AI COPYRIGHT
Germany just challenged how AI learns
A Munich court has ruled that ChatGPT broke German copyright law by training on song lyrics without permission.
GEMA, Germany’s main music rights group, brought the case after saying OpenAI used parts of nine well-known German songs to train its models.
The court agreed and ordered OpenAI to pay damages.
GEMA now hopes this will lead to talks on how creators should be compensated when AI tools learn from their work.
The judge also rejected OpenAI’s argument that users, not the company, should be responsible for what ChatGPT generates.
Three quick points:
A German court ruled that ChatGPT used protected lyrics during training without permission.
GEMA calls it a major step for copyright in the AI era.
OpenAI disagrees with the ruling and may appeal.
When lyrics meet AI
OpenAI says its models learn from large datasets and do not store specific songs.
The company is considering an appeal and stressed that the ruling covers only a small set of lyrics, not everyday use of ChatGPT in Germany.
Creative groups in Germany say this is an important early decision for Europe and could shape how AI companies use copyrighted work in the future.
OpenAI is already dealing with similar cases in the US and says it is speaking with rights holders worldwide.
Das ist nicht gut! - MV


