Setback for Google as EU court adviser warns AI chatbot may breach copyright rules | Reuters
An advocate general at the EU's Court of Justice has advised that Google's Gemini chatbot may infringe publishers' copyright when it reuses news articles in recognisable form. The court's final ruling is expected in the coming months, and Google argues that the EU's text and data mining exceptions are needed for Europe's AI competitiveness.
- The EU court adviser's view is that publishers' rights may be infringed when Gemini reproduces articles in a recognisable form, which is a setback for Google.
- The Court of Justice of the European Union has yet to rule, with a decision due in the coming months, so this is advice, not a binding verdict.
- Google's defence is that text and data exceptions are vital for Europe's competitiveness and AI development.
The core issue is whether Google's Gemini chatbot infringes publishers' copyright when it reuses press articles in recognisable form. An advocate general at the Court of Justice of the European Union, the EU's top court, has given a non-binding opinion saying it may. That's a setback for Google, though not a final ruling. The court itself will decide the case in the coming months, and it typically follows its advocate general's view, so the opinion carries real weight. The dispute sits at the intersection of two things: publishers who want to be paid when their journalism feeds AI products, and AI companies who argue that training and text-and-data-mining exceptions in EU law protect their ability to build these systems.
Google's position is that those exceptions are essential for Europe to stay competitive in AI development, and it has warned that a narrow reading would handicap European innovation. The adviser's opinion cuts against that argument in one specific respect: it's not about whether a model can learn from text, but about whether the output reproduces protected articles closely enough to count as a copy. Reproducing an article in recognisable form, rather than summarising or paraphrasing it, is where the adviser located the potential infringement. That distinction matters because it narrows the fight. Google may be able to defend training on press content while still facing liability for what Gemini spits back out.
The friction here is part of a broader pattern. Publishers across Europe have been pressing platforms to pay for the use of their articles, and the chatbot era has raised the stakes, since users can get news-derived answers without ever clicking through to the source that produced them. Expect the Court of Justice's eventual ruling to set precedent well beyond Google, affecting how every AI developer operating in the EU handles licensing and output filtering. For now the question is open, the adviser has tilted it toward publishers, and the answer will come from Luxembourg in the months ahead.