EU Adviser Says Chatbot Reuse of News May Infringe
The advisory opinion, not yet a court ruling, could shape how generative AI products draw on press content in Europe.
Updated on Nov. 3, 2026 in Artificial Intelligence
Live Poll
Do you trust news organizations to stay healthy as AI chatbots reuse their reporting?
EU Advocate General Maciej Szpunar advised the Court of Justice of the European Union that Google's Gemini chatbot use of press publications may infringe copyright rules. The advisory opinion is research-stage legal guidance, not a court ruling, and the court's judges are not yet bound by it.
Why it matters
The opinion separates two questions: copyright does not protect the mere facts a press publication reports, only its specific form of expression, while training large language models on copyrighted material is freely permitted under the text and data mining exception. Opinions of the advocate-general are usually followed by the court's judges, so the reasoning could shape how generative AI products retrieve and reuse news content in Europe.
The advisory opinion is a legal analysis, not a measured technical result, and it reports no benchmark or performance figures. It characterizes the Gemini workflow as a chatbot that retrieves a press publication from a publisher's website and reuses it to generate an answer for the user.
The players
Court of Justice of the European Union
The EU's highest court, which will rule on the case and whose judges usually follow advocate-general opinions.
Maciej Szpunar
The EU Advocate General who issued the advisory opinion on Google's Gemini copyright question.
The company whose Gemini chatbot, built on its large language model stack, is the subject of the copyright challenge.
Like Company
The Hungarian website operator that brought the case against Google.
The details
An LLM-driven chatbot, meaning a chatbot powered by a large language model (a neural network trained on vast text collections to predict and generate language), retrieves a publication from a press publisher's website. It then reuses that publication to generate an answer and passes the answer to the user. The advocate-general separated this retrieval-and-reuse step from training, finding that using copyrighted material to train such models is freely permitted under the text and data mining exception, a copyright carve-out for automated analysis of works.
Timeline
The advisory opinion was issued on November 3, 2026.
The Tech Race
The opinion lands in a contest over how generative AI products access and reuse news content, with the EU court setting the legal boundary. It frames the training question separately from the retrieval question, which may influence how rival AI developers structure their data pipelines.
The advisory opinion is not a ruling, so Gemini users and publishers see no immediate change. The timeline for a binding decision from the court has not been announced.
The takeaway
Watch for the Court of Justice of the European Union's final judgment, which will confirm or reject the advocate-general's split between training and retrieval-and-reuse. That ruling will set the copyright boundary for how chatbots use press content in Europe.
What happens next
The Court of Justice of the European Union will issue its judgment in the case, which will determine whether the advocate-general's reasoning is adopted.
Further reading
For the broader policy and legal context, explore our Artificial Intelligence coverage.
Live Poll
Do you trust news organizations to stay healthy as AI chatbots reuse their reporting?
