The European Court of Justice (ECJ) is set to rule on critical questions surrounding artificial intelligence (AI) and copyright law. The case, initiated by Hungarian publishing house Like Company against Google, centers on whether Google’s AI chatbot, Gemini, infringes copyright by reproducing copyrighted material.

The Core of the Dispute: Gemini and Copyright

Like Company, which operates the news page Balatonkornyeke.hu,alleges that Gemini reproduced a copyrighted article about a planned dolphin aquarium on Lake Plattensee. The publisher claims that Gemini’s summary was a near-literal reproduction of the original article.

The publisher accuses Google on constant legal violations because the search engine giant uses protected texts again and again without consent.

The Budapest district court (Környéki Törvényszék) has referred several questions to the ECJ, seeking clarification on whether displaying content in a chatbot that mirrors protected content constitutes an act of reproduction and public interaction. A ruling in favor of Like Company could have meaningful implications for AI development and copyright law in the EU.

The court also seeks to understand whether the predictive nature of chatbots like Gemini, which generate text based on observed patterns, affects copyright considerations. Additionally, the ECJ will examine whether training generative AI systems falls under reproduction rights, considering the EU’s exceptions for text and data mining.

Google’s Defense and Industry Concerns

Like Company accuses Google of repeated copyright infringements, arguing that the search engine uses protected texts without consent, exceeding the permissible use of “individual words or very short extracts.” Thay contend that Google’s training of Gemini violates reproduction rights.

Google maintains that its actions fall within the bounds of copyright exceptions, arguing that Gemini’s responses do not constitute public provision or duplication but merely reference facts contained in the protected content.

However, the Office of the EU for Intellectual Property (EuIPO) has indicated that many AI systems “obtain and use online content without prior approval of copyright owners.” Experts at the IPKAT specialist blog also suggest that chatbot operations impact reproduction and public accessibility rights.