ElevenLabs launched a new artificial intelligence model on Tuesday, enabling users to generate music and claiming the output is cleared for commercial use. This marks an expansion beyond the company’s established focus on AI audio tools, including text-to-speech products, conversational bots, and speech translation tools.
The company shared samples of its AI-generated music, which included a synthetic voice rapping about personal experiences. Previous concerns around the training data for AI music generation tools have led to legal challenges for other startups.
Last year, the Recording Industry Association of America (RIAA) initiated lawsuits against Suno and Udio, alleging they trained their music-generation models on copyrighted material. Suno and Udio are reportedly engaged in licensing discussions with major record labels.
ElevenLabs has also announced agreements with Merlin Network and Kobalt Music Group, both digital publishing platforms for independent musicians. These agreements permit the use of their materials for AI training.
Merlin’s website indicates representation for artists such as Adele, Nirvana, Mitski, Carly Rae Jepsen, and Phoebe Bridgers. Kobalt represents artists including Beck, Bon Iver, and Childish Gambino.
A Kobalt representative informed TechCrunch that artists must voluntarily opt-in for their music to be licensed for AI use. The representative stated, “Our clients benefit directly from this agreement in several key ways: it opens a new revenue stream in a growing market, includes revenue sharing so they participate in the upside, provides strong safeguards against infringement and misuse, and offers favorable terms comparable to other publishing and recording rightsholders.”
