09 – Copyright and data protection rights in co-generated input and output of generative AI

The aim of this study is first to review the legal frameworks applying to co-generation scenarios (considering, at least, co-generation of data as such, co-generation of models, and co-generation of content). The conclusions envisaged at the end of the study aim to reflect on the current ethical and policy debates around co-generation. This study focuses on the regulatory frameworks of the EU, Japan and the US, and offers a high-level overview on the hard and soft law developments particularly in the U.S. and Japan on copyright, as well as on the policy discussions regarding rights in co-generated input and output of Generative AI. Before reviewing the legal frameworks, it is important to assess the notion of Generative AI (GenAI) as well as discuss what co-generation, as an overarching concept, means in the context of this study (A.II.). Furthermore, as explained below this study will distinguish co-generated input and co-generated output of GenAI (A.IV.) rather than rely on the differentiation between co-generated data, co-generated models or co-generated content.