About
The book
Copyright Law in the Age of AI is a complete casebook on United States copyright law. Across 28 chapters and 177 extracted decisions, it covers the field from subject matter and originality, the reproduction and derivative work rights, fair use, ownership, the distribution and public performance rights, secondary liability, the safe harbors, and remedies.
This is a general copyright law textbook that focuses on the intersection of copyright law and AI because that is where the most interesting and contested questions in copyright are. AI runs through the book rather than being confined to a single chapter. Comparative notes set the American position against decisions from the United Kingdom, the European Union, Germany and India. The book can be used as a standalone textbook, or a chapter at a time alongside another.
A few things set it apart from comparable copyright textbooks. It tackles fair use relatively early and devotes six chapters to it. That’s a lot of space devoted to fair use, but there are two good reasons for this. First, to really understand fair use, you need to go beyond the headline Supreme Court cases and most recent court of appeals decisions and see how judges apply the doctrine across a range of different fact patterns. Second, the focus on fair use is a fair reflection of where modern copyright law is at. It seems that almost all of the big questions in copyright law are, to some extent, questions about fair use.
Another thing I have tried to do in this book is to integrate issues relating to AI wherever they are relevant, rather than quarantining AI to one or two chapters.
Edition. This site carries the 2026 Edition, version 1.1, published on August 23, 2026. The PDF is the edition of record; the text here is generated from the same source. Changes made after publication are recorded in the changelog.
A note on extracts. All extracts have been edited for concision and readability, often without any indication that they have been changed from the original. If you want a reliable quote from any of these sources, find the original on Westlaw, Google Scholar or the court’s website.
The author
Matthew Sag is Professor of Law, Artificial Intelligence, Machine Learning and Data Science at Emory University School of Law. He writes about copyright law and intellectual property, and is a leading United States authority on the fair use doctrine and its implications for researchers working in text data mining, machine learning and artificial intelligence.
His research, publications and commentary are at matthewsag.com, and most of his academic writing is on SSRN.
How to cite
The book:
Matthew Sag, Copyright Law in the Age of AI (2026 Edition), https://aicopyright.ai.
A chapter, for example chapter 15:
Matthew Sag, Copyright Law in the Age of AI ch. 15, Nonexpressive Use as Fair Use (2026 Edition), https://aicopyright.ai/chapters/15/.
Every chapter page ends with a ready-made citation in this form. For a pinpoint reference, cite the page in the PDF; each chapter page links to where that chapter begins.
Permissions and corrections
© Matthew James Sag (2018–2026). No copyright is claimed in United States government works or in the works of other authors specifically credited in the book. Third-party copyrighted material is included with permission of the rights holder, or more often in reliance on the fair use doctrine. Reading this book will help you understand why that makes sense.
The book is free to read, download, print and assign for teaching, study, and research from this site. You are not authorized to distribute this work through any other channels, except that instructors who have assigned any portion of this book to their students may make it available to those students on their institution’s free electronic course reserve. For any other permission, or to report a correction, email [email protected] with the subject line “Copyright Law in the Age of AI”.
AI training and retrieval
Use of this book to train machine learning models, including large language models, is expressly authorized, as is its use for retrieval-augmented generation, search indexing, and similar computational purposes. No further permission is needed and no fee is payable. This includes crawling the site, downloading the PDF, and reproducing passages of the book in responses to users, with attribution to Copyright Law in the Age of AI and a link to this site where practical.
This authorization covers the author’s own text and the selection and arrangement of the book. It does not, and could not, extend to material owned by others that appears in the book, whether by permission or in reliance on fair use. Whether computational uses of that material are lawful without permission is a question the book itself takes up at length, most directly in chapter 15.
The same permission is signalled to machines: every AI crawler is allowed in robots.txt, the site declares no rights reservation under the TDM Reservation Protocol, and llms.txt lists the chapters in a form intended for language models.
This site
This site, aicopyright.ai, is a static site generated directly from the textbook manuscript. It changes only when the book does. For Matthew Sag’s other writing on copyright and AI, see matthewsag.com.