EU Advocate General Says AI Chatbots May Infringe News Publishers’ Copyright

Luxembourg Advocate General Maciej Szpunar said LLM-powered chatbots can infringe news publishers’ rights when they retrieve an article from a publisher’s website, reuse it to generate an answer, and provide that answer to users. He set out the view in remarks reported by Wallstreetcn, focusing on how generative AI systems interact with news content.

Szpunar said infringement may occur when a chatbot accesses publications by illegally bypassing paywalls or other technical barriers. It may also occur when the system reuses “the expression form of the publication” in an “identifiable way,” rather than merely drawing on the underlying information.

In those circumstances, he said, an answer generated by Gemini would amount to an infringement of the publisher’s reproduction right. The comments place the focus on the output of AI chatbots and on whether it reproduces protected language rather than only conveying facts.