论人工智能生成物著作权归属的判断与解决路径——以AI文生图案为例

On the Judgment and Solution Path of Copyright Attribution of Artificial Intelligence Generated Objects—Taking AI Wenshengtu Case as an Example

摘要: 本文以AI文生图案为切入点探讨人工智能生成物著作权归属问题。通过详细分析作品特性模糊性和创作主体多元性对著作权归属判断造成的复杂性，进一步阐述人工智能生成物构成作品的依据以及常见归属模式的利弊。在此基础上，提出人工智能生成物应被认定为作品，且以归属于人工智能使用者为一般原则、以归属于人工智能投资者为特殊原则的观点，并给出相应的制度建议，旨在为解决人工智能生成物著作权归属争议提供理论支持，促进人工智能技术与著作权保护的协调发展。

Abstract: This paper discusses the issue of copyright attribution of AI-generated objects with AI Wenshengtu case as an entry point. By analyzing in detail the complexity caused by the ambiguity of the characteristics of works and the plurality of creative subjects on the judgment of copyright attribution, the paper further elaborates the basis for AI generated objects to constitute works as well as the advantages and disadvantages of common attribution models. On this basis, it puts forward the view that AI-generated objects should be recognized as works and attributed to AI users as a general principle and AI investors as a special principle, and gives corresponding system suggestions, aiming at providing theoretical support for solving the disputes over copyright attribution of AI-generated objects, and promoting the coordinated development of AI technology and copyright protection.