Authentic Brands Group, the licensing organization controlling Elvis Presley's music catalogue and likeness, has filed a copyright infringement lawsuit against Anthropic, the AI company behind Claude. The suit alleges Anthropic used 274 Presley songs — including Can't Help Falling in Love, All Shook Up, and A Little Less Conversation — to train its models without permission. The plaintiffs are seeking $150,000 per infringed track, totaling approximately $41 million. The structural significance is not the dollar amount but the plaintiff. While AI firms have already faced copyright claims from the RIAA and major labels including Sony, Universal, and Warner Records, this case represents one of the first major legal challenges brought directly by an artist's estate against an AI developer. Attorney Richard Busch frames it in maximalist terms: Anthropic "has built its entire business model based on copyright infringement at a scale never seen before." The lawsuit goes beyond training data. It claims that "from its inception, Anthropic's AI model, 'Claude,' has created infringing output of copyrighted material of virtually every song from every catalog available on the internet." It further alleges that Anthropic's post-hoc guardrails were ineffective and that the company knew they were ineffective, establishing a theory of willful infringement — a distinction that matters enormously for statutory damages. This is the extraction question at the heart of AI development: who captures the value of decades of creative work when it becomes training data? The music industry's position is clear — ingestion without licensing is theft. AI companies have argued that training on publicly available data constitutes fair use, a transformative process akin to a human learning from existing works. No court has yet issued a definitive ruling on this question at scale. The broader context is an industry in defensive mode. Spotify announced last week it had removed approximately 50 million songs considered AI slop from its recommendations over the past year. Musicians have repeatedly expressed concerns that their voices and compositions are being used without permission, and that AI-generated content could displace human-made music on streaming platforms. Some AI bots have begun impersonating real living artists. The case's trajectory matters more than its specifics. If the estate prevails on the willful infringement theory, it creates a template for every rights-holder sitting on a valuable catalogue. If Anthropic successfully argues fair use for training, it effectively legalizes the ingestion of the entire internet's creative output. The middle ground — compulsory licensing regimes, negotiated rates — remains the most likely long-term equilibrium, but it requires either legislation or a string of settlements that establish market norms. For now, the lawsuit is a price-discovery mechanism dressed as a legal complaint. The $41 million figure is modest relative to Anthropic's valuation but significant as a per-song precedent. Multiply $150,000 by every copyrighted song on the internet, and the implied liability is existential for any AI company that trained on unlicensed data — which is all of them.
