EU Court Adviser Says Chatbot Answers Can Infringe Press Publishers' Rights, While AI Training Is Generally Allowed
In the first generative AI case at the EU's top court, Advocate General Maciej Szpunar said on November 3 that a chatbot such as Google's Gemini can infringe press publishers' rights when it reaches articles unlawfully, for example past a paywall, or reuses their wording so closely that readers no longer need the original. He said training on such content is, as a rule, allowed under the EU's text and data mining exception. The opinion is not binding.
The EU's top court has been handed a test that turns less on how AI models are trained and more on what chatbots say when they answer questions about the news. In an opinion delivered on November 3 in Case C-250/25, Like Company v Google, Advocate General Maciej Szpunar advised the Court of Justice that a chatbot pulling press articles in real time to answer users can, in certain circumstances, infringe the rights of press publishers. The court calls it the first generative AI case to reach it.
The case#
Like Company, a Hungarian operator of news portals, sued Google in Budapest. It claims Google used its articles to train the large language models behind Gemini and reproduced and made them available through the chatbot between June 2023 and May 2024. Gemini uses retrieval-augmented generation, fetching web pages live to ground its answers. The Hungarian court asked four questions under the EU's 2019 copyright directive, which gives press publishers a right over online use of their publications.
What the Advocate General advised#
Training. Szpunar told the court to declare the two training questions inadmissible because there is no evidence Like Company's articles were used to train Gemini and training is not "online use" of a press publication. For completeness, he said using protected works to train a model is, as a rule, freely permitted under the text and data mining exception.
Answers. On the chatbot's replies, he said the publisher's right protects the form of expression, not the facts reported. Whether an answer infringes depends on two factors: whether the chatbot reached the article unlawfully, in particular by circumventing technical measures such as a paywall or anti-scraping protection, and whether the answer merely reports facts or reuses the article's expression recognisably enough that the user no longer needs to read the original. The right of making available to the public is, as a rule, infringed where that expression is systematically recognisable in answers.
Liability. The chatbot provider is liable where such uses are inherent to its business model, or where it has not put in place credible, effective measures to counter them.
What happens next#
Advocate General opinions are not binding. The judges have begun deliberating and will rule later; the Budapest court then decides the dispute. Publishers in the US are pursuing similar fights through lawsuits, including the Elvis Presley rights holders' case against Anthropic over lyrics.
Limits#
- This account relies on the court's press release; we did not read the full opinion text.
- Google and Like Company statements on the opinion were not available to us.
- The court often, but not always, follows its Advocates General.