The long-running copyright litigation between Thomson Reuters and ROSS Intelligence is now pending in the 3rd U.S. Circuit Court of Appeals for an interlocutory appeal of the trial judge’s rulings in favor of TR.
Recently here, I reported on the 10 amicus curiae briefs filed in support of ROSS, all arguing that the now-defunct AI legal research startup did not violate copyright law.
Now, nine amicus briefs have been filed in support of TR. Those filing briefs range from major movie studios such as Disney and Paramount, to news media and copyright organizations, to individual copyright law professors, and even to TR’s principal competitor LexisNexis.
(Note: All of my stories covering this case back to 2020 can be found here.)
The primary argument of all nine briefs is that Westlaw’s headnotes are sufficiently original to qualify for copyright and that ROSS’s unauthorized copying of them to build its AI legal research platform was not fair use.
This matter is now before the 3rd Circuit after the trial judge in the case, U.S. Circuit Judge Stephanos Bibas, sitting by designation in the U.S. District Court in Delaware, issued partial summary judgment last February in favor of TR on two key issues of copyright law.
At ROSS’s request, he then postponed the scheduled trial to allow ROSS to file an interlocutory appeal on those two issues.
The two issues Judge Bibas certified for appeal to the 3rd Circuit are:
- Are Westlaw’s editor-created headnotes and Key Number system sufficiently original to qualify for copyright protection? The answer could have profound implications for the legal publishing industry. Judge Bibas concluded that attorney editors exercise enough creativity when they choose which parts of judicial opinions to highlight and how to categorize them, meeting copyright law’s low bar for originality.
- Fair use. Even if the headnotes are copyrightable, did ROSS’s use of them constitute fair use under copyright law? Judge Bibas found that ROSS’s use was not “transformative” because both companies created legal research tools that “identify the important parts of a large body of law.” He concluded that ROSS essentially created a “market substitute” for Westlaw rather than adding new meaning or purpose to the original work.
Below are quick summaries of the amicus briefs, including a “nugget” quote, with links to the full text of each. I published the summaries of the briefs in favor of ROSS here.
Note: I used ChatGPT to help me extract information from the briefs and create these summaries.
1. American National Standards Institute (ANSI) and Six Standards Organizations
Amici Identities
- American National Standards Institute (ANSI).
- ASTM International (American Society for Testing and Materials).
- International Association of Plumbing & Mechanical Officials (IAPMO).
- International Code Council (ICC).
- National Electrical Manufacturers Association (NEMA).
- National Fire Protection Association (NFPA).
- ULSE Inc. (UL Standards).
Background
These are private standards-development organizations (SDOs) that create technical codes and standards used across industry and often incorporated by reference into law. They emphasize that standards are costly to develop and funded largely through copyright-based licensing and sales of standards documents.
Supports
Thomson Reuters.
Principal Arguments
- Headnotes are like standards: TR’s headnotes are privately authored, creative works that help professionals navigate complex legal material, analogous to SDO technical standards that help industry navigate complex technical requirements.
- Government edicts doctrine is narrow: Amicus argue ROSS misreads Banks and Georgia v. Public.Resource.Org; the “government edicts doctrine” is “a straightforward rule based on the identity of the author,” and applies only to works authored by judges/legislators in their official capacities, not to private commentary such as headnotes.
- Supreme Court precedent already answers this: Callaghan v. Myers and Georgia confirm that headnotes and similar explanatory texts created by private parties remain copyrightable, even when closely tracking judicial text.
- Policy concern: If ROSS’s rule were adopted, it would jeopardize the business model of SDOs whose privately authored standards sit “side-by-side with government-authored works” and are frequently incorporated by reference into law.
Counsel for Amicus
- Stanley J. Panikowski, J. Kevin Fee, Jane W. Wise, DLA Piper LLP, San Diego and Washington, D.C.
Nugget Quote
“Like standards authored by standards development organizations, Thomson Reuters’ headnotes are creative, privately-authored works that advance public knowledge by translating complex material into structured, comprehensible guidance. These are precisely the kinds of expressive works that copyright protects.”
2. Copyright & IP Law Professors
Amici Identities
Fifteen copyright and IP professors:
- Sandra Aistars, Distinguished Counselor in Residence & Professor of Law, IPPI: The IP Policy Institute, The University of Akron School of Law.
