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 every lawsuit 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 through the book rather than being confined to a chapter, because that is where copyright's most contested questions now live.
Primer · six chapters
What AI is, how large language models work, why prediction is not truth, and the interests and concerns that motivate regulation — the missing general introduction for any law & AI syllabus, no computer science degree required.
Convenor · annual
The annual roundtable for legal scholarship on artificial intelligence, hosted at Emory Law since 2022 — alongside the Copyright Scholars Roundtable, convened since 2015.
The framework for why machines copying works they never show to humans — search, indexing, mining, training — should generally not infringe. Originated here; 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.
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…
Emory Law’s New AI & Law Concentration
Training Lawyers for a New Era Starting in academic year 2026–27, Emory Law will offer a formal concentration in Artificial Intelligence and the Law — a structured academic pathway for J.D. students…
2026 Legal Scholars Roundtable on Artificial Intelligence
The 5th Annual Legal Scholars Roundtable on Artificial Intelligence at Emory University School of Law starts at Emory Law tomorrow. The Roundtable features a phenomenal lineup of authors, commentators and participants, including:…
David Kemp’s AI Policy Builder for Legal Education
David Kemp has just released a policy builder designed to help people who are struggling to design an AI policy that is relevant to their specific course. According to the website, the…
This post is a very lightly edited extract from my forthcoming article in the Duke Law Journal, Copyright’s Jagged Frontier (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6319379) What does AI memorization prove? Some argue that any evidence of…