Gemini Faces New European Copyright Warning

Europe's highest court has been urged to recognise that artificial intelligence chatbots can infringe publishers' rights when they reproduce protected news articles in responses to users, potentially reshaping the relationship between technology companies and the media.

Advocate General Maciej Szpunar issued the non-binding opinion on Tuesday in a case brought by Hungarian website operator Like Company against Google. The dispute concerns Gemini's use of online articles to produce answers. Szpunar said liability could arise when a chatbot reproduces a publication's recognisable expression or obtains material by circumventing technical restrictions, including paywalls. Reporting facts alone, he argued, does not necessarily infringe copyright.

The adviser drew a distinction between retrieving articles to answer questions and using material to train large language models. He recommended that the court decline to rule on the training questions because the case lacked evidence that Like Company's work had been used for that purpose. He also suggested that training on protected material could generally fall within an EU exception for text and data mining.

The opinion is not a final judgment, and the Court of Justice of the European Union may reach a different conclusion. Nevertheless, it highlights the stakes for publishers whose reporting increasingly appears in AI-generated summaries rather than traditional search results. The case could also influence licensing negotiations between AI companies and news organisations across the continent. The forthcoming judgment could clarify when developers must obtain permission or pay for reproduced content, while preserving the distinction between protected writing and facts available to everyone.