- Robert Brauneis, Michael J. McKeon Professor of Intellectual Property Law, Co-Director of the Intellectual Property Program, The George Washington University Law School.
- Megan Carpenter, Dean and Professor of Law, Franklin Pierce School of Law, University of New Hampshire.
- Kevin Casini, Lecturer, Quinnipiac University School of Law.
- Jon M. Garon, Associate Dean for Technology and Innovation, Director, Goodwin Program for Society, Technology, and the Law, Shepard Broad College of Law, Nova Southeastern University.
- Timothy T. Hsieh, Associate Professor of Law, Oklahoma City University School of Law.
- Joshua Kresh, Research Professor & Executive Director, The IP Policy Institute, the University of Akron School of Law.
- Jake Linford, Associate Dean for Academic Affairs, Loula Fuller & Dan Myers Professor, Florida State University College of Law.
- Philippa Loengard, Executive Director of the Kernochan Center for Law, Media and the Arts and Lecturer in Law, Columbia Law School.
- Loren E. Mulraine, Professor of Law, Director of Music and Entertainment Law Studies, Belmont University College of Law.
- Sean A. Pager, Professor of Law; Associate Director, Intellectual Property, Information & Communications Law Program (IPIC), Michigan State University College of Law.
- Eric Priest, Professor of Law, University of Oregon School of Law.
- Zvi Rosen, Associate Professor of Law, Franklin Pierce School of Law, University of New Hampshire.
- Mark F. Schultz, Goodyear Tire & Rubber Company Endowed Chair in Intellectual Property Law, Faculty Chair, IPPI: The IP Policy Institute, The University of Akron School of Law.
- Bhamati Viswanathan, Fellow, Columbia Law School, Kernochan Center for Law, Media and the Arts, Senior Visitor, University of Cambridge Faculty of Law.
Background
These are academic copyright scholars from law schools across the U.S. who study fair use, AI, and markets for expressive works.
Supports
Thomson Reuters.
Principal Arguments
- Fair use remains holistic and fact-specific: Amicus stress that fair use is a “case-by-case” analysis focusing on substitution, and that ROSS bears the burden of justifying its copying.
- Training AI to perform the same function is not transformative: When AI is trained “to perform the same function as the original work” – here, directing users to relevant portions of judicial opinions – the first factor favors the plaintiff.
- Non-expressive / intermediate use arguments misstate the law: ROSS’s claim that it engaged only in “nonexpressive” or intermediate copying ignores that copyright protects reproduction even if the output is not infringing, and that one cannot justify copying expressive works simply by saying the goal was to reach the unprotected “facts.”
- Oracle, Sega, and Sony are distinguishable: Those cases involved functional computer code and necessary interoperability, not expressive editorial text, and do not create an AI-specific carveout.
Counsel for Amicus
- Felicity S. Kohn, Pryor Cashman LLP, New York, NY.
Nugget Quote
“It makes no difference that the headnotes express facts (albeit creatively); if all defendants could excuse their unauthorized copying by claiming to be seeking the facts contained therein, the exception would swallow copyright law whole. Nor, as Ross and its amici argue, is its use fair because it engaged only in ‘intermediate copying,’ creating no infringing output. The law is clear that copyright holders have the right to control reproduction, irrespective of whether the separate rights to distribute the work or create derivative works therefrom are infringed.”
3. News/Media Alliance (N/MA)
Amicus Identity
News/Media Alliance (N/MA) – a nonprofit trade group representing over 2,200 news and magazine publishers in the U.S.
Background
N/MA represents newspapers, digital-only outlets, and magazines ranging from local papers to national brands. Its members depend on copyright to fund journalism, especially as they negotiate AI and RAG (retrieval-augmented generation) licensing deals.
Supports
Thomson Reuters.
Principal Arguments
- Unauthorized copying to build a competing product is not fair use: The district court correctly held that Ross’s use of more than 2,000 headnotes to build a rival research platform was not fair use, and that “one may not copy another’s copyrighted content to produce a product that seeks to displace the market for that very content.”
- Parallel with news scraping: Amicus warn that if this conduct is blessed, AI developers could treat news reporting the same way—scraping it to build directly competing AI news tools and invoking fair use.
