Article 15 of Directive (EU)2019/790 (Digital Copyright Directive) grants publishers of press publications exclusive rights of reproduction and communication to the public for the online use of their press publications by information society service providers. In recent years, it has often happened that newspaper articles (like other copyright-protected content) have been used by some big tech companies to train their AI systems without the authorization of the rights holders. In this context, doubts have arisen as to whether training activities may fall under the copyright exception provided for in Article 4 of the Digital Copyright Directive for text and data mining operations (TDM).
At present, the issue is not settled: in some cases, the applicability of the TDM exception to the training of AI systems has been recognized, but some believe that training activities should be excluded from this exception.
The Court of Justice is called upon to remedy this uncertainty, to which the Budapest Court has requested a preliminary ruling (C-250/25) to clarify if the process of training an LLM-based chatbot through the use of press publications constitutes an instance of reproduction and, if so, whether such activity falls within the exception provided for in Article 4 of the Digital Copyright Directive.
