This week’s addition to our library is a Primer by Alexander Peukert, focused on how copyright issues are addressed in the European Union’s AI Act. Culture Action Europe, through its Digital & AI Action Group, takes a deep interest in the questions surrounding the intersection of technology and cultural rights.
The European Union’s AI Act, which entered into force on the 1st of August 2024, introduces new regulations aimed at addressing the growing intersection between artificial intelligence and copyright law. Two key provisions in the Act, Articles 53(1)(c) and (d), set out specific requirements for providers of general-purpose AI models to implement policies that comply with the EU copyright laws. Additionally, AI providers must publish sufficiently detailed summaries of the content used to train their models, increasing transparency around the data that fuels AI development.
In his recent article on the AI Act and copyright, Dr Peukert explains how the AI Act merges copyright law and AI regulation. On the one hand, copyright is a private right that allows rightsholders to seek remedies through civil courts for any infringement. On the other hand, the AI Act establishes obligations for AI providers to protect the public interest, such as trustworthy AI and product safety. Here we enter the public law domain. The compliance obligations are enforced by public authorities, namely the European Commission or national authorities, which act if they find AI models to be non-compliant. This overlap could potentially create challenges at the enforcement level.
This resource delves into the origins of these provisions, examining the motivations behind them and exploring how they impact key players in the AI industry and is a must-read for anyone interested in AI and digital transformation in the cultural sector.
Read the full paper below:
read it here
Peukert, Alexander, Copyright in the Artificial Intelligence Act – A Primer (March 25, 2024). GRUR International, Volume 73, Issue 6, June 2024, Available at SSRN: https://ssrn.com/abstract=4771976 or http://dx.doi.org/10.2139/ssrn.4771976