The German copyright collective GEMA has introduced a new product called PLAI by GEMA. This is a music dataset designed for training artificial intelligence tools, which exclusively contains fully licensed copyrighted content. According to company representatives, it is the first offering of its kind on the European market, bringing together audio files, metadata, and rights to musical works and recordings from a single source.
The service is primarily aimed at AI tool developers that assist in production and do not directly compete with original creations. The first customer is the German company Klangio, which specializes in automatic transcription of music into sheet music. This company now uses a database comprising approximately 178,000 audio files covering over 60 different music genres.
Legal Certainty for Developers and Authors
The main goal of the project is to create a transparent environment for technology companies. GEMA emphasizes that instead of complex negotiations with many rights holders, developers can now obtain legal training data centrally. Several partners participated in the database's development, including Bailer Music Publishing, Earmotion Audio Creation, and Sonoton Music.
“Instead of complex rights clearance, AI tool providers now receive high-quality music data from a single source. This creates legal certainty for developers and new income potential for music creators,” said Thomas Theune, GEMA's Director of Broadcasting and Online.
Sebastian Murgul, head of Klangio, confirms that a clear legal framework is crucial for technology companies. “In GEMA’s dataset, we found the solution we had been waiting for: legally sound training data where all rights are actually covered,” said Murgul, adding that this gives the company confidence for further innovation without the risk of future lawsuits.
Awaiting a Key Ruling in the Suno Case
The launch of the PLAI platform comes at a time when the music industry is eagerly awaiting a verdict in the lawsuit between GEMA and Suno, a company that operates an AI-based music generator. GEMA filed a lawsuit in a Munich court, alleging that Suno trained its models on protected repertoire without permission.
This is not the first time this collective management organization, representing over 100,000 members in Germany, has gone to court. In November 2025, GEMA already won a dispute with OpenAI, where the court confirmed that the ChatGPT chatbot used song lyrics without proper licensing. The organization's current activities show an effort to transform the AI market from a threat into a sustainable licensing model that ensures fair compensation for creators whose works are used in the digital age.