- RAG and live news: Because LLMs are static, they rely on RAG that scrapes up-to-date publisher content. Treating that as fair use would divert traffic and revenue away from publishers and undermine a rapidly growing AI licensing market.
- Market harm and Warhol: Amicus lean heavily on Warhol’s rule that using a work for the same commercial purpose as the original is incompatible with fair use, especially where it usurps existing or potential licensing markets.
Counsel for Amicus
- Regan A. Smith, News/Media Alliance, Arlington, VA.
- Jacqueline C. Charlesworth & Nicholas M. Medellin, Frankfurt Kurnit Klein + Selz PC, Los Angeles, CA.
Nugget Quote
“News companies rely on revenue from the stories they publish to pay the journalists who research and write those stories. When news reporting or other copyrighted material is taken and used without permission to develop a competing product — as Ross Intelligence Inc. did here — it undermines the ability of publishers to create and disseminate original content, including quality online journalism.”
4. Copyright Alliance
Amicus Identity
Copyright Alliance is a nonprofit public-interest organization representing over two million individual creators and 15,000 organizations across music, publishing, visual art, film, software, etc.
Background
The Alliance advocates for strong copyright protections and enforcement as essential to the livelihoods of creative professionals and to the U.S. copyright industries’ contribution to the economy.
Supports
Thomson Reuters.
Principal Arguments
- Headnotes easily clear Feist’s originality bar: Amicus argue the West headnotes are independently created by attorney-editors who select, organize, and phrase legal points, making them “quintessentially copyrightable.”
- Narrowing originality would harm many industries: Weakening protection for selections/arrangements of factual material would devastate creators in photography, fashion, databases, scientific publishing, and more, who rely on compilation-based copyright.
- ROSS’s fair-use arguments misapply Google Books, Arriba, and Oracle: Those decisions involved distinct, functional uses (search, thumbnails, interoperability) and were “boundary-testing” cases that do not justify copying to build a direct competitor.
- AI licensing is a growing, legitimate market: AI training is already being licensed at scale (e.g., content deals with OpenAI, Microsoft, and others), and fair use should not be stretched to undermine this emerging revenue stream for rightsholders.
Counsel for Amicus
- Nancy E. Wolff and Elizabeth Safran, Cowan, DeBaets, Abrahams & Sheppard LLP, New York, NY.
Nugget Quote
“Appellant’s activity threatens wide swaths of creative industries and the livelihoods of authors and book publishers, as well as many other types of creative professionals. Permitting Appellant to take and reproduce Appellees’ copyrighted headnotes without authorization for purposes of a competing product would undermine existing licensing markets and strip creators and rightsholders of their statutory rights to control and commercialize their copyrighted works.”
5. Major Film Studios (Disney, Paramount, Sony, Universal, Warner Bros.)
Amici Identities
- Disney Enterprises, Inc.
- Paramount Pictures Corporation
- Sony Pictures Entertainment Inc.
- Universal City Studios LLC
- Warner Bros. Entertainment Inc.
Background
These are five of the largest U.S. film/television studios, each owning extensive libraries of copyrighted works and heavily invested in both content creation and new technologies. They frequently litigate and rely on fair use but worry about over-expansion of the doctrine in the AI context.
Supports
Thomson Reuters.
Principal Arguments
- No AI-specific fair-use rule: Amicus urge the court to reject any “AI exception,” arguing that fair use already accommodates new technologies and should not be rewritten to make AI training presumptively transformative.
- ROSS’s use is classic market substitution: Ross used the headnotes as a “shortcut” to build a competing legal research product serving the same function and customers as Westlaw – precisely the kind of substitution Campbell, Warhol and Video Pipeline disfavor.
- Intermediate copying doesn’t save them: Amicus argue that “intermediate copying” cases (Sega, Sony, Oracle) involved functional code and interoperability, not expressive text, and required necessity. Ross could have created its own headnotes from public cases but chose not to.
- Licensing and potential markets: Fourth-factor analysis must consider Westlaw as the copyrighted work and the emerging derivative market for AI training on editorial datasets. Ross’s refusal to pay for a license and subsequent copying is classic market harm.
