Elvis Presley's Music Owners Sue Anthropic Over AI Training

The owners of Elvis Presley's music have sued Anthropic, the artificial intelligence company behind the Claude chatbot. They accuse it of training its AI models on hundreds of Elvis songs without permission. The lawsuit was filed on Friday, 23 October 2026, and was first reported by Billboard.
The case claims Anthropic infringed 274 copyrighted Elvis Presley songs. Among the titles it cites are "Can't Help Falling in Love," "A Little Less Conversation," "Don't Be Cruel," "All Shook Up," "Return to Sender" and "(You're The) Devil in Disguise."
The lawsuit was filed by a unit of Authentic Brands Group, a company that controls the intellectual property rights to Elvis. Those rights include his song catalogue, a catalogue being the full set of songs a company owns or manages. They also cover the Elvis brand and trademarks, and the rights to his name and image. Other Elvis rights entities are part of the case. Richard Busch is the lead attorney for the plaintiffs.
Why does this matter to music people? AI models learn by processing huge amounts of material, including songs. Rights holders argue that copying recordings or compositions for that process needs a licence, which is formal permission, usually paid for. Anthropic has not been found to have done anything wrong. These are claims made in a court filing, and they have not been tested in court.
Elvis is not the only catalogue owner taking this route. On 17 August 2026, Round Hill Music announced copyright infringement lawsuits against Suno, an AI music generator, and against Anthropic. Those suits allege that the defendants used musical works owned or managed by Round Hill without authorisation, licence or compensation, according to Round Hill's announcement.
Universal, one of the three largest record companies, sued Anthropic in 2024, alleging systematic and extensive infringement. Earlier this year, the singer-songwriter Jason Isbell also sued Suno. A report published on 2 September 2026 described his complaint as alleging that Suno captured and stored artists' voiceprints, the digital patterns that identify a singer's voice. It also alleged that users could get around the tool's guardrails, the built-in limits meant to stop misuse.
For the songwriters and performers whose work sits in these catalogues, the question is a practical one. If a machine learns from a song, who should be paid, and how much? Those questions are now being put to the courts in several places at once, and none of the cases has produced a ruling yet.


