Reservoir sues Suno and Anthropic over AI music copyright

The scope of the lawsuits

Independent music publisher Reservoir Media has initiated two major copyright infringement lawsuits against the AI companies Suno and Anthropic. The legal actions, filed in the US District Court for the Northern District of California, target the unauthorized use of copyrighted material to train generative models. Reservoir is seeking a combined maximum of approximately $338 million in statutory damages, along with injunctions to stop the practices.

The lawsuit against Suno specifically identifies 1,050 musical compositions that the publisher alleges were copied without permission. Reservoir claims that Suno obtained this music from various online sources to build the datasets used for its music-generation platform. The company also argues that Suno's technology can be used to generate new music that derives from these protected works, even when users try to avoid imitating specific artists.

In a separate action, Reservoir is suing Anthropic over the use of 853 sets of copyrighted song lyrics to train its general-purpose AI assistant, Claude. The publisher alleges that the AI can reproduce these lyrics or create unauthorized derivative versions when prompted by users. These works include famous tracks by artists such as Joni Mitchell, Johnny Cash, Madonna, and Evanescence.

How the data was collected

Reservoir alleges that Suno employed automated scraping methods to gather content from several major platforms, including YouTube Music, Deezer, and Genius. The publisher points to publicly reported information from leaked Suno source code as evidence of how these training datasets were structured. These allegations suggest a systematic approach to gathering large amounts of music data without seeking individual licenses from rights holders.

The legal filings also highlight a significant distinction between the types of AI technology being challenged. Suno is a generative music platform that creates songs from text prompts, while Anthropic develops Claude, which is primarily designed for generating written content. Despite their different functions, Reservoir argues that both companies have benefited commercially from copyrighted works without providing proper compensation to the original creators.

The lawsuits also include claims under the Digital Millennium Copyright Act. These claims suggest that the AI companies may have circumvented technological access controls or removed copyright management information during the data collection process. If successful, Reservoir could obtain court orders requiring the removal or destruction of these unauthorized copies from internal repositories and model infrastructure.

The licensing landscape

One of the most notable aspects of the Suno lawsuit involves the company's existing licensing relationships with major music businesses. Suno has already reached agreements with entities like Warner Music Group, BMG, and Believe to use music in its services. Reservoir argues that these existing deals prove a viable commercial market exists for licensing music to AI developers, which supports their claim for compensation.

However, the existence of new licenses does not automatically mean that previous training practices were legal. The court will have to determine if the music used to build the initial models was taken without permission before these agreements were signed. This distinction is critical for AI developers who are trying to navigate the transition from unregulated data scraping to formal licensing structures.

Furthermore, the legal battle highlights the complexity of distinguishing between sound recordings and underlying musical compositions. A license that covers a specific recording does not necessarily grant the rights to the underlying composition, which includes the melody and lyrics. For AI companies, this creates a significant hurdle when trying to negotiate rights across massive and diverse music catalogues.

Implications for creators

For independent musicians and songwriters, these lawsuits center on whether the commercial value of AI technology will lead to fair recognition and pay. While generative AI can create music at a massive scale, it also introduces intense competition for creators who rely on their copyrights for income. The outcome of these cases may help establish the rules for how these new tools can coexist with human creativity.

The litigation underscores the growing importance of music publishing rights, which cover the underlying songwriting rather than just the final audio recording. Independent creators need to understand how their publishing rights are managed and who holds the authority to license those works. As AI companies seek to build more sophisticated models, transparent data and effective rights administration will become essential for smaller players.

Ultimately, the music industry is currently in a period of transition as it tries to define the economics of AI generation. While some developers are already signing deals with major labels, the fate of independent works remains uncertain. These lawsuits could shape the future of how songwriters and composers participate in the value generated by the AI systems built upon their work.