Can AI Infringe Moral Rights of Authors and Should We Do Anything About It: An Australian Perspective
Law, Innovation and Technology, published online 23 March 2023 https://doi.org/10.1080/17579961.2023.2184138
27 Pages Posted: 22 Mar 2022 Last revised: 4 Apr 2023
Date Written: January 24, 2022
Abstract
While artificial intelligence technologies hold significant potential for the Australian economy and social wellbeing, it is unclear to which extent copyright laws stimulate or impede the development of these promising technologies. While the implications of unauthorised use of copyright-protected works in the machine learning process on economic rights of authors have been previously explored, the implications of such use on the moral rights of authors – the rights of attribution and integrity – have not been examined. This paper analyses whether the use of works as training data in the machine learning process could amount to the infringement of moral rights of authors and, if so, whether law reform in the area is needed. As the Australian government is discussing an introduction of a new text and data mining exception, it is important to examine whether such exception should also cover moral rights. The paper concludes that, while the unauthorised use of works in machine learning projects might in some cases lead to prima facie infringements of moral rights, the broad and flexible ‘reasonable use’ exception that exists under Australian copyright law, would shield artificial intelligence developers from liability in many cases, and further limitation to moral rights is not required.
Suggested Citation: Suggested Citation