Sony Music, Warner sue Anthropic, alleging a "brazen campaign" of intellectual property theft

The rapidly evolving world of artificial intelligence has once again collided with established intellectual property rights, as two of the music industry's biggest players, Sony Music Publishing and Warner Chappell Music, have launched a significant lawsuit against AI developer Anthropic. The complaint, filed in the U.S. District Court for the Northern District of California on August 28, 2026, accuses Anthropic of a "brazen campaign" of widespread copyright infringement to train its large language models (LLMs), specifically the Claude series.

This lawsuit is particularly broad in its accusations, alleging "one of the largest and most blatant ongoing thefts of intellectual property in history." It highlights the growing tension between AI companies, which require vast amounts of data for training, and content creators seeking to protect their copyrighted works and ensure fair compensation.

The Core of the "Brazen Campaign" Allegations

The lawsuit by Sony Music Publishing and Warner Chappell Music, along with several affiliated catalog companies including EMI, Hipgnosis, and the Jobete/Motown publishing arm, makes specific and severe claims against Anthropic. The publishers assert that Anthropic, along with its co-founders Dario Amodei (CEO) and Benjamin Mann, engaged in systematic illegal activities to amass training data for its Claude AI models.

According to the court documents, Anthropic allegedly used various illicit methods, including:

• Illegal Torrenting: The complaint heavily leans on findings from a previous class-action lawsuit, Bartz v. Anthropic, where Anthropic settled for $1.5 billion after being accused of illegally downloading over seven million books from pirate sites like Library Genesis (LibGen) and Pirate Library Mirror (PiLiMi). The music publishers argue that the same torrenting spree that targeted books also swept up their songbooks and sheet music. Specifically, the lawsuit alleges that co-founder Benjamin Mann personally used BitTorrent in June 2021 to download at least five million books from LibGen, and Anthropic employees torrented another two million from PiLiLi in July 2022, reportedly with CEO Dario Amodei's knowledge and approval.
• Scraping Licensed Sites: Anthropic is accused of scraping lyrics directly from licensed sites such as MusixMatch and LyricFind, in violation of their terms of service.
• Data Extraction from Datasets: The complaint also claims Anthropic pulled lyrics from large datasets like Common Crawl and "The Pile," which independent researchers have indicated contained copyrighted books and other materials.
• "Destructive Scanning" Operations: The lawsuit points to an alleged "destructive scanning" operation where Anthropic reportedly bought used songbooks, digitized them, and then discarded the physical copies. An internal planning document from 2024 for this project reportedly stated, "We don't want it to be known that we are working on this."
• Removal or Alteration of Copyright Management Information (CMI): A critical part of the suit includes a claim that Anthropic's data-cleaning processes stripped song titles, songwriter names, and copyright notices while retaining the lyrics themselves. This constitutes a separate violation under copyright law.

The music publishers are seeking statutory damages of up to $150,000 per willfully infringed work, plus up to $25,000 for each alleged removal of copyright management information. Given the "thousands upon thousands" of copyrighted musical compositions allegedly harvested, the potential damages could amount to billions of dollars.

Anthropic and Claude AI: A Brief Overview

Anthropic, founded in 2021 by former OpenAI executives Dario Amodei and Daniela Amodei, is an American artificial intelligence public benefit corporation headquartered in San Francisco. The company's mission emphasizes a safety-focused approach to AI research and development, positioning itself as prioritizing positive social impact over profit. They are known for developing "Constitutional AI," a framework designed to align AI systems with human values, ensuring they are helpful, harmless, and honest.

Anthropic's flagship product is Claude, a series of proprietary large language models (LLMs) available through an online chatbot and an API. The Claude family includes models like Haiku, Sonnet, Opus, and Fable, with Claude Opus offering the strongest capabilities for complex reasoning and creative tasks. The company states that it aims to build reliable, interpretable, and steerable AI systems. However, the ongoing legal battles challenge the ethical sourcing of the vast amounts of data required to train these powerful models.

