On Tuesday, A Munich court ruled in favor of Germany’s music rights organization, GEMA, in a significant copyright case against the U.S.-based AI company OpenAI. The court determined that OpenAI cannot use song lyrics without obtaining a proper license and ordered the company to pay damages for using copyrighted material without authorization.

GEMA argued that OpenAI’s chatbot, ChatGPT, reproduces lyrics from copyrighted German songs without permission. The organization also claimed that ChatGPT was trained on protected works from GEMA’s repertoire, which includes approximately 100,000 members such as best-selling musician Herbert Groenemeyer. According to GEMA, this use of copyrighted content constitutes a violation of authors’ rights, as the AI reproduces lyrics without securing proper authorization from the rights holders.

OpenAI responded by stating that GEMA’s arguments reflect a misunderstanding of how ChatGPT operates. The company maintained that the AI does not directly copy content but generates responses based on patterns learned during training. Despite this defense, the Munich court sided with GEMA, reinforcing the importance of copyright compliance in the development and deployment of generative AI technologies.

The ruling is seen as a potentially landmark decision for the regulation of generative AI in Europe. GEMA is advocating for a licensing framework that would require AI developers to pay for the use of musical works both during AI training and in the output generated by the system. The decision can still be appealed, and both OpenAI and GEMA said they plan to release statements regarding the verdict later on Tuesday. This case highlights ongoing tensions between AI innovation and copyright protection, raising questions about how creative works can be used in AI training while respecting intellectual property rights.