In brief
- Warner Music Group settled its copyright infringement lawsuit against AI music generator Udio, announced on Wednesday.
- Udio will transition from its current model to a licensed platform where users can create music using participating artists’ voices and compositions.
- The settlement includes Warner’s recorded music and publishing divisions, with Udio rolling out artist protections ahead of a 2026 launch.
Warner Music Group has resolved its copyright infringement lawsuit against AI music generator Udio, announcing an agreement Wednesday that will convert the latter into a licensed service launching in 2026.
The agreement resolves a lawsuit filed last June, when Warner joined Sony Music Entertainment and UMG Recordings in accusing Udio and competitor platform Suno of mass copyright infringement for allegedly training its AI models on copyrighted recordings without permission.
Under the settlement, Udio will abandon its current model and develop a subscription platform where users can create remixes, covers, and new songs using voices and compositions from participating artists and songwriters, all with proper licensing, credits, and payment, according to the latest statement.
“This partnership is a crucial step towards realizing a future in which technology amplifies creativity and unlocks new opportunities for artists and songwriters,” Andrew Sanchez, Udio’s co-founder and CEO said in the statement.
The agreement spans Warner’s recorded music and publishing divisions, creating new revenue streams while ensuring artist protections.
Udio will also roll out expanded protections and safeguards for artists and songwriters ahead of the 2026 launch, while maintaining access to its current system during the transition.
Artists and songwriters who choose to participate will receive credit and payment when users create content using their work, according to the statement.
Copyright claims
The lawsuit, filed by the Recording Industry Association of America and a coalition of major labels, accused Udio and Suno of copying “decades worth of the world’s most popular sound recordings” to train their models.
The lawsuit initially sought $150,000 per infringement incident and injunctions barring future violations.
“They have both taken steps to hide the scope of their widescale infringement, which we expect to uncover in the litigations. But we know more than just that the snippets ‘sound like’ copyrighted recordings,” an RIAA spokesperson told Decrypt at the time.
Regulatory battle
The settlement comes as AI music generation faces mounting legal and regulatory pressure.
A German court recently ruled against OpenAI for reproducing copyrighted song lyrics, marking the first time a European court found a large language model violated copyright law.
The court held that GPT-4 and GPT-4o contained reproducible lyrics from nine songs, constituting unauthorized reproduction under EU and German copyright law.
Generally Intelligent Newsletter
A weekly AI journey narrated by Gen, a generative AI model.
