EU Adviser Urges Court to Focus on Google AI’s Use of Journalism Beyond Training
EU adviser says Google's AI Gemini may infringe publishers' rights beyond data mining use.
Why it matters: A key CJEU ruling could reshape legal standards for AI's use of copyrighted press content. It impacts how AI developers use journalism and how publishers protect their rights internationally.
- Advocate General Maciej Szpunar advised the EU court that Google's Gemini chatbot may infringe press publishers' rights when reusing their content to answer users.
- Szpunar clarified facts in news are not copyright-protected, but specific expression is protected.
- Using copyrighted material to train AI under the EU's text and data mining exception remains allowed.
- The case was initiated by Hungarian website operator Like Company against Google, with the Advocate General's opinion non-binding but influential.
Advocate General Maciej Szpunar at the Court of Justice of the European Union (CJEU) has recommended the court assess Google's AI chatbot, Gemini, for potential infringements on press publishers' rights beyond the legality of training large language models (LLMs) on copyrighted content. Szpunar emphasized that while extracting "mere facts" from press publications is permissible under EU copyright law, the "specific form of expression" used by publishers is protected.
Szpunar stated, "When an LLM-driven chatbot retrieves a publication from a press publisher's website, reuses that publication to generate an answer, and provides that answer to the user, the press publisher's right may be infringed in certain circumstances." This perspective highlights concerns about Gemini's practice of repackaging journalism to generate user responses, not just the training of AI models on data.
The legal case was brought by Hungarian website operator Like Company against Google, focusing on Google's reuse of press content without explicit authorization. While Szpunar upheld that training AI with copyrighted material remains lawful under the EU's text and data mining exception, he urged the court to examine the narrower but significant issue of how the AI's output might infringe publishers' rights.
Although the Advocate General's opinion is non-binding, it holds considerable influence over the CJEU judges, who often adopt such recommendations in their rulings. The upcoming court decision could establish crucial precedent on how AI technologies interact with copyrighted press material, shaping the legal landscape for AI developers and media publishers across the European Union and beyond.