Publisher:
Martin-Luther-Universität Halle-Wittenberg, Institut für Wirtschaftsrecht, Forschungsstelle für Transnationales Wirtschaftsrecht, Halle (Saale)
Abstract:
On 1 August 2024, the EU Regulation on Artificial Intelligence (AI Act) came into force after a difficult road. Originally conceived purely as a product safety law, the regulation has developed in reaction to public debate to include the protection of copyright used to train generative AI models such as the Large Language Model GPT developed by OpenAI. This article provides a legal classification of the relevant provisions and examines the influence of the AI Act on AI training, in particular how the copyright strategy required from the providers of generative AI models can be justified in terms of legal doctrine as well as economics and how it can be applied in practice.
