Recently, an artificial intelligence startup in San Francisco Anthropic Anthropic is being sued by a group of authors for using a large number of pirated books without authorization when training its popular chatbot Claude. The case is the first time authors have brought a lawsuit against Anthropic, although similar lawsuits have been filed against rival OpenAI over the past year. Anthropic is a small company founded by former OpenAI leaders that bills itself as a more responsible and safety-conscious developer of generative AI models that can compose emails, summarize documents, and communicate naturally with people.
However, the lawsuit, filed on Monday in federal court in San Francisco, alleges that Anthropic's actions "make a mockery of its noble goals" by using pirated books to build its AI products. "It is no exaggeration to say that Anthropic's model seeks to profit from the human expression and creativity behind every work," the complaint states.
The lawsuit is filed by three authors, Andrea Bartz, Charles Graeber and Kirk Wallace Johnson, who hope to represent a group of fiction and nonfiction authors who have also been affected. While this is the first author lawsuit against Anthropic, the company is also facing lawsuits from major music publishers alleging that Claude reused the affected material.copyrightProtection song lyrics.
The case joins a string of other lawsuits against developers of large language models for AI, including OpenAI and Microsoft’s ongoing legal battles with high-profile authors like John Gleason, Jodi Picoult, and A Song of Ice and Fire author George R.R. Martin, as well as lawsuits from media organizations like The New York Times, The Chicago Tribune, and Mother Jones.
What all these cases have in common is that tech companies are using large amounts of human work without authorization to train AI chatbots to produce human-like text. The lawsuits come not only from writers, but also from visual artists, music companies, and other creators, all of whom claim that the profits of generative AI are based on the misappropriation of original works.
Although Anthropic and other technology companies have argued that the training of AI models falls within the scope of "fair use" under US law, the lawsuit alleges that the dataset they used, "The Pile," contains a large number of pirated books. At the same time, the lawsuit also refutes the view that AI systems learn in a similar way to humans, emphasizing that humans learn by purchasing legal books or borrowing books from libraries, which at least brings some financial compensation to authors and creators.