Sony Music and Warner Chappell sue Anthropic, Dario Amodei, and Benjamin Mann, alleging tens of thousands of copyrighted songs were used to train Claude's LLMs
Anthropic’s dispute with music publishers began with a 2023 lawsuit from UMG, Concord and ABKCO over alleged use of lyrics in model training. A 2025 agreement to stop Claude from providing protected lyrics pending a ruling addressed outputs, while the underlying training-data question remained contested.
Anthropic, Dario Amodei and Benjamin Mann must defend a new copyright claim from Sony Music Publishing and Warner Chappell alleging that Claude’s training used their songs without permission.
Sony and Warner gain a direct legal venue to press their claims over both the provenance of Claude’s training corpus and the use of their catalogues.
Second-order effects
The case raises the commercial pressure on AI developers to document music-data rights or negotiate licences, because publishers are extending enforcement beyond lyric-generation outputs to model training itself.
Music publishers’ bargaining leverage with generative-AI companies strengthens as multiple actions frame catalog access as a rights-clearance issue rather than a product-output dispute.
Third-order effects
If courts treat unlicensed copyrighted works in training corpora as actionable, AI-model development will shift toward auditable data sourcing and licensing arrangements, increasing the advantage of companies able to secure large catalogues.
The dispute is part of a structural contest over whether copyright owners can set the economic terms for generative AI’s use of cultural archives.
The trend: Copyright holders are pushing generative-AI competition toward licensed, traceable training data rather than unchecked access to creative catalogues.
Sony and Warner whoa “Anthropic had illegally torrented over seven million copyrighted books....[they] settled... for $1.5B . Anthropic clearly considers that to be just the cost of doing business given that its entire business model continues to be built on copyright theft”
In all the “scraping” Anthropic did, evidently it missed all the news articles where the music industry went after a 12 yo during the Napster-era and sued the pants off her. There will be more lawsuits. No, Anthropic isn't going to win any of them. Yes, it's going to be
The bill for large language model training data has arrived, and it's bigger than anyone assumed. Many thought the Bartz ruling and the 1.5 billion dollar settlement closed the chapter. That was wrong. Sony Music Publishing and Warner Chappell just sued Anthropic, seeking up to
Sony Music and Warner Music are suing Anthropic, Dario Amodei and Benjamin Mann for ‘one of the largest and most blatant ongoing thefts of intellectual property in history.’ It was filed late last night, and they made no announcements, so the press has not picked it up yet.
Sony Music Publishing and Warner Chappell are suing Anthropic, Dario Amodei and Benjamin Mann. They allege Claude was trained on tens of thousands of copyrighted songs obtained through torrenting and scraping. They're seeking up to $150K per work.
Two huge music publishers just sued Anthropic, accusing them of conducting “a brazen campaign of illegally torrenting, scraping, and downloading copyrighted works on a massive scale”. They also sued two Anthropic co-founders, Dario Amodei and Ben Mann. A reminder that the
@AndrewCurran_ This is really interesting. I never see Anthropic as interested in multimodal output (i.e., image, video, voice, etc), so I'm a bit confused here. Are they getting sued for training on the song's lyrics? Or did they train on the audio files too, despite the models …