A collective of professors with expertise in copyright law has submitted an amicus brief supporting authors in a lawsuit against Meta. The authors accuse Meta of using their ebooks without permission to train its Llama AI models.
This brief was filed on Friday in the U.S. District Court for the Northern District of California, San Francisco Division. It critiques Meta’s fair use defense, describing it as a significant overreach compared to legal privileges previously extended to human authors.
The brief argues that using copyrighted works for training generative models is not “transformative.” It equates this use with educating human authors, which is the original intent of the works. Furthermore, the brief states that since the purpose is to create works competing in the same markets, especially by a for-profit entity like Meta, this use is “undeniably commercial.”
The case, known as Kadrey v. Meta, involves authors such as Richard Kadrey, Sarah Silverman, and Ta-Nehisi Coates, who claim Meta infringed on their intellectual property rights. They allege that Meta used their ebooks for training purposes and removed the copyright information to obscure this infringement. Meta contends that its actions fall under fair use and argues for dismissal on the grounds that the authors lack legal standing to sue.
Earlier in the month, U.S. District Judge Vince Chhabria allowed the case to proceed, although he dismissed a portion of it. In his ruling, Chhabria stated that the copyright infringement allegation represents a “concrete injury” sufficient to establish standing, and noted that the authors have convincingly alleged that Meta deliberately removed copyright management information to conceal the infringement.
Currently, the courts are addressing multiple AI-related copyright lawsuits, including a case by The New York Times against OpenAI.
