算法推荐模式下网络服务提供者的版权侵权研究

Research on Copyright Infringement of Network Service Providers under Algorithmic Recommendation Model

摘要: 我国网络服务提供者在较长的一段时间内，作为中立的平台对其平台上存在的侵权作品承担着比较低但也较为合适的责任。而平台之所以承担此种较低的责任正是因为其不直接参与侵权，始终保持在一个中立的位置。然而，算法的出现和发展打破了这一局面，算法推荐的模式愈发成熟，其对用户行为的介入程度也逐步加深，且相较于以往更加具备隐蔽性和专业性，技术似乎已经不再那么中立，版权者在这场权益争夺战中逐渐落入下风。另外，在司法实践中，因为算法的存在，也出现了对网络服务提供者责任认定的分歧，注意义务的适用比较混乱，事后保护的模式也稍显乏力。要在算法时代维持网络服务提供者与版权者的利益平衡就有必要考察不同类型的推荐模式，以及不同平台的抗风险能力，在此基础上明确注意义务的适用层次，也要进一步地加强针对平台上版权侵权的事前保护。

Abstract: For a long time, Chinese internet service providers have been bearing relatively low but appropriate liability as neutral platforms for copyright infringement on their platforms. The reason why platforms bear such lower liability is that they do not directly engage in copyright infringement and always maintain a neutral position. However, the emergence and development of algorithms have changed this situation. The algorithm recommendation model is becoming more and more mature, and its intervention in user behavior is gradually deepening. Compared to the past, it is more hidden and professional, and technology seems to no longer be neutral. Copyright holders are gradually falling into disadvantage in this rights struggle. In addition, in judicial practice, because of the existence of algorithms, there are also disputes on the liability of internet service providers. The application of due diligence is relatively chaotic, and the post-event protection model is somewhat ineffective. In order to maintain a balance between the interests of network service providers and copyright owners in the era of algorithms, it is necessary to examine different types of recommendation models and the anti-risk capabilities of different platforms, and on this basis, clarify the application level of the duty of care, and further strengthen the prior protection against copyright infringement on the platform.