Counsel for Amicus
- Adam G. Unikowsky and Jonathan J. Marshall, Jenner & Block LLP, Washington, D.C.
- David R. Singer & Julie A. Shepard, Jenner & Block LLP, Los Angeles, CA.
Nugget Quote
“[W]hile Defendant seeks shelter behind the fact that its competing product involves purportedly novel machine-learning technology, that does not alter the copyright principles in play. There is no ‘novel technology’ exception to copyright law, and there is no thumb on the fair-use scale for commercially motivated copying just because the competing product in some sense incorporates AI.”
6. Center for Art Law
Amicus Identity
The Center for Art Law is a Brooklyn-based 501(c)(3) nonprofit focused on the intersection of visual art, cultural heritage and law.
Background
Founded in 2009, the Center researches and educates on artists’ rights, AI and copyright, restitution, and cultural property. It represents visual artists’ interests in emerging AI copyright issues.
Supports
Thomson Reuters.
Principal Arguments
- Reject a blanket AI fair-use exemption: Fair use is an “equitable rule of reason,” and courts should not carve out a categorical exemption for AI training; each use must be analyzed under the four factors.
- AI training can seriously harm artists: Unauthorized ingestion of images and visual art for training causes concrete harm to creators—loss of autonomy, licensing opportunities, and value of their styles—especially when outputs mimic original works.
- All four factors usually cut against training on creative works:
- Factor 1: Mere ingestion for pattern recognition is rarely transformative.
- Factor 2: Visual art is at the core of copyright protection.
- Factor 3: Training typically copies entire works at scale.
- Factor 4: AI outputs can function as substitutes, diluting markets for commissioned and licensed works.
- Comparative perspective & licensing: The brief notes EU TDM rules and growing licensing models (Adobe Firefly, Shutterstock, Getty, News Corp–OpenAI) as proof that innovation and rights-respecting licensing can coexist.
Counsel for Amicus
- Irina Tarsis, Center for Art Law, Inc., Brooklyn, NY.
Nugget Quote
“[T]he unauthorized use of copyrighted visual works in AI systems poses significant and concrete harm to artists. Illustrators, photographers, painters, and designers face infringement of their creative autonomy, threats to licensing and market opportunities, and loss of the value of their labor and distinctive styles when their works are incorporated into AI datasets without consent, attribution, or compensation.”
7. AI Coalition for Data Integrity (AICDI)
Amicus Identity
The AI Coalition for Data Integrity is a multi-stakeholder coalition of AI developers, content owners, publishers, trade associations and others focused on transparency, attribution, and licensing in AI data practices.
Background
AICDI promotes ethical AI by advocating for proper data stewardship, including lawful licensing of training data. Several member organizations (writers’ guilds, SAG-AFTRA, music and image licensing groups, Yelp, etc.) are listed as signatories to the brief.
Supports
Thomson Reuters.
Principal Arguments
- Warhol and factor one: Under Warhol, commercial uses serving “substantially the same purpose” as the original weigh sharply against fair use. Ross used headnotes to build an AI research tool that serves the same legal research function as Westlaw.
- Three-fold market harm under factor four:
- Direct substitution: Ross’s platform is marketed as a competitor to Westlaw, usurping TR’s core market.
- Licensing & derivative markets: Ross’s copying bypassed an established licensing market for editorial datasets and training data; Ross even sought a license, was refused, then used the data anyway.
- Market dilution: AI systems can flood the market with substitutes; AICDI leans on Kadrey’s acknowledgment that dilution can decisively defeat fair use when evidence is present.
- Critique of Kadrey and Bartz: Amicus argue those decisions misapplied Warhol by treating AI training as inherently transformative and downplaying licensing and dilution harms; they urge the Third Circuit not to follow that path.
- Protecting the AI-licensing economy: Affirming the district court will reinforce copyright’s incentive structure and the emerging, legitimate market for AI-training licenses.
Counsel for Amicus
- Tod Cohen and Emily Whitely, Manatt, Phelps & Phillips, LLP (San Francisco & New York).
Nugget Quote
“Ross’s conduct … threatens the rapidly developing market for licensing copyrighted works as AI-training data. By ignoring a refused license and taking Thomson Reuters’s proprietary content to power a direct competitor, Ross undermines existing licensing practices and accelerates market dilution, which are both core harms under the Copyright Act’s fair use fourth factor.”
