Research On Copyright Infringement Identification Of Content-sharing Service Provider Under Algorithm Based Recommendation

Algorithm based recommendation service has become an important means for content sharing service platform to obtain and serve users,but if such means do not control the source of works,the problem of copyright infringement will become prominent.The current copyright law system mainly relies on the "safe harbor" rule under the principle of technology neutrality to deal with this problem.In judicial practice,the dispute about whether the platform providing algorithm recommendation service should bear the responsibility mainly lies in the subjective fault of the platform.However,since the "safe harbor" rule is based on a specific historical background,it is difficult to adapt to the new technological environment.There is a lack of unified standards for the identification of the subjective fault of the platform.The new regulations of the the Cyberspace Administration of China on Algorithm recommendation management only make a posture response,and there is a lack of specific institutional arrangements for the platform to bear a higher duty of care.Under the principle of subjective and objective consistency,the content sharing service platform under algorithm recommendation is difficult to enter the "safe harbor" due to the enthusiasm and initiative of algorithm recommendation and its obvious value orientation.In addition,with strong information control ability,it should bear a higher duty of care,that is,take reasonable measures.From the perspective of technology,jurisprudence and economic logic,it is reasonable for the platform to undertake the filtering obligation.Combined with extraterritorial legislative experience and Chinese practice,China should unify the subjective fault determination standards of the content sharing service platform recommended by the algorithm,and improve the copyright infringement determination rules of the content sharing service platform under the algorithm recommendation with the filtering obligation as the core.