An adviser to the European Union’s top court said Google’s Gemini could infringe publishers’ rights if it retrieves press articles and reproduces protected material in answers to users.

The issue is Gemini’s retrieval-augmented generation, which searches online sources in real time; the opinion distinguishes this from training an AI model on large volumes of text.

Advocate General Maciej Szpunar said copyright protects the specific expression used to report news, not the facts themselves.

Szpunar treated AI training separately, saying it is generally allowed under the EU’s text-and-data-mining exception and recommending that the court not address training in this dispute because there was no evidence that Like Company’s articles had been used.

The case was brought by Hungarian website operator Like Company against Google, amid publishers’ concerns that AI-generated summaries could divert traffic and threaten advertising and subscription revenue.

A ruling in publishers’ favor could strengthen their control over how reporting is used, bolster licensing negotiations and push AI companies to answer users without reproducing protected material.

Google argued that Gemini creates original responses rather than copying articles, pointed to publisher opt-out controls and welcomed the opinion’s recognition that EU law permits AI training on publicly available information.

Szpunar’s opinion is non-binding, and the European Union’s court has yet to issue its final ruling; its decision could shape rules for technology platforms, generative AI services and digital publishing across the bloc’s 27 member states.

The court is expected to rule in the coming months, and a decision against Google could require licensing payments to European news outlets; the case will then continue through the Hungarian legal process.