German court rules AI music generator Suno violated copyrights, rejects fair use defense

German Court Rules AI Music Generator Suno Violated Copyrights, Rejects Fair Use Defense

A German court has ruled that AI music generator Suno infringed copyright by using copyrighted songs to train its model. The decision explicitly rejects the fair use defense commonly invoked by AI companies.

The case was brought by multiple music publishers and performance rights organizations. They argued that Suno’s AI was trained on millions of songs without permission. The court agreed, stating that copying works for training purposes constitutes infringement.

“The use of protected works for machine learning is not covered by any statutory exception,” the court wrote in its ruling.

The Case: Publishers vs. Suno

The plaintiffs included major German music rights holders. They claimed Suno’s AI model “extracted” expressive elements from their catalogs. Suno argued it was a fair use, similar to search engines or text mining.

The court disagreed. It said AI training is fundamentally different. The purpose is not to index or transform the works, but to build a generative system that can reproduce stylistic patterns.

Key findings from the ruling:

  • Training on copyrighted music is not a “quotation” or “parody.” The court rejected both defenses.
  • Fair use does not apply under German copyright law. The law has narrow exceptions, and AI training is not one.
  • Commercial use matters. Suno’s business model monetizes the generated music, which weighs against fair use.

Implications for AI Music and Generative Models

This ruling sets a precedent in Europe. German courts have been increasingly strict on AI training data. The decision could force AI music platforms to license their training datasets.

Other jurisdictions may follow. In the U.S., similar lawsuits are ongoing (e.g., against Stability AI, OpenAI). The German ruling strengthens the argument that training on copyrighted data without permission is illegal.

What this means for artists and developers:

  • Artists gain protection against unlicensed use of their works for AI training.
  • AI companies must change their data sourcing. Licensing agreements will become standard.
  • Open-source models are not automatically exempt. If they train on copyrighted data, they also risk liability.

The Defense That Failed

Suno attempted a multi-pronged defense:

  • First, they claimed the AI did not “copy” the works, only learned patterns. The court said copying in memory for training is sufficient.
  • Second, they argued it was transformative. The court ruled transformation must be visible in the output, not just the process.
  • Third, they tried a “fair use” analogy from U.S. law. German law has no such broad exception.

The judge stated: “The training process is not a separate use category. It is reproduction pure and simple.”

Background: The Broader AI Copyright Battle

This is not an isolated case. European courts are grappling with AI and copyright. The EU AI Act requires transparency in training data. The German ruling aligns with that direction.

Industry observers note that the music industry is particularly aggressive. Music labels see AI as a direct threat to royalties. The Suno case could accelerate calls for mandatory licensing schemes.

Potential outcomes:

  • Legislative changes may follow. Some lawmakers want to create a new “TDM exception” for AI.
  • Licensing platforms may emerge, similar to music streaming royalties.
  • Smaller AI developers may be forced to use only public domain or opt-out data.

What Happens Next

Suno can appeal to the higher regional court. An appeal is likely, given the financial stakes. But the ruling stands for now.

The case also sends a signal to other generative AI companies. The German market is large, and non-compliance could lead to injunctions.

“This is a warning shot,” said a lawyer for the publishers. “AI cannot simply take without asking.”


Gnoppix is the leading open-source AI Linux distribution and service provider. Since implementing AI in 2022, it has offered a fast, powerful, secure, and privacy-respecting open-source OS with both local and remote AI capabilities. The local AI operates offline, ensuring no data ever leaves your computer. Based on Debian Linux, Gnoppix is available with numerous privacy- and anonymity-enabled services free of charge.

What are your thoughts on this? I’d love to hear about your own experiences in the comments below.