{"id":137598,"date":"2025-11-25T16:15:54","date_gmt":"2025-11-26T00:15:54","guid":{"rendered":"https:\/\/xira.com\/p\/2025\/11\/25\/thomson-reuters-tells-appeals-court-rosss-copying-was-theft-not-innovation\/"},"modified":"2025-11-25T16:15:54","modified_gmt":"2025-11-26T00:15:54","slug":"thomson-reuters-tells-appeals-court-rosss-copying-was-theft-not-innovation","status":"publish","type":"post","link":"https:\/\/xira.com\/p\/2025\/11\/25\/thomson-reuters-tells-appeals-court-rosss-copying-was-theft-not-innovation\/","title":{"rendered":"Thomson Reuters Tells Appeals Court: ROSS\u2019s Copying Was \u2018Theft, Not Innovation\u2019"},"content":{"rendered":"

The post Thomson Reuters Tells Appeals Court: ROSS\u2019s Copying Was \u2018Theft, Not Innovation\u2019<\/a> appeared first on Above the Law<\/a>.<\/p>\n

In a redacted brief filed Nov. 19 with the 3rd U.S. Circuit Court of Appeals, Thomson Reuters urged the court to affirm the Delaware district court\u2019s ruling that ROSS Intelligence infringed Westlaw\u2019s copyrights by copying thousands of its attorney-written headnotes to train an AI-powered legal research tool.<\/p>\n

\u201cCopying protectable expression to create a competing substitute isn\u2019t innovation: it\u2019s theft,\u201d the brief asserts. \u201cThis basic principle is as true in the AI context as it is in any other.\u201d<\/p>\n

The 85-page brief (which you can read here<\/a>), signed by Kirkland & Ellis partners Dale Cendali, Joshua Simmons and Miranda Means, defends the copyrightability of Westlaw\u2019s headnotes, the editorial summaries written by its attorney-editors, and portrays them as a hallmark of creative legal analysis rather than mere factual summaries.<\/p>\n

\u201cFor over a hundred years and as recently as 2020,\u201d TR\u2019s brief argues, \u201cthe Supreme Court has upheld \u2018the reporter\u2019s copyright interest in explanatory materials including headnotes.\u201d Citing Callaghan v. Myers<\/em> (1888) and Georgia v. Public.Resource.Org<\/em> (2020), TR calls headnotes \u201ca paradigmatic example of protectable material,\u201d and argues that the Delaware court was right to treat 2,243 of them as copyrightable works.<\/p>\n

Related: \u2018No One Can Own the Law\u2019: Amici Come Out In Force to Support ROSS In Appeal of Copyright Ruling Favoring Thomson Reuters<\/a>.<\/strong><\/em><\/p>\n

TR asserts that its headnotes are crafted through numerous creative editorial choices \u2014 how to phrase the point of law, how many headnotes to create, which facts or concepts to include, which case passages to link and how to categorize them within the West Key Number System. These choices, TR says, easily satisfy the minimal creativity required by Feist.<\/p>\n

ROSS, the brief says, \u201cmay want to ignore the Supreme Court\u2019s numerous statements that headnotes are protectable, as it did in its opening brief, but this Court must follow binding precedent.\u201d<\/p>\n

\u2018Knew It Could Not Legally Access Westlaw\u2019<\/h3>\n

TR\u2019s account portrays ROSS as a commercial actor that knowingly copied Westlaw to build a rival product. After being denied a Westlaw license, ROSS allegedly hired the outsourcing firm LegalEase Solutions to scrape Westlaw data and convert headnotes into \u201cquestion and answer\u201d pairs for training its AI model.<\/p>\n

According to exhibits described in the brief, LegalEase contractors \u201ccopied the West Headnotes into the form of questions\u201d and then copied \u201cthe case passages that West\u2019s attorney-editors had selected to link to those headnotes.\u201d TR accuses ROSS of using bots to \u201cscrape Westlaw en masse<\/em>,\u201d creating \u201cthousands of Bulk Memos quickly\u201d and copying \u201chundreds of thousands of annotated cases.\u201d<\/p>\n

(Two days before using ROSS in 2020, TR settled litigation against LegalEase based on similar facts, with the two parties agreeing to entry of a consent judgment and stipulated permanent injunction<\/a>\u00a0in the U.S. District Court in Minnesota.)<\/p>\n

The brief asserts that ROSS used the resulting material multiple times in training its AI system. It cites testimony that ROSS already possessed a repository of case law but needed Westlaw\u2019s editorial analysis to build a functional search tool capable of mapping natural-language questions to relevant case passages.<\/p>\n

ROSS\u2019s conduct, TR contends, was not inadvertent: \u201cROSS knew it could not legally access Westlaw. When ROSS directly asked TR for a Westlaw subscription, TR expressly declined.\u201d Yet after learning this, the brief says, ROSS induced first another company (whose name is redacted) and then LegalEase to get ROSS access anyway.<\/p>\n

\u2018A Direct Substitute, Not a Transformative Use\u2019<\/h3>\n
\"\"<\/a><\/p>\n

