AI Affairs, home

Friday 30 October 2026

Government · Brief

Sag argues training on copyrighted books is fair use in OpenAI case

His 29 October brief, submitted across the consolidated cases, supports neither side and challenges plaintiffs’ arguments about expressive substitution and licensing markets.

A magnifying glass enlarges the OpenAI logo on a browser screen.
Photo: Jernej Furman from Slovenia, CC BY 2.0, via Wikimedia Commons (cropped)

A filing on CourtListener shows that law professor Matthew Sag submitted an amicus brief on 29 October 2026 to the United States District Court for the Southern District of New York in the consolidated OpenAI copyright infringement litigation. The document, numbered 2308-1 in case 1:25-md-03143-SHS-OTW, says it relates to all cases. Sag offers the court a view without being a party.

Sag, the Jonas Robitscher Professor of Law in Artificial Intelligence, Machine Learning, and Data Science at Emory University School of Law, says he supports neither side and has no direct interest in the outcome. He argues that training large language models on copyrighted books is highly transformative and strongly favours fair use under the purpose-and-character consideration.

AI Affairs previously covered Suskind and Frye’s brief on news and OpenAI and Microsoft’s dispute with news plaintiffs in the same litigation. Sag disputes the plaintiffs’ asserted harms to expressive-substitution and licensing markets, while accepting that indirect expressive substitution may be a cognisable market harm in appropriate circumstances.

Sources

Topics: Copyright, Foundation models