Who Owns AI-Generated Works?A Comparative Analysis of International Copyright Frameworks

Abstract

The rapid proliferation of generative artificial intelligence has forced a fundamental re-examination of a question copyright law has never had to answer: can authorship exist without a human author? This article undertakes a comparative analysis of the international copyright frameworks operative in five jurisdictions — the United States, the European Union, the United Kingdom, Japan, and the People's Republic of China — set against the foundational international instruments of global copyright protection: the Berne Convention, the TRIPS Agreement, and the WIPO Copyright Treaty. Drawing on landmark judicial decisions including Thaler v. Perlmutter (D.C. Circuit, 2025), Li Yunkai v. Liu Yuanchun (Beijing Internet Court, 2023), and the CJEU's Infopaq line of cases, as well as the U.S. Copyright Office's January 2025 Part 2 Report on Copyrightability, this article demonstrates that the current international legal architecture is anchored to a human-centred conception of authorship that precludes autonomous AI systems from holding copyright. It further identifies significant and growing divergence in how jurisdictions treat the human-AI collaborative spectrum, and argues that this fragmentation creates regulatory arbitrage, impedes cross-border creative industries, and demands multilateral reform. The article concludes with a proposal for a WIPO-led harmonisation instrument built on the concept of qualified human creative control.