Who Owns Ai-Generated Content? Examining Copyright Challenges in The Era of Generative Artificial Intelligence

Abstract

The emergence of generative artificial intelligence technologies such as large language models and image diffusion models has fundamentally changed the landscape of creative work, as machines are now capable of producing text, images, music, and audiovisuals with a level of sophistication that resembles human work, thus giving rise to legal uncertainty regarding who is entitled to be called the creator and copyright holder of such content. This article aims to critically examine the conceptual and normative challenges faced by the copyright regime, particularly Indonesian copyright law, in responding to the phenomenon of works produced by generative artificial intelligence, by placing the principles of legal flexibility and adaptability as the main analytical framework amidst global technological disruption. This research uses a normative juridical method with a statutory approach, a conceptual approach, and a comparative legal approach, by examining relevant primary, secondary, and tertiary legal materials, including developments in decisions and policies in the United States, China, and Indonesia. The study results show that the requirements of authenticity and human authorization that are the foundation of Law Number 28 of 2014 concerning Copyright have not explicitly anticipated works born from autonomous generative processes, thus creating a normative vacuum regarding the legal status of artificial intelligence outputs, responsibility for copyright infringement in the data training process, and the division of ownership between users, platform developers, and training data providers.