A Familiar Battleground: AI Training Data and Copyright

This lawsuit is not an isolated incident but rather part of a growing wave of legal challenges against AI companies concerning the use of copyrighted material for training their models. The underlying issue is fundamental: generative AI models, like Claude, require immense datasets to learn patterns, language, and creativity. Much of this data, especially from the internet, includes copyright-protected works.

The music industry has been particularly vocal. Prior to this lawsuit, other music publishers, including Universal Music Publishing Group, Concord Music Group, and ABKCO, had already filed two pending cases against Anthropic. An October 2023 suit covered around 500 songs, and a January 2026 suit sought over $3 billion in damages for more than 20,000 works. BMG and Round Hill Music also brought separate cases in March and August 2026, respectively. Major labels like Capitol Records and Sony Music Entertainment have also launched copyright infringement lawsuits against generative AI music companies Udio and Suno, accusing them of "willful copyright infringement on an almost unimaginable scale."

The central legal argument often revolves around "fair use." AI companies typically argue that their use of copyrighted material for training is "transformative" – meaning it creates something new and different from the original work, thus qualifying for fair use protection. For instance, in a June 2025 ruling in the Bartz v. Anthropic case concerning books, a judge stated that Anthropic's use of books to train Claude was "exceedingly transformative" and could qualify as fair use. However, the same judge also ruled that Anthropic's copying and storage of over 7 million pirated books in a central library constituted infringement, regardless of the "transformative" argument for training.

This distinction is crucial: while the act of training an AI might, in some contexts, be argued as transformative, the method of acquiring the training data (e.g., through illegal torrenting) is often where the direct infringement claims arise.

Industry Implications and the Road Ahead

The lawsuit by Sony Music and Warner Chappell against Anthropic underscores the high stakes involved in the intersection of AI and intellectual property. The music industry fears that AI-generated songs could flood the market with imitations, reducing income for human artists and songwriters. The outcome of these cases could set significant precedents for how AI models are trained and how content creators are compensated in the future.

Anthropic, which reportedly plans an initial public offering (IPO) in the fall of 2026 and was valued at $965 billion in May 2026, faces substantial financial and reputational risks. The prior $1.5 billion settlement with authors, while significant, was described in the current lawsuit as "just the cost of doing business" for a company with a projected multi-trillion-dollar valuation, and "not large enough to deter infringing conduct."

The legal process will likely be protracted, involving extensive discovery to determine the full scale of alleged infringement. The publishers' decision to name Anthropic's co-founders, Dario Amodei and Benjamin Mann, personally as defendants further emphasizes the gravity of their accusations, alleging direct involvement in the "brazen campaign."

As AI technology continues to advance, the legal framework surrounding its use of copyrighted content will continue to be tested and shaped by these landmark cases. The industry is watching closely to see if AI companies will be compelled to adopt more rigorous and transparent methods for sourcing their training data, potentially leading to new licensing models and compensation structures for creators. Some AI companies are already exploring licensing agreements to use high-quality, legally obtained data.

Frequently Asked Questions

What are Sony Music and Warner Chappell Music accusing Anthropic of?

Sony Music Publishing and Warner Chappell Music accuse Anthropic of a "brazen campaign" of intellectual property theft. They allege that Anthropic illegally torrented, scraped, and downloaded "thousands upon thousands" of copyrighted musical compositions, including lyrics, to train its Claude AI models.

Who are the specific defendants in the lawsuit?

The lawsuit names Anthropic, PBC, as well as its CEO Dario Amodei and co-founder Benjamin Mann, as defendants. The publishers allege that Amodei and Mann personally directed and approved the alleged infringement.

What kind of damages are the music publishers seeking?

The music publishers are seeking statutory damages of up to $150,000 for each willfully infringed work. Additionally, they are seeking up to $25,000 for each instance where Anthropic allegedly removed or altered copyright management information from the works.

Has Anthropic faced similar lawsuits before?

Yes, Anthropic previously settled a class-action lawsuit for $1.5 billion with a group of authors who claimed the company used pirated copies of their books to train its AI models. Other music publishers, including Universal Music Publishing Group, Concord Music Group, ABKCO, BMG, and Round Hill Music, also have pending lawsuits against Anthropic.