US Court Rejects Fair Use Argument as EU Considers Copyright Reform for Generative AI
date: 2026-10-12

A recent US court ruling in a copyright case highlights the challenges policymakers face in balancing the interests of content creators and AI developers, an expert has said.


Cerys Wyn Davies, a lawyer specializing in AI and copyright, made the comments after the European Commission indicated that it was considering copyright reform to address the various challenges arising from the use of copyright-protected content by generative AI.


The European Commission has invited content creators, AI developers and other stakeholders to propose measures they believe should be taken. The intervention options listed include introducing new mandatory provisions within the EU copyright framework, strengthening voluntary cooperation through “soft law instruments,” and adopting non-legislative measures to support AI access to and use of copyright-protected content.


The European Commission’s public consultation runs until November 3, 2026. On the day the consultation opened, a US court of appeals rejected the fair use defense in the dispute between media company Thomson Reuters and Ross Intelligence Inc.


Ross Intelligence developed an AI-powered legal research engine. The training data used for the product included legal headnotes published in the Westlaw database. These headnotes distill points of law and relevant details of court decisions. Thomson Reuters, which owns the copyright in works in the Westlaw database, had previously refused to grant Ross Intelligence a license to use the content to train its competing product. It argued that Ross Intelligence should be liable for infringement because it had used Westlaw headnotes to train its product without permission.


In its recent ruling, the US court of appeals held that the relevant Thomson Reuters materials were eligible for copyright protection and that Ross Intelligence’s use of those materials did not fall within the fair use exception under US copyright law. This case is one of several lawsuits worldwide concerning copyright issues relating to AI data inputs and model training.



返回顶部图标

京公网安备11010802050249号