This page collects documented examples of the influence of Matthew Sag’s scholarship:
In the courts
- His scholarship has been cited in dozens of federal court decisions (a 2026 Westlaw review found 39 genuine citations). His empirical copyright trolling work supplies the standard judicial definition of a copyright troll, cited from Malibu Media v. Doe (S.D.N.Y. 2015) through Strike 3 Holdings v. Doe, 351 F. Supp. 3d 160 (D.D.C. 2018), and by the Seventh Circuit in Design Basics, LLC v. Lexington Homes, Inc., 858 F.3d 1093 (7th Cir. 2017).
- The Ninth Circuit has cited his work twice: The Pre-History of Fair Use in Monge v. Maya Magazines, 688 F.3d 1164 (9th Cir. 2012), and the stadium rent-seeking study in City of Oakland v. Oakland Raiders, 20 F.4th 441 (9th Cir. 2021).
- In the generative AI litigation, Kadrey v. Meta Platforms, 788 F. Supp. 3d 1026 (N.D. Cal. 2025) engages Fairness and Fair Use in Generative AI directly — adopting the article’s terminology of “indirect substitution” in analyzing market dilution, while disagreeing with its conclusion on the fourth fair use factor.
- Amicus briefs of digital humanities and law scholars, which he co-authored, were before the courts in Authors Guild v. HathiTrust and Authors Guild v. Google — the decisions that established the fair use foundation for library digitization, search, and text data mining in the United States.
- Brief of Copyright Law Professors as Amici Curiae in Thomson Reuters v. Ross Intelligence (3d Cir. 2025), the first appellate case to address fair use for AI training.
- In September 2026 the United States filed a statement of interest in the OpenAI multidistrict litigation arguing that training large language models on copyrighted works is transformative fair use — the non-expressive use position, though the brief does not use the term.
- Abroad, in ANI Media Pvt. Ltd. v. OpenAI, the High Court of Delhi (India) declined to enjoin the use of copyrighted news content to train AI models, holding the practice prima facie fair dealing under India’s Copyright Act (July 24, 2026). The written submissions of the court-appointed amicus curiae, Prof. Arul George Scaria, drew at length on Copyright Safety for Generative AI to explain why LLM training is generally non-expressive and only rarely produces infringing outputs, and the court’s opinion cites The Globalization of Copyright Exceptions for AI Training (with Peter K. Yu) in its account of how models are trained.
Before Congress and the agencies
- U.S. Senate testimony — Judiciary Subcommittee on Intellectual Property, hearing on Artificial Intelligence and Copyright (July 12, 2023).
- U.S. Copyright Office — comments and reply comments in the Notice of Inquiry on Artificial Intelligence and Copyright (2023, with Pamela Samuelson and Christopher Jon Sprigman); invited presentations to the Office’s international AI webinar and listening sessions.
- White House OSTP — submission on copyright for the U.S. AI Action Plan (2025).
- International briefings for policymakers in Australia, Korea, and Argentina, and submissions to the Australian Law Reform Commission.
In the field
- Originated the concept of non-expressive use (2009), which has been adopted, debated, and extended in copyright scholarship and now frames the AI training debate.
- Convenes two recurring scholarly roundtables: the Copyright Scholars Roundtable (since 2015) and the Legal Scholars Roundtable on Artificial Intelligence at Emory (since 2022).
- His open-access casebook, Copyright Law in the Age of AI, is used in copyright courses in the United States and overseas.
- Research published in Nature and Science as well as leading law reviews (California, Georgetown, Northwestern, Notre Dame, Vanderbilt, Iowa, Fordham, Houston, Emory, Duke).
- Elected member of the American Law Institute (2018). Named to the Top 50 Legal Innovators in Academia (The National Law Review & Wickard, 2026) for work defining the copyright framework for AI training and generative AI outputs.
By the numbers
- 3,456 Google Scholar citations (1,808 in the last five years); h-index 27 (21 in the last five years).
- More than 49,000 SSRN downloads across 50 papers.
- Over 1,097 citing articles on HeinOnline (461 in the last five years).
In the media
Sag’s analysis of copyright and AI appears regularly in outlets including WIRED, Bloomberg Law, The Economist, and the Wall Street Journal, and on legal and technology podcasts. Journalists seeking comment can reach him at msag@emory.edu.