论人工智能生成内容的著作权保护模式——以我国首例“AI文生图案”为例

The Copyright Protection Model for Artificial Intelligence-Generated Content——A Case Study of China's First"AI-Generated Image"Case

探讨我国人工智能生成内容适用狭义著作权保护面临的困境和难题,提出对人工智能生成内容采用邻接权保护的制度设计,为我国人工智能技术发展提供更为坚实的法治保障。通过梳理判断人工智能生成内容独创性存在的两大标准:“以作者为中心”的主观主义标准和“以作品为中心”的客观主义标准,分析当前存在的独创性判断困境。梳理关于人工智能生成内容作者认定的三种学说:将人工智能拟制为民事主体说、将人工智能研发者视为作者说和将人工智能使用者视为作者说,展现当前关于人工智能生成内容作者认定存在的难题。采用法律推理的方法,论证运用邻接权保护人工智能生成内容的可行性,并对人工智能生成内容的邻接权保护在制度设计上提出以下建议:人工智能生成内容的权利归属人工智能使用者;人工智能生成内容的权利内容以财产权为主;应对人工智能生成内容的权利设置必要限制。

This paper analyzes the dilemmas and challenges faced by the application of narrow-sense copyright protection to artificial intelligence-generated(AI-generated)content in China,and proposes a system design for protecting AI-generated content through neighboring rights,thereby cultivating the fertile legal soil for the development of artificial intelligence technology in China.By combing the two major criteria for determining the originality of AI-generated content:the"author-centered"subjective standard and the"work-centered"objective standard,the paper analyzes the current predicament in the judgment of originality.The article also reviews the three doctrines regarding the recognition of the authorship of AI-generated content:the doctrine of personifying AI as a civil subject,the doctrine of considering AI developers as authors,and the doctrine of considering AI users as authors,highlighting the current difficulties in recognizing the authorship of AI-generated content.Using legal logical reasoning,the paper argues for the feasibility of protecting AI-generated content through neighboring rights and makes the following suggestions for the institutional design of neighboring rights protection for AI-generated content:the rights of AI-generated content should be vested in the AI users;the rights content should be primarily proprietary;and necessary restrictions should be set for the rights of AI-generated content.