Text and data mining exceptions in the development of generative AI models: What the EU member states could learn from the Japanese “nonenjoyment” purposes?

• DOI: 10.1111/jwip.12285
• Publication Date: 2023-05-27T18:54:57Z
• Authors: Artha Dermawan
• Keywords: Exclusive right

Abstract The European Union (EU) text and data mining (TDM) provisions are a progressive move, but the horizon is still uncertain for both generative artificial intelligence (GenAI) models researchers developers. This article suggests that to drive innovation further commitment digital single market, during national implementation, EU Member States could consider taking Japanese broad, all‐encompassing “nonenjoyment‐based” TDM as an example. “nonenjoyment” purposes, however, not foreign continental view of copyright. A similar concept can be found under German “Freier Werkgenuss” or enjoyment work. flexible exception built upon notion nonenjoyment purposes become opening clause foster creativity in age GenAI. Moreover, argues allowing with permissible so‐called three‐step test. suggests, if there no political will safeguard “the right read should mine” provide welcoming environment GenAI developers, when shaping legal interpretation through case law, following: (1) advocate 72 h response technological protection measures (TPMs) preventing TDM, (2) Robot Exclusion Standard (robot.txt) warning allowed on website. It now hands States, whether protect interests rightholders create balance between safeguarding mine,” protecting exclusivity, creating supportive researcher