EU court adviser: Google Gemini risks copyright breach in AI case

• EU legal opinion says Gemini chatbot output could infringe on publisher rights
• First major EU guidance splits rules for AI model training and generated content

On November 3, 2026 (UTC), CoinDesk reported that an adviser to the European Court of Justice said Google’s Gemini chatbot may be breaching EU copyright law by reproducing the language of news articles in its responses. The case, brought by Hungarian publisher Like Company, directly challenges the legality of generative AI systems using online news content and highlights a core copyright issue: whether Gemini’s replies unlawfully copy protected expression from publishers.

Advocate General Maciej Szpunar clarified that while EU law permits large language models like Gemini to use protected press articles for training under text and data mining exceptions, copyright infringement occurs if a chatbot reproduces “the specific form of expression” from news articles during user interactions. This distinction allows AI to reuse facts but prohibits copying distinctive language or phrasing from press content in responses, which is the crux of the Hungarian publisher’s claim.

Reuters noted Google’s response that Gemini’s outputs are algorithmically generated and do not store or retrieve entire articles. Google also pointed out the technical mechanisms publishers can use to block their content from being accessed by chatbots.

Law Society Gazette reported that while the adviser’s opinion is influential, it is not binding. The final judgment from the European Court of Justice, expected in six months, will set a major precedent for how generative AI systems handle online news content throughout the EU.

This case marks the first substantive legal clarification of generative AI’s status under EU news publisher copyright, and its outcome could shape future rules for AI developers and rights holders across Europe.