Advocate General Maciej Szpunar delivered his opinion in Like Company v Google Ireland (C-250/25) on 3 November 2026, on a reference for a preliminary ruling from the Budapest Regional Court in Hungary. The opinion does not bind the Court of Justice, which will give judgment at a later date.
The reference turns on Article 15(1) of Directive (EU) 2019/790 on copyright and related rights in the Digital Single Market, which gives press publishers rights over the online use of their press publications, and on Article 4 of the same directive, the exception for text and data mining. Four questions were referred: two asking whether training a large language model on press publications engages the publisher's reproduction right or falls within Article 4, and two asking whether the answers a chatbot gives to users engage the reproduction right or the right of making available to the public.
For chatbot providers, including Google Ireland, the opinion sets out two factors that decide whether an answer engages the reproduction right: whether the chatbot accessed the press publication unlawfully, in particular by circumventing effective technical measures including a paywall or scraping restrictions, and whether the answer merely states facts or instead reuses the publication's form of expression so recognisably that the user no longer needs to go to the original. Advocate General Szpunar concluded that the right of making available to the public is generally infringed where that form of expression is systematically recognisable in answers to related user queries. He considered the provider liable for those acts where they are inherent to its business model, or where it has failed to put appropriate measures in place to counter them credibly and effectively.
Advocate General Szpunar treated the two questions on model training as inadmissible, finding them hypothetical because there was no evidence that the articles had been used to train the Gemini models, and because training is not online use of a press publication. For completeness he stated that training a large language model is, as a rule, permitted under the text and data mining exception in Article 4. The Court of Justice will deliver its judgment at a later date.
Licentium advises press publishers, platform operators and AI model providers on copyright and related rights in the European Union. Work we undertake includes assessing exposure under Article 15 of Directive (EU) 2019/790, reviewing text and data mining reservations and technical access controls, and preparing licensing positions for generative AI outputs.