The EU Copyright Exceptions for Text and Data Mining: (Un)suitable Model Solution for the Development of Artificial Intelligence?

Abstract

Two copyright exceptions for text and data mining (TDM) have been recently introduced in European Union (EU) law. The first "stronger" exception covers TDM conducted for the narrow purposes of scientific research by research organisations (non-profit or tasked with a public interest mission) and cultural heritage institutions. The second " weaker" exception refers to TDM in general, but the rightholders have the power to opt out from it (and after doing so, to license TDM of their content). These exceptions have become a target of immense criticism concerning their (un)suitability to adequately support further development of artificial intelligence in the European Union. This article offers an overview of both the EU TDM copyright exceptions and the ongoing critical discussions accompanying them. The aim is to determine what lessons and/or guidelines could be drawn from this EU experience for other countries that might be considering the introduction of a TDM copyright exception in their national laws.