How Did Anthropic Use the Songs in Its AI Training?
Major music publishers Sony Music Publishing and Warner Chappell have filed a multibillion-dollar lawsuit against Anthropic, the AI startup behind the Claude chatbot, accusing it of illegally using tens of thousands of copyrighted songs to train its artificial intelligence models without permission or compensation. The legal action, filed in a U. S. federal court, claims the company engaged in widespread copyright infringement by scraping lyrics and musical compositions from publishers’ catalogs to develop its generative AI systems. The publishers argue that Anthropic’s actions undermine the rights of songwriters and composers whose works were used without authorization.
The lawsuit centers on the allegation that Anthropic copied vast quantities of song lyrics and metadata from copyrighted works to train its large language models, enabling the AI to generate or reproduce protected content. Sony Music Publishing and Warner Chappell state that the startup did not obtain licenses for the use of these materials, despite the clear commercial value of the data in improving AI performance. The publishers emphasize that songwriters rely on licensing revenue, and unauthorized AI training threatens their livelihoods. They are seeking damages that could reach into the billions, reflecting the scale of the alleged infringement.
What Are the Publishers Asking For in the Lawsuit?
According to the complaint, Anthropic’s models were trained on datasets that included lyrics extracted from music publisher databases, allowing the AI to recognize patterns in songwriting and generate text resembling copyrighted lyrics. The publishers claim this was done systematically and at scale, forming a core part of the AI’s training process. They argue that Anthropic benefited from this use without providing compensation, constituting a violation of copyright law. The lawsuit also alleges that the company ignored takedown requests and continued to use the material despite knowing its protected status.
Sony Music Publishing and Warner Chappell are seeking both actual damages and statutory damages for each instance of alleged infringement, which could amount to significant sums given the number of works involved. They are also requesting an injunction to prevent Anthropic from further using their copyrighted materials in AI training. The publishers say the case is not just about compensation but about setting a precedent for how AI companies must respect intellectual property rights. They hope the lawsuit will encourage AI developers to seek proper licenses before using creative works.
Why are music publishers suing Anthropic over AI training? The publishers allege that Anthropic used tens of thousands of copyrighted songs without permission to train its AI models, violating copyright law and depriving songwriters of rightful compensation.
Frequently Asked Questions
What kind of damages are the publishers seeking? They are seeking both actual and statutory damages for each infringed work, which could total billions of dollars, along with a court order to stop further unauthorized use of their catalogs.
Could this lawsuit affect how other AI companies train their models? Yes, a ruling against Anthropic could set a legal precedent requiring AI firms to obtain licenses for copyrighted content used in training, potentially changing industry practices across the sector.