EU Court Adviser Warns Google Gemini May Violate Copyright via News Usage
An advocate general at the EU Court of Justice states that Google's Gemini chatbot could infringe on copyright by reproducing journalistic content in its responses.
A legal adviser to the European Court of Justice (CJEU), Maciej Szpunar, has issued an opinion suggesting that Google's Gemini AI could violate copyright law when it uses news reporting to generate answers. The opinion follows a lawsuit from Hungarian website operator Like Company, which challenged how AI models process and present news content.
Szpunar noted that while using copyrighted material for training large language models is generally permissible under data mining exceptions, the output of these models is subject to different standards. He argued that if an AI bypasses paywalls or reproduces the specific expression of a journalist, it may constitute a violation of the publisher's reproduction rights. Facts themselves, however, remain unprotected by copyright.
This development comes as publishers globally express concern that AI-generated summaries are eroding traffic to original news sites. Although the advocate general's opinion is not legally binding, judges at the CJEU frequently adopt such guidance when delivering their final rulings. If the court rules against Google, it could establish a precedent requiring technology companies to compensate news organizations for content use.