IP Law Daily, COPYRIGHT NEWS: Authors file class action suit against Salesforce over use of their novels to train AI, (Oct 17, 2025)

By Kevin M. Finson, J.D.

The complaint alleges the plaintiff authors’ works were included in the RedPajama and The Pile datasets used by Salesforce to develop its XGen AI models.

Novelists E. Molly Tanzer and Jennifer Gilmore, on behalf of themselves and other similarly situated, brought suit against Salesforce, Inc. (Salesforce) in the U.S. District Court in San Francisco. The authors allege that Salesforce used two datasets, RedPajama and The Pile, containing their copyright-protected books, as well as thousands of others, to train its artificial intelligence XGen series of large language models (LLMs) (Tanzer v. Salesforce, Inc., No. 3:25-cv-08862-CRB (N.D. Cal. Oct. 15, 2025)).

E. Molly Tanzer is an author living in Colorado and the owner of registered copyrights in her book Creatures of Will and Temper, among others. Jennifer Gilmore is an author living in Pennsylvania and the owner of registered copyrights in her book If Only, among others. The pair filed suit on behalf of themselves and others similarly situated against Salesforce, a company that provides cloud-based services and produces the Xgen large language model AI products.

The authors allege that their works, among others, were included in two “notorious” online databases of pirated copyright-protected works, known as RedPajama and The Pile. According to the complaint, when Salesforce announced the release of the XGen models, it uploaded the models to GitHub, a website where companies can provide access to open-source models and users can ask questions to its creators. The authors allege that a Salesforce developer working on the XGen series LLM products admitted in a post on GitHub that Salesforce used RedPajama and The Pile datasets as training data in producing the XGen AI products. This use as training data is alleged to have included impermissible copying of the protected works, and the authors had never given authorization for such use.

The authors assert one claim of direct copyright infringement and request that the action be certified as a class action. They are seeking statutory and other damages, attorney fees, a declaration that the alleged infringement was willful, and destruction of all infringing copies of the copyrighted works made by Salesforce.

The Case is No. 3:25-cv-08862-CRB.

Judge: Breyer, C.

Attorneys: Joseph R. Saveri (Joseph Saveri Law Firm, LLP) for E. Molly Tanzer and Jennifer Gilmore.

Companies: Salesforce, Inc.

News: AINews Copyright TechnologyInternet CaliforniaNews