A court in Munich has ruled that OpenAI infringed German copyright law by using song lyrics to train its artificial intelligence models. The court determined that both the memorization of lyrics within the language models and their reproduction in the AI chatbot’s outputs constitute copyright violations.
The case was brought by GEMA, Germany’s music rights organization representing more than 100,000 composers, songwriters, and publishers. The lawsuit, filed in November 2024, concerns nine German songs whose creators allege that OpenAI reproduced their protected lyrics without purchasing licenses or providing compensation. GEMA argued that OpenAI had “systematically” used its repertoire to train AI models.
OpenAI defended itself by claiming that its language models do not store or copy specific data but rather encode learned patterns, and that the AI chatbot’s outputs are generated by users, who are responsible for them.
However, the Munich court rejected these arguments, ruling that the plaintiffs are entitled to compensation both for the reproduction of lyrics within the AI models and for their reproduction in the chatbot’s outputs.
OpenAI has faced similar legal challenges in the United States, with media companies and authors alleging that its ChatGPT system was trained on copyrighted material without permission. The Munich ruling marks the first major European case of its kind and could set a significant precedent for AI training practices and the compensation of creative artists.
“This case could have vital implications for the remuneration of creative artists,” said GEMA representative Kai Welp following a September hearing.
