The owners of rights to songs recorded by legendary singer Elvis Presley have sued artificial intelligence company Anthropic, alleging that it used copyrighted musical compositions to train its AI models without permission. The lawsuit targets the company’s Claude AI assistant and raises fresh questions about copyright protection, AI training practices and the commercial use of creative works.
The case involves 274 musical compositions associated with Presley and seeks approximately $48.6 million in damages. The plaintiffs allege that Anthropic used copyrighted lyrics without obtaining licences or compensating the relevant rights holders. They also claim that the company’s AI systems can generate outputs that reproduce protected lyrics or closely resemble existing compositions.
The lawsuit adds to the growing number of legal disputes between copyright owners and AI developers over the use of protected material to train generative AI models.
What Does the Lawsuit Claim?
The lawsuit was filed by music rights holders, including entities associated with Authentic Brands Group, which manages significant intellectual property rights connected to Elvis Presley. Other plaintiffs include music publishers and trusts representing interests in copyrighted compositions.
According to the complaint, Anthropic allegedly copied 274 copyrighted musical works without authorisation. The catalogue includes well-known songs recorded by Presley, such as Can’t Help Falling in Love, All Shook Up, Don’t Be Cruel and Return to Sender.
The plaintiffs argue that the alleged infringement extends beyond using copyrighted lyrics as training material. They claim that Claude can generate responses containing protected lyrics and produce adaptations that retain recognisable elements of existing compositions.
The rights holders contend that using their works without permission allows Anthropic to develop and commercialise its AI products while depriving creators and publishers of licensing revenue.
Anthropic has disputed the allegations and indicated that it intends to defend itself against the claims. The allegations have not been established in court.
Why Are the Plaintiffs Seeking $48.6 Million?
The plaintiffs are seeking statutory damages of up to $150,000 for each of the 274 compositions, potentially amounting to $41.1 million. Additional claims concerning alleged copyright management and technological protection violations bring the total potential damages sought to approximately $48.6 million.
Under US copyright law, statutory damages can reach $150,000 per work in cases involving wilful infringement. However, the maximum amount is not automatically awarded. The court would need to assess the evidence, determine whether infringement occurred and establish the appropriate damages.
The plaintiffs are also seeking legal costs and other remedies intended to prevent further alleged infringement.
The financial claim highlights the potential consequences of using copyrighted material without authorisation. For music publishers and other rights holders, licensing agreements provide an established mechanism for allowing commercial use of their works while ensuring compensation.
The lawsuit argues that AI companies should not be permitted to bypass these arrangements simply because copyrighted material is being used to develop new technology.
Concerns About AI-Generated Lyrics
Another significant aspect of the lawsuit concerns the ability of AI systems to reproduce copyrighted lyrics.
The plaintiffs allege that Claude can generate material resembling protected songs, including lyrics and rewritten versions of existing compositions. They also question whether the safeguards implemented by Anthropic are sufficient to prevent users from obtaining copyrighted material through the chatbot.
These claims raise an important distinction between training an AI model on copyrighted works and generating outputs that reproduce protected expression.
AI developers have argued in other copyright disputes that training models on existing material may qualify as fair use under US law. Copyright holders, however, maintain that copying and commercially exploiting their works without permission can violate their exclusive rights.
Courts may need to examine these questions separately, considering how copyrighted material was obtained, how it was used during training and whether generated outputs infringe protected works.
The Presley-related lawsuit could therefore contribute to the broader legal debate over the boundaries of AI-generated content.
A Wider Copyright Battle Facing AI Companies
Anthropic is not the only AI developer facing legal challenges over copyrighted material. Music publishers, authors, news organisations and other content owners have brought lawsuits against technology companies over their use of protected works in AI training.
In 2023, major music publishers filed a separate lawsuit against Anthropic concerning the alleged use and reproduction of copyrighted song lyrics. The dispute reflected growing concerns that generative AI systems could learn from creative works without appropriate licensing arrangements.
The latest case involving compositions associated with Presley adds another prominent music catalogue to the debate.
Rights holders argue that unauthorised AI training could undermine traditional licensing models and create competition for the original creators. They also contend that technology companies should be held to the same copyright standards as other businesses that commercially use protected creative material.
For AI developers, the challenge is determining how to build capable systems while complying with copyright law and responding to the expectations of creators.
What Could the Lawsuit Mean for Anthropic?
The outcome could influence how Anthropic obtains training material, manages copyrighted content and prevents its chatbot from generating protected lyrics.
If the plaintiffs succeed, the case could encourage AI companies to pursue licensing agreements with music publishers and strengthen safeguards against reproducing copyrighted compositions. It could also prompt additional rights holders to examine whether their works have been used without permission.
However, the lawsuit does not establish that Anthropic has violated copyright law. The company will have an opportunity to challenge the allegations, and the court will need to evaluate the evidence before reaching a conclusion.
The dispute also highlights the potential financial risks associated with AI training practices. Significant damages, legal expenses and restrictions on the use of copyrighted material could affect how technology companies develop and commercialise their products.
What Happens Next?
The case is part of a broader effort by copyright owners to establish clearer boundaries around the use of creative works in AI development.
The plaintiffs must establish their claims concerning the alleged unauthorised use of musical compositions and other copyright-related violations. Anthropic, meanwhile, is expected to contest the allegations.
The legal proceedings could help clarify how existing copyright protections apply to AI training and generated content.
Ultimately, the dispute centres on a question affecting the entertainment and technology industries: whether AI developers can use copyrighted creative works to build commercial products without obtaining permission or paying the rights holders.
As generative AI becomes increasingly widespread, the outcome could influence future licensing arrangements and shape how technology companies balance innovation with the protection of intellectual property.
