论人工智能生成内容的可版权性:实务、法理与制度

On the Copyability of Content Generated by Artificial Intelligence:Practice,Jurisprudence and System

生成式人工智能对著作权法带来挑战,涉及作品独创性和作者主体性的核心问题。国内外司法机关、行政主管部门以现行法为依据,采取法教义学的解决方法,对人工智能作品可版权性争端作出回应。尽管裁判结果有别,但其主旨都表现了强调人类作者的著作权主体地位,维系主体(作者)与客体(作品)二分原则的法政策取向。面对人工智能技术迭变,著作权法理论需要反思和重构,从法理学层面化解独创性标准与传统人格理论的紧张关系,正视人类作者与机器作者的共创事实,重塑事实作者、拟制作者与著作权人的主体构成。在法律构造方面,不宜根本颠覆现有基本制度框架,但可对相关法律规范进行补充和完善,包括人工智能作品的客体规范、人工智能创作的主体规范和人工智能著作权归属的本体规范。

Generative artificial intelligence poses challenges to copyright law,including the core issues of the originality of works and the eligibility of the author.Domestic and international judicial and administrative authorities use methods of dogmatics of law to respond to disputes over the copyrightability of Al-generated works based on current laws.Although the judgments vary,they all emphasize the copyright subject status of human authors and maintain a legal policy orientation that upholds the dichotomy between subjects(authors)and objects(works).Faced with the rapid evolution of AI technology,copyright law theory needs to be reflected and reconstructed.It is necessary to resolve the tension between the originality standard and traditional personality theory from a jurisprudential perspective,to acknowledge the reality of co-creation between human and machine authors,and to reshape the composition of factual authors,constructive authors,and copyright holders.In terms of legal structure,it is not advisable to fundamentally overturn the current institutional framework.However,it is feasible to improve relevant legal norms including the object norms of AI-generated works,the subject norms of AI creation,and the ontological norms of AI copyright ownership.