Google’s Gemini chatbot could infringe publishers’ copyright in certain cases when it uses press articles to generate answers, an adviser to the European Union’s highest court said on Tuesday, 3 November 2026. The opinion concerns a case brought by Hungarian publisher Like Company against Google and is the first generative AI dispute of its kind to reach the Court of Justice of the European Union.
Advocate General Maciej Szpunar said the court should distinguish between the material used to train an AI system and the way a chatbot later retrieves and presents information to users. His recommendation is not binding on the judges, but such opinions are influential and are frequently followed. A final judgment is expected in approximately six months.
“When an LLM-driven chatbot retrieves a publication from a press publisher’s website, reuses that publication to generate an answer, and provides that answer to the user, the press publisher’s right may be infringed in certain circumstances.”
— Advocate General Maciej Szpunar, as quoted in the court’s statement
The opinion focuses on situations in which a chatbot may access an article by bypassing a paywall or another technical restriction, or reproduce the article’s recognisable form of expression in its answer. Copyright protection would not extend to the mere facts reported in a publication, according to Szpunar. It would instead apply to the particular way those facts are expressed.
Training and chatbot answers treated separately
Szpunar said that using copyright-protected material to train large language models is “freely permitted under the text and data mining exception” in EU law. However, he also said there was no evidence that Like Company’s articles had been used to train Gemini. The case therefore turns on the chatbot’s handling of press content when responding to users, rather than establishing that the articles were part of Gemini’s training data.
The legal distinction could be important for publishers whose articles are summarised by AI systems instead of being displayed as conventional search results with links. Like Company’s case seeks to establish whether Google must pay for the use of its content. A decision against Google could strengthen the negotiating position of online media organisations, although the advocate general’s recommendation does not itself order compensation or establish a final liability finding.
- Training: AI training on publicly available material is covered by the text and data mining exception, according to the opinion.
- Generated responses: Reusing a publication’s protected expression, or accessing it by circumventing restrictions, may infringe a publisher’s rights.
- Next step: The EU court is expected to issue its judgment in around six months.
Google said Gemini creates original outputs by learning linguistic patterns and pointed to controls that allow publishers to opt out. The company also welcomed the opinion’s recognition that EU law permits AI training on publicly available information. For Indian users and publishers, the case is significant as a closely watched test of how courts may draw the line between factual reporting, protected expression and AI-generated summaries, even though the pending judgment concerns EU copyright rules.
