Research on Copyright Ownership of AI-Generated Content
人工智能生成内容的著作权归属问题研究

摘要

生成式人工智能介入文字、图像和音乐创作后，作品形成过程由传统的人类直接表达转向人机协同， 著作权归属因此面临新的判断难题。现行著作权法仍以人的智力创作为保护基础，人工智能本身不具备作者 资格。对于人工智能辅助生成的内容，是否构成作品及权利归属于谁，应考察使用者是否对最终表达形成可 识别的实质性创作贡献。司法实践已经出现保护与不保护两种裁判结果。应建立以人类创作贡献为核心的分 层认定标准，并完善生成记录和权属证明规则。

Abstract

With generative artificial intelligence increasingly involved in the creation of text, images, and music, the process of producing works has shifted from direct human expression to human–AI collaboration, giving rise to new challenges in determining copyright ownership. The current copyright law remains grounded in human intellectual creation, and artificial intelligence itself does not qualify as an author. For content generated with the assistance of artificial intelligence, whether it constitutes a copyrightable work and to whom the rights should belong should be determined by examining whether the user has made an identifiable and substantial creative contribution to the final expression. Judicial practice has already produced both protective and non-protective rulings. A tiered framework centered on human creative contribution should therefore be established, together with improved rules on generation records and proof of ownership.