This is an informal case summary prepared for the purposes of facilitating exchange during the 2024 WIPO IP Judges Forum.
Session 1 of the 2024 IP Judges Forum
Thomson Reuters Enter. Centre GmbH v. Ross Intelligence Inc., 694 F. Supp. 3d 467 (D. Del. 2023)
Date of judgment: September 25, 2023
Issuing authority: United States District Court for the District of Delaware
Level of the issuing authority: First Instance
Type of procedure: Judicial (Civil)
Subject matter: Copyright and Related Rights (Neighboring Rights)
Plaintiffs: Thomson Reuters Enterprise Centre GmbH and West Publishing Corp.
Defendant: Ross Intelligence Inc.
Keywords: Copyright Infringement, Fair Use, Preemption, Summary Judgment, Artificial Intelligence, AI, Machine Learning, Training Data
Basic facts: Thomson Reuters's Westlaw platform organizes judicial opinions according to its Key Number System, which is based on the type of law at issue. Westlaw also adds “headnotes” to opinions, which are short summaries of points of law that appear in the opinion. Ross Intelligence is a legal-research industry start-up who sought to create an artificial intelligence tool that could return quotations from judicial opinions in response to a user’s query. To build its AI tool, Ross used machine learning. Machine learning is a method by which a program studies extensive amounts of “training data,” and then uses that training data to create an output upon request that is based on the training data the program has studied. Ross tried to get a license to Westlaw to obtain training data, but Westlaw refused. Ross then turned to a third-party company to provide training data, much of which Thomson claims came directly from its headnotes on Westlaw. Thomson brings claims for copyright infringement and tortious interference with contract. Thomson moved for summary judgment on its copyright infringement claim. Both sides moved for summary judgment on Ross's fair use defense. Thomson also moved for summary judgment on its tortious interference with contract claim, and Ross has counter-moved on its preemption defense to that claim.
Held: For copyright infringement, there is a genuine dispute of material fact on the originality of Thomson’s headnotes and whether an ordinary observer would view the copyrighted work and the reproduced work as substantially similar. Thus, summary judgment is improper on at least two prongs of the legal test for copyright infringement. The third prong, proof of actual copying, is resolved in Thomson’s favor because there is no genuine dispute of material fact that Ross actually copied at least portions of the copyrighted work. For fair use, significant factual disputes remain that preclude summary judgment. All four fair use factors (the purpose and character of the use, the nature of the copyrighted work, the amount and substantiality of the portion used in relation to the copyrighted work as a whole, and the effect of the use upon the potential market for the copyrighted work, 17 U.S.C. § 107) must go to the jury. On tortious interference, Thomson presents several claims, each of which has different elements. Thomson’s first claim is preempted by federal copyright law because those claims do not include additional elements that must be proven beyond the elements required by federal copyright law. Thomson’s two remaining claims are not preempted, but cannot be entirely resolved at this junction due to genuine disputes of material facts.
Relevant holdings in relation to Frontier Technologies and Intellectual Property Adjudication: Plaintiffs alleging that their independent creations have been used as training data must be prepared to prove sufficient similarity to a jury. This could prove to be very challenging, as AI tools are often trained on millions of data points, yet to prove copyright infringement, Plaintiffs must show substantial similarity between their independent creation, which is likely one of many data points, and the output of the AI tool at issue.
Relevant legislation: Digital Millennium Copyright Act