The Extended Brief

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

Brief by The AI News AI newsroom · Aug 1, 2026, 7:11 AM EDT edition

Original reporting by The Decoder — Matthias Bastian · published Aug 1, 2026, 6:40 AM EDT

Updated Aug 1, 2026, 11:28 AM EDT

AI music companies can no longer assume training on copyrighted songs is legally safe in Germany after a Munich court ruled against Suno.

Key points

• A Munich court ruled that Suno violated copyrights through both its model training and its generated output. source ↗
• The court found six copyrighted songs were reproducibly stored inside Suno's models. source ↗
• Judges rejected both Germany's text-and-data-mining exception and the US fair use defense. source ↗
• The ruling is not final, and several key legal questions remain open. source ↗

Practical applications

• Reassess legal exposure for generative-audio products trained on copyrighted music, particularly for German and wider EU operations.
• Test whether your own generative models reproducibly store training works, since memorization of six songs was central to the finding against Suno.
• Review reliance on text-and-data-mining exceptions in EU training-data strategy, as the Munich court rejected that defense.
• Track this case before signing licensing or indemnification terms with AI music vendors, since the ruling is not yet final.

Context

AI music generators like Suno are trained on large collections of recorded music, and whether that training requires rights-holder permission is a central open question in AI law. A Munich court ruled Suno infringed copyright in both training and output, finding six copyrighted songs reproducibly stored inside its models — undercutting the argument that models only learn abstract patterns — and rejecting both Germany's text-and-data-mining exception and the US fair use defense. The ruling is not final, but it removes the presumption of legal safety for music-model training in Germany.

What to watch

• The appeal outcome, since the ruling is not final and several key legal questions remain open.
• Whether courts in other EU jurisdictions follow the Munich reasoning on the text-and-data-mining exception.

Related briefs

• Have it both ways: stay discoverable in search while disallowing AI training
• Houthis used Claude Code to develop missile guidance software: Anthropic
• Anthropic Makes $13.7 Compute Deal With Trump-Linked Rum Group
• OpenAI agents attacked RubyGems back in May

This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.