Elvis Presley rights holders sue Anthropic over copyright infringement

The legal filing against Anthropic

A subsidiary of Authentic Brands Group, the entity that controls the intellectual property for Elvis Presley, filed a lawsuit against Anthropic on Friday, October 23. The legal action is joined by other companies that hold a stake in the music of the legendary performer. The primary allegation is that the AI company engaged in copyright infringement by using 274 of Presley's songs to train its models without permission.

The lawsuit specifically identifies some of Presley's most famous hits as being included in the training data. These include well-known tracks such as "All Shook Up," "Can't Help Falling in Love," and "A Little Less Conversation." The plaintiffs argue that Anthropic failed to obtain any licenses or provide payment for the use of these copyrighted works.

According to the filing, the AI company's business model is built on a scale of copyright infringement that has never been seen before. Richard Busch, the lead attorney for the plaintiffs, stated that the suit aims to enforce exclusive rights against the defendant. Busch is a notable figure in music copyright, having previously represented the estate of Marvin Gaye in a high-profile lawsuit against Robin Thicke and Pharrell Williams.

How the AI models were trained

The core of the dispute involves how Anthropic developed its Claude chatbot. The plaintiffs claim that the company ingested a vast amount of copyrighted music to teach the AI how to understand and generate content. This process allows the models to learn the patterns, styles, and structures of existing popular music without the creators' consent.

The lawsuit alleges that these models are now capable of generating unauthorized versions of existing songs. Even though Anthropic has claimed to implement safeguards to stop the chatbot from reproducing specific lyrics or compositions, the plaintiffs argue these protections are insufficient. They contend that the technology still allows for the creation of content that mimics the protected works of Elvis Presley.

The legal documents describe specific instances where the Claude chatbot was asked to recreate songs for a modern audience. In these cases, the AI reportedly produced material that closely resembled the original Elvis tracks. This suggests that the underlying training data remains influential in the output the chatbot provides to its users.

The defense of fair use

Anthropic has not yet spoken publicly regarding this specific lawsuit, but the company has maintained a consistent stance on the matter. They argue that training AI models on copyrighted material is a practice that should be protected under U.S. fair use law. Fair use is a legal doctrine that allows the use of copyrighted material without permission under certain circumstances, such as for purposes of criticism, comment, or news reporting.

The company's position suggests that the transformative nature of AI training—where the data is used to create a new tool rather than just copying the original work—qualifies as legal. However, the Elvis plaintiffs are challenging this interpretation by focusing on the scale and commercial nature of the AI's development. They argue that the use of 274 specific songs constitutes a direct infringement on their exclusive rights.

The legal battle will likely hinge on whether the courts view AI training as a transformative use or as a wholesale appropriation of intellectual property. If the court finds against Anthropic, it could create a significant hurdle for AI companies that rely on large datasets of human-created content. Conversely, a win for Anthropic could provide a broader shield for the data-scraping practices currently used by many AI developers.

A growing trend in music litigation

This lawsuit is not an isolated incident for Anthropic, which is facing a growing number of legal challenges from the music industry. The company is currently embroiled in copyright infringement allegations from several major entities, including Universal Music Publishing Group, Concord Music Group, Sony Music Publishing, and Warner Chappell Music.

The sheer number of plaintiffs suggests a coordinated effort by the music industry to establish boundaries for AI development. These companies represent a significant portion of the world's recorded music and are looking to protect their revenue streams from being diluted by AI-generated content. The Elvis Presley case adds a high-profile cultural icon to this list of targets.

The plaintiffs are seeking statutory damages of up to $150,000 for each of the 274 songs infringed. This calculation brings the total amount they are seeking to approximately $41.1 million. The outcome of this case may determine how much financial risk AI companies face when using popular cultural assets to build their products.

Implications for the AI industry

The success of the Elvis Presley plaintiffs could lead to a significant shift in how AI companies source their training data. If the court rules that using these 274 songs without a license is a violation, other AI companies may be forced to seek formal agreements with labels and publishers. This could lead to a more regulated environment where high-quality data comes with a high price tag.

For the average user, this might mean that AI tools become more restricted in what they can generate regarding specific artists. It could also lead to the development of 'licensed' AI models where users pay a premium to access content that has been legally cleared. This would create a clear distinction between open-source training and commercially licensed training.

As the legal process unfolds, the industry will be watching closely to see how the court balances the interests of technological innovation with the rights of creators. The case serves as a landmark test for the intersection of modern technology and traditional copyright law. It will likely influence how the next generation of AI models is built and deployed globally.