Background
GEMA brought proceedings against OpenAI before the Munich Regional Court I, alleging that OpenAI reproduced protected musical works without authorisation in the course of training its AI systems. The court’s official press release explains that the training process involved “das Abspeichern von urheberrechtlich geschützten Werken” [the storing of copyright-protected works], which it classified as “eine Vervielfältigung im Sinne des § 16 UrhG” [a reproduction within the meaning of § 16 of the German Copyright Act]. The externally available judgment text also references “öffentliche Wiedergabe” [communication to the public] in assessing user-facing outputs. The court ruled in favour of GEMA, finding that OpenAI’s training and output-related uses breached German copyright law, and the case concluded on 11 November 2025.
AI Interaction
The judgment directly addressed how OpenAI’s AI systems interacted with protected material during both training and output generation. The court noted that the training corpus contained protected works through “das Abspeichern von urheberrechtlich geschützten Werken” [the storing of copyright-protected works], classifying this internal data handling as “eine Vervielfältigung im Sinne des § 16 UrhG” [a reproduction under § 16 UrhG]. It further considered whether model outputs accessible to users qualified as “öffentliche Wiedergabe” [communication to the public], treating the AI system’s functioning as a legally relevant act attributable to OpenAI.
Note/Update:
• Final judgment issued on 11 November 2025 confirming infringement by OpenAI.
• Summary relies primarily on the official press release; German-language quotations from the externally circulated Endurteil are included with English translations in brackets for clarity.
Other relevant documents: https://ifrro.org/resources/documents/General/Decision_German_court_42_O_14139-24_Endurteil.pdf