Google’s Gemini may infringe EU publishers’ rights when it reuses a news article to answer a user’s question, Advocate General Maciej Szpunar said in a non-binding opinion for the Court of Justice of the European Union. Szpunar said that applies when a chatbot gets past paywalls or similar blocks, or repeats an article’s wording in a recognizable way, since copyright covers how a story is written but not its facts. However, he said the EU’s text and data mining exception freely allows training AI on copyrighted material, and he found no evidence Google trained Gemini on articles from Like Company, the Hungarian publisher that sued. Google said Gemini produces original answers and lets publishers opt out. The court usually sides with its advisers and is expected to decide within months.
