A German court has ruled that OpenAI violated copyright law through its ChatGPT chatbot program. This time, OpenAI’s generative AI services were found to have illegally reproduced the lyrics of various popular German artists within its content.
The judge determined that OpenAI infringed upon several copyright laws and must pay an undisclosed amount in damages to the affected copyright holders. This outcome was reported by Reuters following a ruling earlier this week.
The plaintiff in the case was the German music rights society GEMA, which represented a number of musicians whose works were used without permission to train OpenAI’s ChatGPT language models. Among the artists impacted was best-selling musician Herbert Groenemeyer. His singles “Maenner” and “Bochum” were reportedly among nine songs included in the lawsuit.
Judge Elke Schwager ultimately ruled in favor of GEMA, ordering OpenAI to pay damages for using copyrighted material without authorization. However, the final amount of the damages has not been disclosed.
This case marks yet another legal challenge for OpenAI related to its generative AI services. The company’s Sora 2 video generation tool has faced several complaints and threats of lawsuits from organizations such as the Motion Picture Association, SAG-AFTRA, and Japan’s CODA group, among others.
In its defense during this recent case, OpenAI argued that its language models did not store training data directly but instead sometimes produced outputs that reflected similar material seen during training. However, the court found that the outputs were too similar to the copyrighted works and constituted infringement.
This ruling underscores ongoing legal complexities surrounding the use of copyrighted content in training AI models and sets a precedent for how generative AI services must navigate intellectual property rights moving forward.
https://www.shacknews.com/article/146798/openai-chatgpt-german-music-copyright-lawsuit
