AI Exposes Flaws in Copyright’s Focus on Transformativeness in Fair Use

In this commentary piece for Pro Market author Shishene Jing discusses how generative AI is exposing weaknesses in the traditional approach to fair use under U.S. copyright law. For decades, courts largely emphasized whether an unauthorized use was sufficiently “transformative,” but AI can produce highly transformative outputs that nonetheless compete directly with the markets for the copyrighted works used to train it. The author argues that this tension requires a reassessment of how fair use is analyzed.

The piece contrasts two approaches to fair use: a market-focused framework associated with Wendy Gordon and the transformative-use framework developed by Judge Pierre Leval. The former views fair use primarily as a response to market failures and licensing barriers, while the latter emphasizes whether a use adds new meaning or purpose. The author notes that transformative use became dominant partly because most earlier fair-use cases involved uses that did not meaningfully substitute for the original works’ markets.

AI, however, can make both conditions exist simultaneously. The author points to 2025 cases involving AI training, including Bartz v. Anthropic, Kadrey v. Meta, and Thomson Reuters v. Ross Intelligence, which illustrate the tension between highly transformative uses and potential market substitution. The commentary also highlights the Supreme Court’s 2023 Andy Warhol Foundation v. Goldsmith decision as a course correction that emphasized whether a new work shares substantially the same commercial purpose or market as the original.

The author argues that courts should give greater weight to the Copyright Act’s fourth factor, which examines the effect of an unauthorized use on the potential market or value of the copyrighted work. Market displacement, the author contends, is more objectively assessable through economic evidence and licensing practices than the inherently subjective question of whether a work is sufficiently transformative. A market-focused approach would still protect parody, criticism, scholarship, and genuinely non-substitutive uses while addressing AI uses that are both transformative and commercially competitive…

CONTINUE READING…