Photo by Nathan Howard (Bloomberg) / Getty Images
As highlighted by Techcrunch
In Germany, a court has ruled that OpenAI violated copyright law by using licensed musical works to train its language models without permission, according to several outlets, including The Guardian.
The ruling was issued in the lawsuit brought by GEMA, the association responsible for music rights in Germany, filed last November against OpenAI. The company was ordered to pay GEMA compensation, the amount of which was not disclosed. OpenAI says it disagrees with the ruling and is considering further steps, while GEMA called it “the first AI precedent in Europe”.
“Today we have established a precedent that protects and clarifies authors’ rights: even operators of AI tools, such as ChatGPT, must comply with copyright law,”
“Today we successfully protected the livelihoods of music creators.”
Context and further actions
OpenAI faces further lawsuits from other creators and media groups on the same topic.
The company says it is considering possible additional steps, and the ruling in Germany could have implications for other cases of using data to train AI.
This ruling is seen as one of the first major precedents in Europe establishing accountability for using data to train AI.
Useful reading:
- Munich court rules OpenAI must pay damages for using copyrighted song lyrics without license, impacting AI regulation in Europe.
- Japanese trade association CODA urges OpenAI to stop using Studio Ghibli content for AI training without permission, citing copyright concerns and legal requirements under Japanese law.
