摘要：

以ChatGPT 为代表的生成式人工智能技术的发展，正在重塑著作权领域内容创作的底层逻辑，并对当前的著作权制度带来一系列颠覆性挑战，这些挑战集中于著作权法“只保护自然人创作”的前提假设、“思想表达二分法”的制度基石、“授权许可模式”的适用可能、“接触 + 实质性相似”的侵权标准，以及“避风港制度”的规制范畴等方面。从创作行为、规制主体、保护客体、权利范畴、侵权判定以及平台治理等层面进行考察可以发现，既有著作权制度已难以有效回应生成式人工智能技术所带来的底层冲击，未来的著作权法要进行彻底改革；而具体应当如何变革，则有待理论和实践层面的进一步观察和论证。

关键词:

生成式人工智能, 著作权法, 思想表达二分法, 侵权标准

Abstract:

The development of generative artificial intelligence technologies represented by ChatGPT is reshaping the underlying logic of content creation in the field of copyright, and it poses a series of disruptive challenges to the current copyright system. These challenges are primarily focused on the assumptions of the copyright law that “only protect works created by natural persons”, the institutional foundation of the “dichotomy between ideas and expressions”, the applicability of the “authorization and licensing model”, the infringement standards of “access + substantial similarity”, and the regulatory scope of the “safe harbor system”. By examining the aspects of creative behavior, regulatory subjects, protected objects, rights categories, infringement determination, and platform governance, it can be observed that the existing copyright system is struggling to effectively respond to the underlying impact brought by generative artificial intelligence technologies, necessitating thorough reform in the future of copyright law. However, the specific path of transformation awaits further observation and argumentation on both theoretical and practical levels.

Key words:

generative artificial intelligence,  copyright law,  dichotomy between ideas and expressions, infringement standards