A ROSS ad reproduced in TR\u2019s brief.<\/p>\n<\/div>\n

On the question of fair use, TR\u2019s central argument is that ROSS\u2019s platform \u201csubstituted for and competed with Westlaw in the legal research platform market.\u201d<\/p>\n

It says ROSS\u2019s marketing materials explicitly positioned its AI as a \u201cWestlaw replacement,\u201d even using slogans like \u201cROSS or Westlaw?\u201d<\/em> alongside a price comparison ad \u2014 a copy of which is reproduced in the brief.<\/p>\n

Under the Supreme Court\u2019s 2023 decision Andy Warhol Found. for the Visual Arts v. Goldsmith<\/em>, TR says, ROSS\u2019s use was not \u201ctransformative\u201d because it served \u201cthe same purpose as the original,\u201d which was to \u201chelp researchers find and understand the law.\u201d<\/p>\n

It draws a contrast with other cases, such as one involving Google Books, which merely indexed books and drove users back to the originals.<\/p>\n

See all my coverage of this litigation here<\/a>.<\/strong><\/em><\/p>\n

Here, it contends, ROSS \u201ccopied the Westlaw content that already<\/em> provided a way for researchers to find and understand law to develop a competing<\/em> way to find and understand law.\u201d<\/p>\n

TR also accuses ROSS of acting in bad faith, noting a similar case in which the court found bad faith when the defendant \u201crequested a license, was refused one, and then obtained a copy from a third party rather than paying the requisite fee.\u201d<\/p>\n

That, it says, \u201cis precisely what happened here, where ROSS was refused a license and then illicitly went through a third party.\u201d<\/p>\n

Harm to Westlaw\u2019s Markets<\/h3>\n

Much of TR\u2019s brief focuses on market harm, which it argues is the most important of the fair use factors. It argues that ROSS\u2019s copying deprived TR of several valuable markets:<\/p>\n

    \n
  • The existing market for Westlaw subscriptions.<\/li>\n
  • The potential market for licensing Westlaw content as AI training material.<\/li>\n
  • The exclusive ability to train its own AI using that content.<\/li>\n<\/ul>\n

    \u201cROSS harmed the original market for Westlaw by substituting therefor,\u201d TR argues, and it \u201cdiminished the value of the Westlaw content by depriving TR of its exclusive ability to train its own AI on that content.\u201d<\/p>\n

    A ruling in ROSS\u2019s favor would have broad consequences, the brief argues. \u201cIf any competitor could copy the Westlaw content to train their own legal research platform, why on earth would anyone pay TR for it?\u201d<\/p>\n

    AI Innovation or \u2018Parasitic Copying\u2019?<\/h3>\n

    Responding to arguments from ROSS and others that enforcing TR\u2019s copyright in this case would hinder AI progress, TR suggest that is alarmist, pointing out that Westlaw itself has used artificial intelligence \u201clong before the founders of ROSS were in school.\u201d The company cites milestones from its own AI history dating back to the 1990s, including its 1992 launch of the \u201cfirst commercially available search engine with probabilistic rank retrieval\u201d and the 2018 launch of WestSearch Plus, an AI-powered research feature.<\/p>\n

    \u201cAI development has moved forward at a rapid pace since the decision below was entered, and will surely continue to do so,\u201d the brief says.<\/p>\n

    While there may be scenarios where training an AI algorithm using copyrighted material is fair use, \u201cthis scenario \u2014 where the copying was for purposes of creating a commercial substitute for the original \u2014 is not one of them.\u201d<\/p>\n

    The brief\u2019s concluding paragraph drives home the theme that ROSS\u2019s behavior is not about innovation but misappropriation:<\/p>\n

    \u201cThis case may involve AI, but it is far from novel. ROSS indisputably pilfered the creativity of a competitor to bring to market a substitute. ROSS\u2019s copying was not technological advancement. It was theft.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"

    The post Thomson Reuters Tells Appeals Court: ROSS\u2019s Copying Was \u2018Theft, Not Innovation\u2019 appeared first on Above the Law. In a redacted brief filed Nov. 19 with the 3rd U.S. Circuit Court of Appeals, Thomson Reuters urged the court to affirm the Delaware district court\u2019s ruling that ROSS Intelligence infringed Westlaw\u2019s copyrights by copying thousands […]<\/p>\n","protected":false},"author":3,"featured_media":137486,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_et_pb_use_builder":"","_et_pb_old_content":"","_et_gb_content_width":"","_jetpack_memberships_contains_paid_content":false,"footnotes":""},"categories":[16],"tags":[],"class_list":["post-137598","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-above_the_law"],"jetpack_featured_media_url":"https:\/\/i0.wp.com\/xira.com\/p\/wp-content\/uploads\/2025\/11\/TR-ROSS-Featured-Image-1024x576-iDzciG.png?fit=1024%2C576&ssl=1","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/137598","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/comments?post=137598"}],"version-history":[{"count":0,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/posts\/137598\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media\/137486"}],"wp:attachment":[{"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/media?parent=137598"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/categories?post=137598"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/xira.com\/p\/wp-json\/wp\/v2\/tags?post=137598"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}