There are so many pending AI lawsuits that you’d be forgiven for not keeping them straight. Even some of the plaintiffs seem confused, with major record labels partnering with AI startups while suing those exact same startups. Anthropic AI has been sued a number of times (including by at least one of those record companies, Universal Music Group), and last summer, the writers Andrea Bartz, Charles Graeber and Kirk Wallace Johnson alleged that the company was “strip-mining the human expression and ingenuity behind” human-written books when it trained its Claude chatbot on millions of copyrighted works.
Well, Anthropic received a legal win this week when a court ruled that it didn’t break the law by training Claude on the books, according to ABC News. Judge William Alsup ruled that because what the generative AI produced was “quintessentially transformative” it qualified as fair use, writing in his decision, “Like any reader aspiring to be a writer, Anthropic’s (AI large language models) trained upon works not to race ahead and replicate or supplant them — but to turn a hard corner and create something different.”
However, the company is not out of the woods and will go to trial in December over how it obtained those works. It’s been alleged that the books Claude was trained on were from “shadow libraries” full of stolen books, which does seem like a pretty big deal. (Reading books is one thing—stealing them is quite another.) Alsup continues in his decision, “Anthropic had no entitlement to use pirated copies for its central library.” Anthropic faced similar allegations from the Writers Guild late last year when the company was one of many cited in an Atlantic article about film and TV subtitles being used to train AI models without permission. The UMG lawsuit also saw Anthropic facing allegations of using lyrics without permission; that case is also still ongoing.
