人工智能生成物的著作权保护研究

Research on the Copyright Protection of Artificial Intelligence-Generated Works

摘要: 随着生成式人工智能技术的迅速发展，人工智能生成的文字、图像、音频、视频等内容已广泛进入学习、工作和商业场景。人工智能生成物在提升内容生产效率的同时，也对传统著作权法提出了挑战。现行著作权制度以人类作者为基础，而人工智能生成物在作品认定、独创性判断、作者资格及权利归属等方面均存在较大争议。对于人工智能生成物，既不能一概否定其法律保护，也不宜不加区分地赋予著作权。应当结合具体生成过程，重点考察人类是否对最终表达作出了实质性智力贡献。只有在人的创作意图、选择、安排、修改等因素足以体现个性化表达时，相关成果才具备进入著作权保护范围的基础。基于此，本文主张坚持人类作者中心立场，对人工智能生成物采取分类保护路径，并通过细化独创性标准、明确权利归属规则、完善标识与留痕制度等方式，回应人工智能时代著作权保护面临的新问题。

Abstract: With the rapid advancement of generative artificial intelligence technology, texts, images, audio, and videos generated by artificial intelligence have been extensively utilized in learning, work, and business contexts. The products generated by artificial intelligence have not only enhanced the efficiency of content production but also presented challenges to traditional copyright laws. The current copyright system is centered around human authors. However, there are substantial disputes regarding the identification of works, the assessment of originality, authorship qualifications, and rights attribution in the context of artificial intelligence-generated works. Concerning artificial intelligence-generated works, it is not advisable to simply deny their legal protection, nor should copyright be granted indiscriminately. Instead, the specific generation process should be considered, with a focus on examining whether humans have made substantial intellectual contributions to the final expression. Only when factors such as human creative intention, choice, arrangement, and modification adequately reflect individualized expression can the relevant results have a basis for entering the scope of copyright protection. Based on this, this article advocates adhering to the human author-centered position, adopting a classified protection approach for artificial intelligence-generated works, and addressing the new problems faced by copyright protection in the era of artificial intelligence through measures such as refining the originality standard, clarifying the rights attribution rules, and improving the identification and traceability system.