The copyright protection of AI-generated content in video games

Abstract The increasing use of artificial intelligence in video game development, particularly through advanced procedural content generation, challenges traditional copyright frameworks. While AI-generated content is now integral to enhancing efficiency and player experience, its copyright status remains disputed, especially regarding the copyrightability threshold and ownership allocation. Existing discussions have largely focused on general works such as music, text, and images, with comparatively little attention to the complex, integrated nature of video game works. This article examines AI-generated content in video games from a comparative law perspective, focusing on four jurisdictions: the United States, China, the United Kingdom, and the European Union. It argues that, despite reduced human input, AI-generated content in video games involves human intellectual contributions at multiple stages, meeting prevailing copyrightability requirements. For ownership allocation, this article proposes a dual-structure model: creations led by video game companies fall under regimes that recognise the video game company as the author; creations led by individuals are considered as the work of AI users. This framework reconciles legal consistency with practical applicability, offering approaches of copyright allocation of AI-generated content in video game creation.