The Copyrightability of Artificial Intelligence-Generated Content in China
Abstract
The commercial application of generative artificial intelligence has rapidly brought to the fore the structural tension between machine-generated content (AIGC) and the core presuppositions of copyright law concerning the “author”. Adopting a Sino-US comparative law approach, this paper examines the institutional divergences between the two countries on the question of AIGC copyrightability and their jurisprudential roots. The research posits that artificial intelligence possesses a dual attribute in the creative process—“transcending tools yet lacking complete creativity”—which constitutes a meta-constraint for institutional design. Through cases such as the “Feilin Case”, “Tencent Case”, “Li v. Liu”, and “Huan Zhi Yi Case”, Chinese jurisprudence has developed a three-stage evolution: from denying AI authorship to establishing human-machine collaboration, and then to strengthening evidentiary review, thereby establishing a process-oriented originality standard centered on substantial human intellectual input. On this basis, and through comparative analysis, this paper proposes a localized path for refining AIGC copyright protection tailored to China’s institutional context: establishing differentiated protection rules through tiered classification, constructing an originality assessment framework centered on human intellectual input, and introducing supporting mechanisms such as content labeling, good-faith disclosure, and extended collective licensing, in order to achieve an institutional objective that balances technological development with rights protection.
