In a groundbreaking decision, a Munich court has ruled that OpenAI violated German copyright law by using song lyrics to train its artificial intelligence models without proper authorization. The lawsuit, filed by GEMA, Germany’s leading music rights organization, marks one of the most significant legal challenges in Europe regarding how AI systems use creative content to build their capabilities.
According to the court’s ruling, “both the memorization in the language models and the reproduction of song lyrics in chatbot outputs constitute infringements of copyright law.” The decision highlights a growing global debate over how AI training data is sourced and whether such use falls under fair use or direct copyright infringement.
GEMA, which represents more than 100,000 composers, songwriters, and music publishers, brought the case on behalf of the creators of nine German songs. The organization accused OpenAI of systematically using its repertoire to train large language models like ChatGPT without purchasing licenses or paying royalties to artists. The lawsuit, filed in November 2024, emphasized that OpenAI’s AI systems had reproduced protected works during both training and output stages.
OpenAI defended itself by claiming that its models do not store or replicate specific copyrighted materials, but rather generate responses based on learned patterns. The company also argued that the users of its chatbot are the ones responsible for the content generated, as they prompt the system to produce specific outputs. However, the Munich court rejected these arguments, ruling that OpenAI’s use of copyrighted lyrics during both training and reproduction constitutes unauthorized use under German copyright law.
The court’s decision entitles the plaintiffs to financial compensation, setting a precedent that could reshape the legal landscape for AI training and creative rights across Europe. Experts suggest that this ruling could lead to new licensing models for AI developers, ensuring that musicians, writers, and other creators receive payment when their work is used to train machine learning systems.
GEMA spokesperson Kai Welp called the verdict “a vital step toward fair remuneration for creative artists.” He emphasized that this case will likely influence future legislation concerning AI transparency, data usage, and copyright compliance.
Conclusion:
The Munich ruling represents a turning point for the intersection of artificial intelligence and intellectual property law. As AI continues to evolve, courts worldwide will face increasing pressure to define where innovation ends and infringement begins. For now, the German verdict signals a clear message: AI developers must respect creative ownership or face legal consequences.
