AI unicorn, backed by Google and Amazon, faces copyright battle
What's the story
Anthropic, one of the world's hottest AI start-ups, is facing legal action from three authors who allege that their copyrighted work was unlawfully used to train the company's artificial intelligence (AI) model, Claude.
The lawsuit was filed on behalf of Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson in California.
The authors accuse Anthropic of building its business by "stealing hundreds of thousands of copyrighted books," without obtaining permission or compensating the creators.
AI exploitation
Authors claim AI model exploits their work
The authors assert that Anthropic's AI model, Claude, has been used to generate inexpensive book content.
They cite an instance from May 2023 when a man named Tim Boucher reportedly authored 97 books using Claude and another AI tool, ChatGPT from OpenAI.
Each book was completed in just "six to eight hours," and sold for prices ranging between $1.99 (roughly ₹167) and $5.99 (roughly ₹500).
Piracy allegations
Anthropic accused of using pirated collections
The lawsuit further alleges that Anthropic knowingly used datasets known as The Pile and Books3, which include Bibliotik, a collection described as "a notorious pirated collection." This was allegedly done to evade licensing costs.
The authors claim that this action violates US copyright law and are seeking damages from the San Francisco-based AI company, which has raised over $7 billion from prominent names such as Google, Amazon, and Salesforce.
Legal trend
AI copyright infringement cases on the rise
Since 2022, there has been a surge in lawsuits related to generative AI services like GitHub Copilot, Midjourney, and ChatGPT.
Several cases involving authors have been consolidated into single cases. However, it remains unclear how US copyright law will apply to AI training or output.
Last year, The New York Times sued OpenAI over similar allegations of copying journalists' work and profiting unfairly from it.