算法推薦下版權過濾義務的構建

Construction of Copyright Filteration Obligation Based on Algorithm Recommendation

算法推薦行為於版權法中的定性是網絡服務提供者承擔侵權責任的邏輯起點。雖然利用算法推薦技術的網絡平臺仍屬網絡服務提供者，但技術發展使其與傳統的網絡服務提供者有所區別，故在侵權責任的承擔上不宜當然適用“ 避風港” 規則。算法推薦提升了網絡平臺的信息管理能力，為網絡平臺帶來了巨額流量經濟;相應地，網絡平臺應當承擔與之相匹配的義務。版權過濾義務作為制度回應技術的具象化舉措，能夠有效應對算法推薦下的版權保護困境。引入版權過濾義務首要面臨且難以回避的問題是其與“避風港”規則如何有效銜接，而限縮版權過濾義務承擔主體的範圍不失為解決良策。相應地，有必要構建起“內容過濾+必要措施+異議救濟”的版權責任體系。

The copyright legality of algorithm recommendation behavior is the logical starting point for internet service providers to assume copyright infringement liability. Although the use of algorithm recommendation by network platform belongs to the behavior of network service provision, the development of technology makes the network platform different from the traditional internet service provider, which means that the safe harbor principle cannot be applied naturally. The algorithm recommendation improves the information management ability of the network platform, and brings a huge amount of traffic economy. Accordingly, the network platform should assume the corresponding obligations. As the embodiment of institutional response technology, the copyright filteration obligation can effectively solve the dilemma of copyright protection under the network environment. The connection with the safe harbor principle is the first and unavoidable problem in introducing the copyright filteration obligation. The copyright liability system of “content filteration + necessary measures + objection relief” is constructed on the basis of the effective connection between the subject of copyright filteration obligations and the safe harbor principle.