Legal Regulation Of Personal Credit Scoring Algorithms On Online Platforms

With the rapid development of Internet technology and artificial intelligence technology,a cyberspace society centered on online platforms has been formed,and personal credit scoring has presented a new model with online platforms as the main body to collect massive data from the platforms and implement it by using algorithmic technology.Personal credit scoring on online platforms has changed the basic model of traditional credit scoring and induced many legal problems in the use of algorithms and data processing.The limitations of algorithm technology have created problems of algorithm instability,algorithm black box,and algorithm discrimination,which greatly affect the accuracy and reliability of personal credit scoring on online platforms,violate the legal rights of platform users,and also cause harm to the construction of China’s credit system.The operation of the personal credit scoring algorithm model of the online platform is based on the massive data in the platform,which is different from the traditional personal credit data,and the interaction between the user side and the platform side leaves a rich type and huge volume of big data,which contains not only a large number of users’ complete credit information,but also a large amount of fragmented information that indirectly reflects users’ credit Although the analysis of fragmented information by algorithms can enhance the richness of credit samples to a certain extent,the indirect credit information also affects the accuracy and reliability of the score.Personal credit score is not only a reflection of the platform user’s personal credit,but also a matter of whether the user can normally obtain the services provided by the platform.The technical and data limitations of the algorithm increase the inequality between the platform and the user,and the related problems need to be regulated by law.Although the credit systems in the European Union and the United States are more perfect,they do not specifically regulate credit scoring on online platforms,but they have legal measures to regulate similar algorithmic automated decision-making,and they have introduced many legal regulations related to the regulation of algorithmic technology and data.From the level of regulation of algorithm technology itself and data regulation of algorithms,we analyze the legal regulation measures of algorithms in the EU and the US in depth,and provide help for China to build a legal path to regulate personal credit scoring on online platforms.It can be found that there are some differences between the legal paths of algorithm governance in Europe and the United States,with the EU forming an algorithm legal regulation system with the GDPR and the Artificial Intelligence Act as the core,and the United States forming an algorithm legal regulation system with the“autonomous driving” and “deep forgery” as the core.In the United States,the legal regulation is carried out in specific areas of artificial intelligence and algorithm applications,showing the characteristics of local governance.In the legal regulation of algorithm technology,we should promote the establishment of a credible algorithm evaluation system,provide legal guarantee for algorithm transparency,pierce the black box of algorithm,and create a mechanism for algorithm supervision and accountability to improve the stability and reliability of algorithm operation.On the regulation of algorithm-applicable data,this paper advocates improving the principle of informed consent in data information collection,effectively safeguarding users’ right to know,promoting network platforms to conduct data compliance,clarifying the responsibilities and obligations of network platforms,reducing inequalities brought by the algorithm divide,balancing the contradiction between enterprise data utilization and protection with data compliance,promoting enterprises to implement the independent fulfillment of data safety and security obligations,and realizing network platforms To achieve multiple comanagement of personal credit scoring algorithm.