Substantive AI Litigation · November 2026
Thomson Reuters Enter. Ctr. GmbH v. ROSS Intelligence Inc.
No. 25-2153, Doc. 214 (3d Cir. Sept. 30, 2026)
Third Circuit holds that copying Westlaw headnotes to train an AI legal-research tool is not fair use.
Defendant ROSS used about 25,000 Westlaw headnotes, through training memos prepared by a contractor, to train an AI legal-research platform meant to compete with Westlaw. The Third Circuit held that the 2,243 headnotes at issue were original and that ROSS's use was not fair.
ROSS's use was commercial and “minimally transformative, at best,” because both platforms serve the same purpose of helping users find responsive legal material; copying was not necessary because the underlying opinions were freely available, and “[u]nlike necessity, ease is not a justification for copying.” On market harm, the court recognized that “the market for licensing headnotes as text to train AI is rapidly developing” and that ROSS usurped Thomson Reuters's opportunity to license its headnotes for that purpose.
The court distinguished generative AI training cases such as Bartz v. Anthropic PBC, 787 F. Supp. 3d 1007, 1014 (N.D. Cal. 2025), and In re OpenAI, Inc. Copyright Infringement Litig., No. 1:25-md-3143 (S.D.N.Y. Sept. 1, 2026), noting that ROSS's platform “cannot generate original expression.”