OpenAI tried to blame users but court didn’t buy it
A court in Munich ruled that ChatGPT violated German copyright laws. OpenAI used song lyrics of popular musicians to train its large language models. And creative industry advocates called this decision landmark for Europe.
The Munich Regional Court sided with the German society for protecting musical rights GEMA. Which reported that ChatGPT collected copyright-protected song lyrics to “learn” from them. GEMA manages the rights of composers, lyricists and music labels. The society includes about 100 thousand artists.
The judge ordered OpenAI to pay compensation for unauthorized use of copyright-protected materials.
And now the most interesting. OpenAI tried to defend itself with a strange argument, quote: “Since the output data is generated by ChatGPT users through queries, the latter should bear legal responsibility for this”. The company stated that not they but the users are to blame. However, the court rejected this argument.
The law firm Raue, representing GEMA’s interests, pointed out that the court decision “creates an important precedent for protecting creative works and sends a clear signal to the global technology industry”.
It turns out OpenAI trained models on others’ content without permission, and when caught, tried to shift responsibility onto users. The German court didn’t buy it. And now this decision could become a model for courts around the world.