EU court advisor suggests gemini's use of news content may infringe copyright

Maciej Szpunar, Legal Counsel to the Court of Justice of the European Union (CJEU) in Luxembourg, indicated that Google's Gemini chatbot could potentially violate copyright when utilizing news reports to answer user queries under specific circumstances. This opinion arose from a lawsuit filed by the Hungarian website operator Like Company against Google, focusing on how the AI chatbot uses journalistic material. Szpunar stated that when an LLM retrieves a report from a publisher's site to generate an answer for a user, it may infringe on the publisher's rights. Specifically, the infringement could occur if the chatbot bypasses paywalls illegally or reproduces the specific expression of the news report in a recognizable way. However, Szpunar clarified that copyright protects the expression, not the underlying facts themselves. He also noted that training LLMs on copyrighted material is currently permissible under existing text and data mining exceptions. While the opinion from the CJEU's Legal Counsel is not legally binding, court judges often reference and adopt such advisory opinions. A ruling against Google could force the tech giant to compensate media outlets or could create new opportunities for online publishers facing declining traffic due to AI summaries replacing traditional search results.