Elvis Rights Holders Demand $41M From Anthropic Over AI Training Allegations

Key points

• Authentic Brands Group, along with other rights holders, filed a lawsuit against Anthropic on Friday in the U.S. District Court for the Northern District of California.
• The complaint alleges Anthropic used 274 Elvis Presley songs, including hits like 'All Shook Up' and 'Can't Help Falling in Love,' to train its Claude model without licensing.
• Plaintiffs seek $150,000 per infringed song, totaling approximately $41 million in statutory damages, and request a jury trial.
• The lawsuit claims Claude continued to produce outputs closely resembling original songs even after Anthropic implemented safeguards, with the AI admitting to reproducing structures and hooks.
• Anthropic stated it disagrees with the claims and intends to defend the case fully, while facing similar lawsuits from Universal Music Group and Reservoir.

Background

This lawsuit follows a series of high-profile copyright actions against AI companies. In 2025, Anthropic agreed to pay $1.5 billion to settle a class action lawsuit from authors alleging its models were trained on their work, marking the largest publicly reported copyright recovery in history. Additionally, Anthropic has faced recent suits from Universal Music Group and Reservoir, indicating a broader industry trend of legal challenges over AI training data.

How outlets are covering it

BBC and Fox News emphasize the specific claim that Claude continued to infringe on copyrights even after Anthropic introduced 'guardrails,' highlighting a failure of technical safeguards. The New York Times provides additional legal detail, noting that plaintiffs accused Anthropic of circumventing protection measures and altering copyright management information, a point not detailed in the other sources. All outlets agree on the core facts of the 274 songs and $41 million demand, but differ in their focus on the ongoing nature of the infringement versus the procedural aspects of the lawsuit.

Why it matters

This case sets a significant precedent for AI companies regarding the use of copyrighted music in training data. It tests the effectiveness of AI safeguards and the legal boundaries of 'derivative works' generated by large language models, potentially influencing future regulations and business practices in the AI industry.

What to watch

Anthropic has stated it will defend the case fully. The lawsuit seeks a jury trial, and the outcome could impact how AI companies handle copyrighted material in their training processes and the development of safeguards to prevent infringing outputs.