人工智能生成物的可版权性问题——日本经验与中国镜鉴
The Copyrightability of AI-Generated Works in Japan:Japanese Experience and Insights for China
张笑尘1
作者信息
- 1. 中南财经政法大学知识产权研究中心,湖北 武汉 430073
- 折叠
摘要
日本在人工智能生成物可版权性方面的研究,最早可追溯至20世纪90年代关于计算机生成创作物相关的学术讨论.通过对日本政府的多份官方报告与学术界各类观点进行梳理之后,不难发现在援引 日本著作权法中"选择的幅度"理论的背景下,即对著作物潜在的表达方式具有较大选择空间时,人工智能生成物是有可能被认定为受到著作权法所保护的著作物的,从而得到著作权法保护,对相关利益主体的利用行为起到规范作用.此外,日本在技术监管和著作权保护之间的平衡措施,例如侵权检测机制和数据使用的透明度要求等提案,也存在可操作性较高的参考价值,以期规范人工智能技术健康发展,同时有效保护著作权人利益.
Abstract
Research on the copyrightability of AI-generated works in Japan dates back to the 1990s,fo-cusing on academic discussions related to computer-generated works.A review of multiple official re-ports from the Japanese government and various academic viewpoints reveals that,under the framework of the"scope of choice"theory in Japanese copyright law-where a work's potential forms of expression al-low for considerable creative freedom-AI-generated works could potentially be recognized as copyright-protected works.This recognition affords them protection and regulation under copyright law,thereby standardizing the behavior of relevant stakeholders in utilizing these works.Furthermore,Japan's balan-cing measures between technological regulation and copyright protection,such as proposals for infringe-ment detection mechanisms and transparency requirements for data usage,offer valuable insights for Chi-na.These measures aim to regulate the healthy development of AI technology while effectively protecting the interests of copyright holders.
关键词
人工智能生成物/可版权性/著作权/创作性/日本著作权法Key words
AI-generated works/copyrightability/copyright/creativity/Japanese Copyright Law引用本文复制引用
出版年
2025
NETL
NSTL
万方数据