Case C-250/25, Like Company v Google, was referred to the Court of Justice of the European Union by a Hungarian regional court (Budapest Környéki Törvényszék) in April 2025. It concerns the operation of chatbots in the context of copyright law, mainly the use of content posted on the internet by chatbots operating on the basis of large language models (LLMs), such as Google Gemini (formerly Bard).
The CJEU will have to answer questions concerning the relationship between AI and copyright, in particular in the context of related rights of press publishers and exceptions for text and data mining (TDM) under Directive 2019/790 (‘DSM Directive’).
Summary of the facts
The Hungarian media company Like Company, publisher of copyright-protected content posted, i.a., on the balatonkornyeke.hu portal, sued Google Ireland Ltd., operator of the AI chatbot Google Gemini, alleging infringement of copyright to press publications. The chatbot used data from Google Search to generate summaries and quotations from this content, allowing users to access it without having to visit the publisher’s websites, which generate revenue from advertising displayed to visitors.
According to Like Company, between 13 June 2023 and 7 February 2024, Google electronically reproduced and made available to the public its press content without the publisher’s consent, thereby infringing Article 15 of the DSM Directive. Google disputes these allegations, arguing that the chatbot’s actions do not constitute reproduction or making press content available to the public, and therefore the plaintiff’s rights have not been infringed.
The dispute therefore comes down to whether the chatbot’s actions require the publisher’s consent and payment of remuneration.
Legal context
Article 15 of the DSM Directive grants press publishers a related right to control the reproduction and making available of their press publications by information society service providers. The scope of protection corresponds to the rights provided for in Articles 2 and 3(2) of Directive 2001/29/EC. In addition, Article 4 of the DSM Directive requires Member States to introduce an exception or limitation to the rights of extraction and reproduction for the purposes of text and data mining (TDM).
Request for a preliminary ruling – questions to CJEU
In its request to the CJEU, the Hungarian court asks, i.a., whether chatbots may infringe the related rights of press publishers by reproducing and making available copyright-protected content. One of the issues is also whether training artificial intelligence on copyright-protected material is legal and whether it can be subject to the TDM (text and data mining) exception provided for in the DSM Directive. The Hungarian court seeks clarification on the following issues:
- Does the display of excerpts from press publications by a chatbot constitute making them available to the public (within the meaning of Article 3 of Directive 2001/29/EC)?
- Does training a chatbot on copyright-protected texts constitute illegal reproduction of those works?
- Does the exception for text and data mining (TDM) provided for in the DSM Directive allow the copying of copyright-protected works for the purpose of training artificial intelligence systems?
- Does a response generated by a chatbot based on a user prompt indicating specific texts and articles constitute reproduction of those works within the meaning of copyright law?
Significance of the case
The present preliminary ruling proceedings concern key issues of copyright law in the context of the development and use of artificial intelligence. The CJEU’s answers may be important for assessing the legal status of the ways in which chatbots and other AI tools use copyright-protected content, as well as potential infringements of those rights. They will also have an impact on the interpretation of exceptions relating to text and data mining.
This case is important not only for digital content publishers, but also for entities developing artificial intelligence models, from the largest technology companies to start-ups and research institutions. The CJEU’s response may play a key role in defining the limits of liability and the rules for the permissible use of copyright-protected content in the context of automation and the development of generative AI.
We will closely monitor the course of these proceedings and their potential consequences for market and regulatory practice.