Does Using In-Copyright Works as Training Data Infringe?

Saved in:

Bibliographic Details
Title:	Does Using In-Copyright Works as Training Data Infringe?
Authors:	Samuelson, Pamela¹ (AUTHOR) pam@law.berkeley.edu
Source:	Communications of the ACM. Nov2025, Vol. 68 Issue 11, p28-30. 3p.
Subjects:	Copyright infringement, Generative artificial intelligence, Training, Copyright cases, Fair use (Copyright), Book piracy, Copyright, Data
Abstract:	The article discusses the legal aspects of copyrights in relation to generative artificial intelligence (GenAI) developments, and it mentions copyright infringement claims involving the use of in-copyright materials in GenAI training data. The copyright court cases Bartz v. Anthropic and Kadrey v, Meta are assessed, as well as copyright's fair use doctrine, legal defenses, and, copyright-related piracy of books. Lost sales and market dilution are addressed.
Database:	Engineering Source

Description
Abstract:	The article discusses the legal aspects of copyrights in relation to generative artificial intelligence (GenAI) developments, and it mentions copyright infringement claims involving the use of in-copyright materials in GenAI training data. The copyright court cases Bartz v. Anthropic and Kadrey v, Meta are assessed, as well as copyright's fair use doctrine, legal defenses, and, copyright-related piracy of books. Lost sales and market dilution are addressed.
ISSN:	00010782
DOI:	10.1145/3764097