Construction Of Civil Legal Obligation System For Online Algorithmic Recommendation Service Provider

As network technology advanced,algorithms served as the underpinning logic.These algorithms were gradually updated,and a deeper level of algorithm technology was derived.As personalized recommendation algorithms became more sophisticated and the uses of these algorithms to provide algorithm recommendation services expanded,a number of issues also arose.In response to this circumstance,the legislator initially proposed the "Provisions on the Administration of Algorithm-generated Recommendations for Internet Information Services" to regulate the algorithm recommendation service providers,but due to a lack of theoretical and practical experience,the "Provisions on the Administration of Algorithm-generated Recommendations for Internet Information Services" legislative level is low and many other reasons,the regulation of algorithm recommendation service providers was ineffective,and the legal obligations required of service.In this paper,we examine the issues surrounding the legal obligations of algorithmic recommendation service providers under the current legal framework,combine existing legal norms and judicial practice of algorithmic recommendation services,summarize relevant theories and laws of other nations,conduct research using normative analysis and comparative study,and finally propose a construction plan of an algorithmic recommendation legal obligation system.This paper begins by outlining the special characteristics of algorithmic recommendation civil subjects and the need to broaden the scope of legal protection for new legal interests brought about by the current social development in order to explain why algorithmic recommendation service providers must assume civil statutory obligations.The paper then conducts a thorough analysis of the sources and current legal norms of algorithmic recommendation in China from a variety of legal effect levels,comes to the conclusion that the norms are fragmented and unsystematic,and suggests a preliminary framework for developing the civil legal obligation of algorithmic recommendation as a foundation for developing the subsequent obligation system.The second part of this essay examines the challenges involved in creating a system of civil law obligations for algorithmic recommendation service providers from two levels of legal precedent.At the level of legal standards,our analysis focuses primarily on the relationship and case law between the Civil Code and other legislation pertaining to algorithmic recommendation.There are currently few unambiguous legal requirements regarding algorithmic recommendation services,and the "Provisions on the Administration of Algorithm-generated Recommendations for Internet Information Services",which governs these services at the legislative level,is inadequate.For the jurisprudence level of system construction,the main issues are the inadequacy of the jurisprudence of the pertinent legal norms examined above for the regulation of algorithmic recommendation services,the lack of coordination between the traditional civil basic law and the development of science and technology,and the insensitivity to science and technology law.In three areas—principles generalization,legal argumentation,and empirical summary—this study concludes by suggesting improvements to the civil legal duty framework for algorithmic recommendation service providers.In order to complete and improve the entire system in detail,the framework of the aforementioned system is divided into distinct boundaries in the two dimensions of resolving the issue of "technology neutrality" and the limit of algorithmic transparency.The legal obligations of service providers are set in a hierarchical manner based on the number of users,technical capability,and revenue capacity of the services provided.