Major music publishers have filed a lawsuit against artificial intelligence company Anthropic, seeking more than US$3 billion in damages over what they describe as the “flagrant piracy” of tens of thousands of copyrighted musical works used to train its AI models.
According to sources, the lawsuit, led by Universal Music Group and other leading publishers, significantly expands an earlier legal action that initially accused Anthropic of infringing on around 500 copyrighted works. The updated filing now alleges the unauthorised use of more than 20,000 compositions, dramatically increasing the scale and potential financial exposure of the case.
The publishers argue that Anthropic copied, stored, and used protected lyrics and musical content without permission or compensation, enabling its AI systems to generate outputs that closely resemble original works. They contend that this practice undermines the value of copyrighted music and threatens the livelihoods of songwriters and composers globally.
Court documents indicate that the US$3 billion figure is based on statutory damages under US copyright law, which allows rights holders to seek up to US$150,000 per infringed work in cases of wilful violation. The plaintiffs maintain that Anthropic was fully aware it was using copyrighted material and proceeded without securing licenses or negotiating agreements.
Anthropic has previously defended its practices by arguing that training AI models on large datasets constitutes fair use and is necessary for technological development. However, the publishers counter that AI training is a commercial activity that directly benefits the company and its investors, and therefore requires proper licensing like any other use of copyrighted material.
The case adds to a growing wave of legal challenges facing AI companies from creators across music, publishing, journalism, and visual arts. Rights holders are increasingly pushing back against what they see as unchecked data extraction fueling the rapid growth of generative AI systems.
Legal analysts say the outcome could have far-reaching implications for the AI industry, particularly around how training data is sourced and whether licensing becomes a standard requirement rather than an exception. A ruling in favour of the publishers could reshape the economics of AI development and accelerate negotiations between tech firms and content owners.
As the case moves forward, it is expected to become a key test of how copyright law applies in the age of generative artificial intelligence, with potential ripple effects well beyond the music industry.
Anthropic launches Claude for Healthcare following OpenAI’s ChatGPT Health reveal
