivity can be regulated.At the same time,the algorithm in Automatic decisionmaking is not a tool to assist the information processor to make decision unilaterally,but as a role to optimize the communication between the information subject and the information processor,the algorithm should be regulated.User profile is the basis for personalized recommendation and determines the result of personalized recommendation.The accuracy of user profile makes personalized recommendation deeply affect the concept and self-development of information subjects.Therefore,personalized recommendation is not only an advertisement for information processors,but also a tool for information subjects to understand themselves and develop themselves.When the user profile pursued by the information processor conflicts with that pursued by the information subject,the information subject should not only be given the right to reject,but also be given the right to manage the user profile.The concept of "context" should be introduced into the protection of personal information in Automatic decision-making.The principle of informed consent realizes the purpose of protecting the autonomy of the information subject by ensuring that the information subject understands the scene and decides whether to enter the scene of Automatic decision-making,rather than strictly ensuring that the information subject controls every piece of personal information independently.Algorithmic regulation is an important aspect of personal information protection in Automatic decision-making.It should go beyond the restriction that only when the decision result has a significant impact on personal rights and interests,and give the information subject a more comprehensive algorithmic interpretation right to help the information subject fully understand the scene.In the personalized recommendation,in the process of personal information collection,the information processor should provide the privacy mode,and the personal information generated by the information subject in the privacy mode will not be used to form user profiles.By controlling the range of personal information used for user profile,the information subject can get the recommendation information that really meets its own needs and realize the control of its own development direction.