GEMA vs. ChatGPT
GEMA's Lawsuit Against OpenAI
GEMA has become the first collecting society in the world to file a landmark lawsuit against OpenAI, the company behind the AI-powered language model ChatGPT. The lawsuit centers on the unlicensed use of copyrighted song lyrics from the repertoire of GEMA’s approximately 95,000 members, including well-known artists such as Reinhard Mey and Kristina Bach. The key question is whether the training and reproduction by generative AI systems constitute a use subject to copyright law. The lawsuit sends a clear message: copyrights must be respected even in the age of AI.
What happened?
Upon request, ChatGPT provides song lyrics that were apparently part of the AI’s training data. However, OpenAI has neither obtained licenses nor compensated the copyright holders, as required by Section 19a of the German Copyright Act (UrhG) for making protected works publicly available. While other providers pay license fees, OpenAI, in GEMA’s view, systematically operates without the necessary legal basis. A lawsuit has therefore been filed with the Munich Regional Court to put a stop to these practices and to establish a precedent.
Who is GEMA?
GEMA (Society for Musical Performance and Mechanical Reproduction Rights) is one of the largest collecting societies in the world. Its primary mission is to manage and enforce the copyrights of music creators. It represents approximately 95,000 members, including composers, lyricists, and music publishers, as well as over two million rights holders from around the world.
GEMA ensures that its members are compensated for their creative works when they are performed, broadcast, reproduced, and, more recently, used by AI systems. This is achieved through licensing models that legally safeguard the use of copyrighted works and ensure fair compensation for creators. With the introduction of a special licensing model for generative AI, GEMA has taken an important step toward safeguarding the rights and interests of its members in this area as well.
Is this use covered by text and data mining?
Under Section 44b of the German Copyright Act (UrhG), the text and data mining exception does indeed permit the automated evaluation of copyright-protected content for analytical purposes. Whether this provision also applies to the training of generative AI models remains a matter of debate. Recent court rulings, such as that by the Hamburg Regional Court, suggest that this is generally possible—but only if the copyright holders have not expressly reserved the right to use the content. GEMA has clearly communicated such usage restrictions on behalf of its members, which could play a decisive role in the proceedings.
Economic Impact of the Lawsuit?
The lawsuit has far-reaching economic implications for both the AI industry and the creative sector. OpenAI, with annual revenue of over $2 billion, could face a significant financial burden due to license fees and potential claims for damages. At the same time, the lawsuit signals that the use of copyrighted content without compensation for creators is not sustainable in the long term.
For the music industry, this approach offers an opportunity to establish fair revenue-sharing models for AI applications. GEMA has already developed a licensing model designed to ensure fair compensation for music creators when AI is used to train models and generate new content. This could also compel other AI providers to enter into similar agreements.
GEMA’s lawsuit against OpenAI marks a turning point in the tension between technological innovation and copyright protection. While courts must determine whether the exceptions to copyright law apply, GEMA makes it clear: Artists’ creative input is not a free raw material for AI models. The ruling could not only redefine how generative AI is handled but also have a lasting impact on the economic framework for the entire creative industry.
Please contact us if you have any questions
