Reservoir Media sues Anthropic and Suno over music copyright

The legal filings

Reservoir Media Management, a major independent music publisher, has taken legal action against two prominent AI companies. On October 21, 2026, the company filed two separate federal copyright infringement lawsuits in the Northern District of California. One suit targets the music generation platform Suno, while the other targets the AI company Anthropic.

The lawsuits are related but distinct in their focus and the specific works they cover. The Suno complaint involves 1,050 musical compositions, whereas the Anthropic complaint focuses on the lyrics of 853 compositions. Collectively, the potential statutory damages from these two cases could reach nearly $340 million.

Reservoir Media Management is the plaintiff in both cases, along with its UK subsidiary and Shapiro, Bernstein & Co. Limited. The company owns a vast catalog of music spanning many genres, including works by Joni Mitchell and John Denver. They are seeking to address how their copyrighted material was allegedly used to build commercial AI products.

The allegations against Suno involve the unauthorized reproduction and distribution of musical compositions to train its generative models. Reservoir Media claims that Suno created unauthorized derivative works without obtaining any licenses or permissions. The complaint further accuses Suno of removing copyright management information and circumventing digital access controls.

To support these claims, the lawsuit cites evidence found in Suno's own leaked source code. This analysis suggests that multiple Reservoir-owned compositions appear directly in the data used for training. These include well-known tracks such as "Both Sides Now" by Joni Mitchell and "Ring of Fire" as performed by Johnny Cash.

The complaint also points to a second internal file that contained references to other catalog titles. These include works by artists like Big & Rich and Kings of Leon, as well as more recordings by Joni Mitchell. These findings suggest a systematic use of the publisher's catalog to develop Suno's music generation capabilities.

Lyrics and the Claude models

The lawsuit against Anthropic takes a different approach by focusing specifically on song lyrics rather than the musical recordings themselves. Reservoir Media accuses Anthropic of copying the lyrics of 853 compositions to train its Claude family of models. The filing argues that the company used these lyrics as a shortcut to avoid the expense of training its models from scratch.

This case highlights a specific legal strategy regarding the text-based components of music. By focusing on lyrics, the publisher is targeting the way large language models process and learn from human-written poetry and prose. The complaint characterizes this conduct as a direct violation of copyright protections for the written words of the songs.

Anthropic responded to these specific allegations in a statement shared with Billboard. The company stated that it disagrees with the claims and intends to defend the case vigorously. This response marks a formal stance in a growing series of disputes over how AI companies utilize creative content to improve their products.

Impacted artists and works

The catalog of works affected by these lawsuits includes some of the most recognizable songs in popular music history. Specifically named works include "Papa Don't Preach" by Madonna and "So What" by Miles Davis. Other mentioned titles include "In the Mood" as recorded by Glenn Miller and His Orchestra and "Lush Life" by Billy Strayhorn.

Joni Mitchell appears most frequently in the Suno complaint's exhibits. Approximately 149 of the 1,050 compositions listed in that suit are credited to her. This represents roughly one in seven of the total works identified in the filing by the publisher.

The lawsuits are not isolated incidents for these AI companies. Suno has previously faced a major copyright lawsuit from the Recording Industry Association of America in 2024. Additionally, Anthropic has faced ongoing litigation from a broader group of music publishers regarding the use of copyrighted lyrics in its models.

The broader legal landscape

The current legal standing of using copyrighted music to train AI models is still being debated. The core of the dispute is whether training an AI model on copyrighted works qualifies as fair use. AI companies typically argue that training is transformative, while rights holders argue that mass reproduction for commercial gain is not protected.

As of October 2026, no US federal court has issued a definitive ruling on this specific issue in the context of music AI. This leaves the question open across dozens of pending cases currently moving through the court system. The Reservoir lawsuits add to more than 80 AI-related copyright cases filed since 2023.

The outcome of these cases will likely determine the foundational legal questions for the entire industry. Courts are currently working through these issues to decide how much of the internet's creative output can be used for training. These rulings will eventually dictate how tech companies and creators interact in the age of generative AI.