Quizlet's Copyright Fight: A New Front In The Generative AI Legal Wars

The recent ruling in Barkley & Associates, Inc. v. Quizlet, Inc., No. 2:24-cv-05964-WLH (C.D. Cal. Sept. 2025), adds fuel to the growing legal fire over how generative AI companies acquire and use copyrighted content. At issue: whether Quizlet, a study platform turned AI content generator, unlawfully trained its "Quizlet+ AI" on proprietary test prep materials created by Barkley & Associates, a publisher of nurse practitioner exam study guides.

The court denied Quizlet's motion to dismiss, allowing Barkley's claims of direct and vicarious copyright infringement to move forward. Barkley alleges that Quizlet not only hosted infringing uploads from users but also directly copied its materials to train AI systems that generate educational outputs mimicking Barkley's proprietary content. Notably, the court accepted Barkley's theory that copyright protection can extend to the selection, arrangement, and pedagogical framing of even factual or test-based content—a clear signal that "thin" protection is still protection.

Quizlet's argument leaned on familiar defenses: the materials are too factual to be copyrightable, user uploads aren't its responsibility, and AI outputs fall under fair use. But the court rejected these defenses at this early stage, emphasizing that Barkley plausibly alleged originality in its curation of test content. The case now tees up critical questions about the role of educational compilations in the AI training pipeline and whether AI-generated outputs that "closely mimic" copyrighted materials cross the infringement line.

The ruling adds to early decisions among courts grappling with similar AI copyright issues. In Ross Intelligence, the court found AI training on legal headnotes infringed copyrights. In contrast, Kadrey v. Meta granted Meta summary judgment, finding plaintiffs failed to show market harm, allowing Meta's fair use defense to prevail. Meanwhile, Bartz v. Anthropic remains notable for Judge Alsup's controversial fair use finding in favor of AI defendants despite clear copying of entire books, while not granting summary judgment on intermediate copies that may not have been properly acquired by Anthropic. The Quizlet case signals skepticism toward AI's blanket fair use defenses, particularly at early stages and in educational and commercial contexts.

This ruling raises the stakes for AI companies training on structured proprietary content, particularly in education. For plaintiffs, it reinforces the viability of bringing copyright claims rooted in the arrangement and instructional design of factual material. For platforms, it's a warning: thin copyright can cut deep when wielded strategically. As the case proceeds to discovery, it could become a bellwether for AI training litigation in academic publishing and beyond.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.