ChatGPT accused of copyright infringement on German songs: precedent set

A court in Munich ruled that the developers of the chatbot OpenAI ChatGPT infringed copyright by using the lyrics of nine German songs without permission. Expert assessment of the language model revealed that it copied fragments of songs, including compositions by Herbert Grönemeier "Maenner" and "Bochum". The lawsuit was filed by GEMA, Germany’s rights agency. The case has attracted international attention as it could set a precedent for similar lawsuits around the world.

The main problem is that developers of generative AI, learning from copyrighted data, do not provide compensation to copyright holders. OpenAI claimed that its models did not copy data, but only “reflected” learning. However, the Munich court found such an explanation insufficient, stressing that memorizing and reproducing lyrics is a copyright infringement.

GEMA CEO Tobias Holzmüller said that “the Internet is not a self-service store, and human creativity is not free templates,” and that the lawsuit makes it mandatory for AI operators to comply with copyright laws. Despite OpenAI's opposition to the court's decision, the company continues to explore further steps, stressing that the ruling only applies to a limited number of songs.