Matthew Sag · Jonas Robitscher Professor of Law in Artificial Intelligence, Machine Learning, and Data Science · Emory University School of Law
In 2009 I argued that copying works in order to analyze them — not to read them — should generally not infringe copyright. That concept, non-expressive use, now sits at the center of the copyright disputes over AI training. I have been following the thread from search engines to text data mining to generative AI ever since.
Casebook · 2026 edition
A complete casebook on United States copyright law — 28 chapters, 177 extracted decisions, current to August 2026. AI runs throughout the book rather than being confined to a single chapter, because many of copyright’s most contested questions now involve AI.
Primer · six chapters
What AI is, how large language models work, why prediction is not truth, and the interests and concerns that motivate regulation — a general introduction for a law & AI syllabus, with no computer science background required.
Since 2022 I have convened the Legal Scholars Roundtable on Artificial Intelligence at Emory Law — an annual gathering for legal scholarship on AI — alongside the Copyright Scholars Roundtable, which I have co-convened since 2015. I convene both as forums for discussing current scholarship in their respective fields.
About the roundtable, and the current call for papers →A framework explaining why machine copying for search, indexing, mining, and training should generally not infringe when the copied expression is not shown to humans. I introduced the idea in 2009; it is now central to the AI training cases.
Fair use and generative AI, copyright safety, the "Snoopy problem," the globalization of training exceptions, and testimony before the U.S. Senate.
From the HathiTrust and Google Books amicus briefs to legal reform proposals in Science: securing the right to do computational research on copyrighted works.
The doctrine's structure, its eighteenth-century origins, and the empirical evidence that it is more predictable than its critics claim.
Empirical scholarship that documented and named the mass-litigation business model — cited by federal courts confronting it.
With Tonja Jacobi: mining oral argument transcripts to measure ideology, interruptions, advocacy, and even laughter at the Court.
Named one of the Top 50 Legal Innovators in Academia by The National Law Review & Wickard (2026), recognized for work defining the copyright framework for AI training and generative AI outputs.
Generative AI is not just a subject I study — it is a technology the legal profession, and legal education, now has to metabolize. Much of my recent writing is about how law schools should adapt:
An 18-page paper arguing that our students deserve an education that prepares them for the legal profession as it will be, not as it was.
AI, Accommodations, and Assessment Integrity
Why generative AI has made most forms of assessment worthless as a measure of what a student knows — and what to do about it.
Emory Law's New AI & Law Concentration
The concentration I drafted and negotiated — what it covers and why it exists.
Public lectures
I give public lectures on understanding and using generative AI —
for universities, courts, bar associations, firms, and general audiences. Recent venues include Yale Law School, Harvard Law School (for an audience of federal judges), the U.S. Senate, and the National Library of Australia.
My amicus brief in the OpenAI/Microsoft generative AI copyright litigation
Today I filed an amicus brief in In re OpenAI, Inc. Copyright Infringement Litigation. There are a lot of different generative AI copyright lawsuits floating around, but this is arguably one of…
Some advice on effective and responsible use of generative AI in memos and presentations
These comments are specific to Current Topics in the Law and Policy of AI, a two-credit Fall 2026 course that I teach at Emory University School of Law. Unlike many law school…
I am the co-convenor of two academic roundtables: one devoted to Law and AI, another devoted to copyright. Between them, I read about 60–80 submissions last year that were supposed to be,…
DOJ Enters the AI Copyright Fight — and Mostly Gets It Right
So, this just happened. On September 1, the U.S. Department of Justice filed a statement of interest in the sprawling OpenAI copyright litigation, In re OpenAI, Inc., Copyright Infringement Litigation, the multidistrict…
I Wrote a Primer on Law and AI
I have taught several different courses on law and AI at Emory Law, and I’m currently teaching “Current Topics in the Law and Policy of AI.” I’m not wild about any of…
AI, Accommodations, and Assessment Integrity
An open letter to my colleagues at Emory Law Dear colleagues, Generative AI has made every form of assessment other than a supervised exam (written or oral) essentially worthless as a measure…