Like Company v Google (C-250\/25) is the first case in which the Court of Justice of the European Union (CJEU) has been asked how EU copyright law applies to generative AI. A Hungarian press publisher claims that Google’s chatbot reproduced and displayed its articles and was trained on them without permission, and a Budapest court has asked the CJEU four questions about outputs, training and the text and data mining exception. The answer will bind courts in every EU Member State, so it matters to publishers, AI developers and any business that deploys chatbots in the EU.<\/p>\n
Like Company runs Hungarian online news portals. According to Bird & Bird’s account of the hearing<\/a>, it alleges that between 13 June 2023 and 7 February 2024 Google’s Gemini chatbot extracted and displayed substantial parts of its press publications in answer to user prompts, and that the content was also used to train the underlying large language model (LLM), all without authorisation or payment. The Budapest court hearing the claim stayed the proceedings and referred questions to the CJEU, as national courts can do when the interpretation of EU law is needed.<\/p>\n The claim rests mainly on the press publishers’ right. Under Article 15 of Directive (EU) 2019\/790<\/a> (the Digital Single Market or DSM Directive), publishers established in the EU have reproduction and making-available rights over the online use of their press publications by information society service providers. The right does not apply to hyperlinks or to individual words or very short extracts, and it expires two years after publication (counted from 1 January of the following year).<\/p>\n The four questions, as published in the Official Journal (C\/2025\/3039, 10 June 2025)<\/a>, are:<\/p>\nWhat exactly has the CJEU been asked?<\/h2>\n