“The complaint noted that despite Anthropic’s purported adoption of so-called guardrails, the models still generate infringing lyrics and can be easily jailbroken by users to output copyrighted content.”
Concord Music Group, Inc., Universal Music Group, and ABKCO Music, Inc. filed a complaint on Wednesday for copyright and Digital Millennium Copyright Act (DMCA) violations against Anthropic PBC, Dario Amodei, and Benjamin Mann in the U.S. District Court for the Northern District of California, adding another lawsuit against generative artificial intelligence companies.
The publishers alleged that Anthropic engaged in mass piracy by downloading millions of unauthorized copies of books containing their copyrighted musical compositions from notorious pirate library websites, including Library Genesis (LibGen) and Pirate Library Mirror (PiLiMi). The complaint argued that Anthropic used BitTorrent to acquire these works and subsequently trained its Claude AI models on the stolen content, thereby directly infringing the publishers’ exclusive rights and undermining the music licensing market.
According to the filing, evidence revealed in a separate case, Bartz v. Anthropic PBC, showed that Benjamin Mann personally engaged in illegal torrenting activities. The filing further alleged that Dario Amodei personally discussed and authorized this conduct. Concord, Universal, and ABKCO asserted that Anthropic concealed these torrenting violations during discovery in an earlier case, Concord Music Group, Inc., et al. v. Anthropic PBC (Concord I), which addressed the exploitation of 499 musical compositions in Claude AI models.
The publishers identified 714 works in Exhibit A of the complaint, including well-known compositions such as “Wild Horses,” “Sweet Caroline,” “Bennie and the Jets,” and “Eye of the Tiger.” An additional 20,517 works were identified in Exhibit B, and the complaint stated that Anthropic unlawfully torrented these works to build a vast central library of written texts. When Anthropic used BitTorrent to download the works, the company simultaneously uploaded unauthorized copies to the public at large, separately violating the exclusive right of distribution and encouraging further infringement.
The publishers alleged that Anthropic’s new AI models, including Claude 4.5 Sonnet, Claude 4.5 Haiku, and Claude 4.5 Opus, were trained to memorize the publishers’ works. As a result, these models generate outputs that copy the publishers’ lyrics in various ways. The complaint noted that despite Anthropic’s purported adoption of so-called guardrails, the models still generate infringing lyrics and can be easily jailbroken by users to output copyrighted content. The publishers argued that any claimed use for AI training was irrelevant and did not qualify as fair use because the act of torrenting constituted a standalone act of unmistakable infringement.
In addition to copyright infringement claims, the complaint included a count for the removal or alteration of Copyright Management Information (CMI) under the DMCA. Concord, Universal, and ABKCO alleged that Anthropic intentionally removed or altered CMI from their works during the training process. They claimed that Anthropic applied algorithms known to remove copyright notices and other identifying information from the copied text. The filing stated that Anthropic did so by design because its objective was for the models to reproduce expressive content while treating CMI as useless junk.
The publishers are pursuing relief on multiple counts, including direct copyright infringement, contributory and vicarious copyright infringement, and DMCA violations. Concord, Universal, and ABKCO are seeking statutory damages under 17 U.S.C. § 504(c) up to $150,000 per work infringed for willful violations. For the CMI violations, they are seeking statutory damages up to $25,000 per violation pursuant to 17 U.S.C. § 1203(c)(3)(B). Alternatively, the publishers are seeking their actual damages and Anthropic’s profits from the infringement. The requested relief also includes a permanent injunction, an order requiring Anthropic to provide an accounting of its training data and methods, and an order requiring the destruction of all infringing copies.
The complaint emphasized that Anthropic’s conduct had far-reaching consequences for the music licensing market. By taking the works without authorization, Anthropic undercut the entire market and lowered the overall value of the compositions. The publishers recognized the potential of ethical AI and stated that they had already begun exploring licenses for authorized uses of their musical compositions. However, they maintained that AI technology must be developed responsibly protecting the rights of songwriters and the creative ecosystem.
Concord, Universal, and ABKCO argued that Anthropic had the means to take simple steps to avoid contributing to infringement but failed to do so. Instead, Anthropic gathered training data that included the publishers’ copyrighted works and provided the site and facilities necessary for users of its AI models to commit direct infringement. The filing concluded that “Anthropic has created a tool, trained on unauthorized copies of Publishers’ works, that permits users to generate vast quantities of AI-generated lyrics and songs that compete with Publishers’ legitimate copyrighted works and harm the market for and value of those works.”
The action is the latest in a series of more than 50 copyright-related suits filed in the United States against AI developers. In September 2025, Anthropic agreed to pay at least $1.5 billion to settle Bartz v. Anthropic, in which authors alleged the AI company trained its models on hundreds of thousands of pirated books. In October 2025, Reddit filed a suit against Perplexity AI for allegedly bypassing its security measures to scrape user-generated content. In December 2025, The New York Times Company filed a complaint against Perplexity AI for copyright and trademark infringement related to the unauthorized use of millions of its articles.
Image Source: Deposit Photos
Image ID: 771714100
Author: alexeynovikov
‘ The preceding article may include information circulated by third parties ’
‘ Some details of this article were extracted from the following source ipwatchdog.com ’
