A group of authors has filed a lawsuit against the start-up company Anthropic. They accuse the company of "theft on a grand scale" for allegedly used pirated copies of copyrighted books to train its chatbot Claude. The lawsuit, filed in a federal court in San Francisco on Monday, is the first time writers have taken legal action against Anthropic. The company advertises itself as a responsible and safety-oriented developer of generative AI models. These models can compose emails, summarize documents and converse in natural conversations.
The San Francisco-based company Anthropic was founded by former executives of OpenAI, another prominent AI company that is facing legal battles over similar issues. The lawsuit alleges that Anthropic, despite its stated commitment to ethical AI development, undermined its principles by allegedly using pirated software to develop its AI products.
The plaintiffs, authors Andrea Bartz, Charles Graeber and Kirk Wallace Johnson, seek to represent a broader group of authors whose fiction and nonfiction books may have been used without permission. They argue that Anthropic's practices amount to profiting from the unauthorized use of their creative works.
Anthropic has not yet responded to requests for comment on the lawsuit. The case against the company is part of a broader wave of litigation targeting AI language model developers, particularly in San Francisco and New York. OpenAI, the maker of ChatGPT, along with its business partner Microsoft, is already facing multiple copyright infringement lawsuits from well-known authors such as John Grisham, Jodi Picoult and George R. R. Martin, as well as major media outlets such as the New York Times and the Chicago Tribune.
The main issue in these cases is the allegation that AI companies have used large amounts of human-authored text to train their chatbots without asking the authors for permission or compensating them. This has led to a growing chorus of complaints from writers and visual artists, music publishers and other creators who argue that the rise of generative AI is based on the misappropriation of their works.
Copyright Battle
Like other tech companies, Anthropic has defended its practices by invoking the "fair use" doctrine, which allows limited use of copyrighted material under certain conditions, such as for research or transformational purposes. However, the lawsuit against Anthropic challenges this defense, particularly with respect to its alleged use of a dataset known as "The Pile," which allegedly contains many pirated books.
The lawsuit also disputes the notion that AI systems learn in a manner comparable to humans. It argues that while humans buy or borrow legal copies of books and thus compensate the authors, AI models like Claude are trained with vast amounts of data, including potentially pirated material, without receiving similar compensation.
This lawsuit underscores the intensifying debate over the ethical and legal limits of AI development, particularly with regard to the use of copyrighted content. As the industry continues to grow, the outcomes of these lawsuits could significantly impact the future of AI and the rights of creators.
