A Munich court has determined that Suno, an artificial intelligence music platform, unlawfully trained its models on six protected musical compositions from the GEMA repertoire without obtaining the necessary permissions or licenses. The ruling marks a significant moment for European copyright enforcement in the generative AI sector.
Copyright Infringement Confirmed
The German court’s judgment found that Suno’s training practices violated both German and United States copyright law. The tribunal ordered the company to secure a formal license agreement before continuing to use works from GEMA’s catalog in its AI training processes and for what it termed “systematic exploitation of the repertoire.”
The decision underscores growing tensions between the rapid development of generative AI technologies and established intellectual property protections across Europe. As AI music generation platforms proliferate, questions about how companies may legally access training data have become increasingly urgent for rights holders and technology developers alike.
Rights Organizations Welcome Clarity
Dutch rights management organisation BumaStemra, which manages performance and mechanical rights for creators across the Netherlands and Belgium, welcomed the court’s decision as an important clarification of the legal landscape. According to Tobias Holzmüller, associated with the organization’s position on this matter, “AI music platforms that want to use the repertoire managed by BumaStemra for generative AI must obtain prior permission and agree on a fair remuneration.”
The statement reflects a coordinated approach among European collecting societies to ensure that AI developers recognize existing licensing frameworks rather than treating protected works as freely available for machine learning purposes.
Broader Implications for EU AI Regulation
While the case was adjudicated in a German court, it carries implications that extend across the European Union’s regulatory environment. The judgment reinforces the principle that European copyright protections apply regardless of where AI platforms operate, even when the technology companies themselves are based in the United States.
The ruling also provides concrete judicial support for the European Union’s approach to generative AI governance, which emphasizes that existing intellectual property laws remain applicable to new technologies. This stands in contrast to some arguments from AI developers who have suggested that copyright rules should be relaxed or reinterpreted for training purposes.
Setting Precedent in Emerging Field
As courts across Europe increasingly address how artificial intelligence intersects with copyright law, decisions like Munich’s ruling establish important precedents. The case demonstrates that major music-generating AI platforms cannot assume automatic rights to training data simply because such data is publicly accessible or because the training occurs at scale.
For the European startup ecosystem and the broader creative industries, the judgment signals that sustainable AI development in music and likely other creative sectors will require negotiated licensing agreements and cooperation with established rights holders, rather than unilateral decisions by technology companies.