Google AI chatbot’s use of press items may infringe copyright rules, EU court adviser says

In a case brought by Hungarian website operator Like Company against Google's artificial intelligence chatbot Gemini, an EU court adviser has stated that Google's use of press stories to generate responses to user queries may infringe copyright rules in certain situations. Advocate General Maciej Szpunar's advice to the Court of Justice of the European Union could potentially provide a lifeline to online media, which has seen declining traffic due to AI chatbots like Gemini replacing traditional search engine results with news summaries.

Szpunar noted that when an LLM-driven chatbot retrieves a publication from a press publisher's website, reuses it to generate an answer, and provides that answer to the user, the press publisher's right of reproduction may be infringed. This can occur when chatbots access publications by bypassing paywalls or other technical barriers, or by reusing the form of expression of the publication in a recognizable manner.

While copyright rules do not protect the mere facts reported by press publications, they do protect the specific form of expression. The use of copyrighted material to train large language models, however, is permitted under the text and data mining exception. The advocate general's opinions, while not binding, are typically followed by the court's judges.