The Tension Between Humanism and Capital: The Binary Dilemma in Copyright Protection for AI-Generated Audiovisual Works and Graded Solutions

Abstract

The rapid advancements in generative AI have intensified the theoretical conflict between “human-centered” principles and “investment incentives” in copyright protection for AI-generated audiovisual works. Analysis of judicial practices, such as the Ultraman AI copyright case, reveals structural contradictions within the traditional copyright system—particularly in determining rights holders and establishing protection criteria. To address this, we propose a tiered rights-determination model based on the “Human Intervention Index (HII).” This framework classifies AI-generated works into three tiers according to the depth of human intervention, granting them either full copyright, specific neighboring rights, or public-domain status. The model offers an institutional pathway to balance creator rights and industrial innovation.