算法推荐时代版权过滤义务的法理证成与制度构建

Construction of Copyright Filtering Obligations in the Era of Algorithmic Recommendations

摘要: 算法推荐技术的广泛应用引发了新时代的版权治理困境，形成于网络发展初期的“避风港”规则也逐渐显露出其在适用上的不足。算法推荐提升了网络服务提供者的信息管理能力，为网络服务提供者带来了巨额流量经济；相应地，网络服务提供者应当承担与之相匹配的义务。在此背景下，作为平衡各方当事人的有力工具，引入版权过滤义务是立法的应然转向。为避免普遍过滤的诘难，引入版权过滤义务首先要对义务主体、过滤标准和成本分担进行明确的规定，同时还要注意其与“避风港”规则的有效衔接以及对错误过滤的救济，以此确保对侵权行为的有效预防。

Abstract: The widespread application of algorithmic recommendation technology has triggered the copyright governance dilemma in the new era, and the “safe harbor” rule, which was formed at the early stage of network development, has gradually revealed its inadequacy in application. Algorithmic recommendation enhances the information management ability of network service providers, and brings huge traffic economy for network service providers; accordingly, network service providers should undertake obligations to match. In this context, as a powerful tool to balance the parties, the introduction of copyright filtering obligations is a natural shift in legislation. In order to avoid the challenge of universal filtering, the introduction of the copyright filtering obligation should firstly make clear provisions on the subject of the obligation, the filtering standard and the cost sharing, and at the same time pay attention to the effective connection with the “safe harbor” rule and the relief for wrong filtering, so as to ensure the effective prevention of infringement.