The Munich Regional Court has issued a ruling that could become an important precedent for the entire generative AI industry. The lawsuit filed by the German rights management society GEMA against the music startup Suno was upheld in full. The court found that the company used six protected works from GEMA's catalog to train its neural networks and subsequently generate tracks without a corresponding licensing agreement.

At the center of the dispute were hits such as Daddy Cool, Rasputin, Forever Young, and Mambo No. 5. According to the court, these compositions were included in Suno's training datasets, which directly violates the exclusive rights of the rights holders. As an analyst, it is especially telling to me that the court did not delve into the technical details of "transformative use," but instead focused on the basic principle: any commercial use of copyrighted content requires permission, even if the end result is new music.

The startup itself has already stated its disagreement with the verdict and is preparing an appeal. It is expected that the key defense argument will be the fair use doctrine or its European equivalents, but the German legal system has historically been stricter on intellectual property matters than the American one.

This is not the first success for GEMA in its fight against AI companies. Earlier this year, the organization won a similar case against OpenAI related to the reproduction of song lyrics in ChatGPT responses. Thus, we are witnessing the formation of a consistent judicial practice in Germany that could serve as a reference for other European jurisdictions.

My expert commentary: This ruling is a clear signal for all AI service developers: the era of "gray" data collection methods is coming to an end. Investors and startup founders in this field should already be allocating significant sums in their budgets for content licensing, otherwise legal risks could prove more fatal than any technical issues. I believe that in the next 12–18 months, we will see a wave of similar lawsuits across Europe, which will ultimately lead to market consolidation and the emergence of legal data marketplaces for AI training.