Research On Copyright Infringement Of Short Video Platform Based On Algorithm Recommendation Technology

Copyright has always been the child of technology from the beginning.With the development of the Internet,the copyright system of digital technology has once again ushered in the era of artificial intelligence and algorithm.The application of algorithm recommendation behavior is changing the ecological environment and interest pattern of the Internet.Can algorithm recommendation behavior be equivalent to the behavior of providing works? What is the impact of algorithm recommendation on the platforms attention obligation?These issues need to be further explored.The application of algorithm technology on short video platforms has led to rapid development of the short video industry,but the problem of copyright infringement has become increasingly severe.On the one hand,algorithm recommendation technology solves the problem of information overload on the Internet and greatly improves the matching efficiency of information.On the other hand,through algorithm recommendation,it is possible to gain a deep understanding of users needs and provide more personalized video content for users,which improves users’ satisfaction and loyalty to the platform.Information flow recommendation service is a typical algorithm recommendation technology,which is based on the historical selection records of platform users to discover and push personalized content to platform users.In the information flow recommendation service of short video platforms,the platform and users form a “multi-modal interconnection” mode with the purpose of providing specific content to the public: after the user clicks the button to agree to upload specific content,the “relay baton” is handed over to the short video platform,which uses technical means to recommend the content to the public.If the content uploaded by the user involves works protected by copyright,and the platform recommends the infringing works to other users of the platform through the information flow recommendation service,then there may be a problem of determining the infringement liability of the short video platform.In this “multi-modal interconnection” mode,the direct infringer is the user who publishes the infringing content,and the algorithm recommendation itself does not participate in the editing of the infringing content.According to the principle of technological neutrality,it cannot be determined that the short video platform has the intention to help infringement due to the application of technology in such infringement.However,the fault liability of“should know” requires the information flow recommendation service provider to bear the corresponding attention obligation for the infringing content.Especially in the context of the increasingly mature algorithm filtering technology,the platform should take the legality of the recommended content seriously and maximize the avoidance of copyright infringement risks.In addition,due to the complexity of infringement determination and the limitations of algorithm technology,we cannot impose too heavy attention obligations on short video platforms,which will cause the erosion of the copyright public domain by allowing the platform to replace public power agencies to carry out private law enforcement in the field of copyright,thus compressing personal use space,depriving reasonable use opportunities,and overriding users right to “counter-notification”.In the determination of copyright infringement on short video platforms under algorithm recommendation technology,it is particularly important to clarify the standard of determining the indirect infringement liability of the platform in order to avoid the platform’s evasion of legal liability by using the “safe harbor” rule.The determination of the platform’s indirect infringement liability is the compliance with the network copyright protection rules,which embodies the legal basis of the principles of technological neutrality,predictability,and the principle of equal rights and obligations.