8. Jonathan Iwry, Fellow, Wharton Accountable AI Lab
Amicus Identity
Jonathan Iwry is a fellow at the Wharton Accountable AI Lab (University of Pennsylvania).
Background
Iwry researches AI governance and accountability, especially how AI intersects with foundational legal concepts such as copyright, liability, and fair use. He appears as an individual scholar, not on behalf of Wharton.
Supports
Thomson Reuters.
Principal Arguments
- AI must not become a loophole for appropriation: If copying expressive works “at commercial scale” and diffusing them into model weights were automatically insulated from liability, any competitor could ingest thousands of works and claim non-infringement, undermining copyright’s incentives.
- Headnotes are expressive, not mere facts: Editorial judgments in headnotes—selection, framing, emphasis—are expressive under Feist and should not be redefined as “facts” simply because they concern legal holdings.
- True purpose, not intermediate technical steps, governs factor one: Converting text into vectors is just another encoding; training an AI is not a distinct purpose when the ultimate goal is a competing commercial research tool.
- Market harm and AI accountability: Allowing this sort of copying would erode incentives for editors and researchers and encourage AI developers to “rely on opacity and litigation risk rather than legitimate acquisition of data.” Affirming TR would promote accountability in AI development.
Counsel for Amicus
- Jonathan Iwry, Potomac, MD (appearing pro se as amicus).
Nugget Quote
“The internal processing of expressive content by an AI system does not thereby count as transforming that expressive content for fair use purposes. The model’s encoding of the headnotes’ content might take a numerical form, but what it encodes is still the rightsholder’s expressive judgments and organization, which ROSS copied to provide a substitutive service.”
9. RELX Inc. (LexisNexis)
Amicus Identity
RELX Inc., doing business as LexisNexis, one of the two dominant U.S. legal research platforms.
Background
LexisNexis is a global legal-information and analytics provider and the principal competitor to Westlaw. It creates and licenses:
- Case law databases
- Legal commentary
- Topical taxonomies
- LexisNexis Headnotes, which it argues are expressive works protected by copyright
LexisNexis notes that it owns its own copyright registrations for headnotes and that it uses a team of attorney-editors who synthesize case law and create summarizing annotations.
This gives the company a strong competitive and doctrinal interest in the outcome: if Westlaw’s headnotes were declared uncopyrightable, LexisNexis’s competing headnotes would be equally jeopardized.
Supports
Thomson Reuters.
Principal Arguments
- Headnotes are copyrightable under long-standing precedent. LexisNexis emphasizes that Callaghan v. Myers and Georgia v. Public.Resource.Org confirm headnotes created by private publishers are protectable expression. It cites a long history of courts treating annotations as expressive works.
- Editorial headnotes serve essential public-facing functions. Lexis describes headnotes as the “bridge” between complex case law and the public—judges, lawyers, students, journalists, and pro se litigants.
- Headnotes require sustained human creativity. The brief outlines the labor-intensive editorial process, in which attorneys: 1. read full opinions, 2. identify and paraphrase key legal issues, draft accessible summaries, apply proprietary taxonomies, and cross-link to citators and research tools. Because LexisNexis and Westlaw often produce different headnotes from the same case, the brief argues, this proves headnotes are not dictated by convention or mere fact.
- Eliminating protection would collapse the market. If the court ruled headnotes uncopyrightable, the economic foundation that supports continuous editorial creation would collapse. Among the predicted consequences: publishers would stop producing headnotes; research platforms would degrade; errors and doctrinal drift would increase; and the harm would fall hardest on small firms, courts, legal aid and pro se litigants. LexisNexis warns this outcome would result in “fewer, lower-quality options” in legal research.
- Fair use does not excuse ROSS’s appropriation. Lexis argues the fourth factor — market effect — decisively weighs against ROSS, in that: ROSS’s copying targeted the core market for editorial annotations, it deprived publishers of licensing revenues and emerging AI-training markets, and allowing this would privilege “copyists over creators” and remove incentives for legal-editorial investment.
Counsel for Amicus
All from Troutman Pepper Locke LLP: Michael D. Hobbs, Jr., Atlanta, GA; Austin Padgett, Atlanta, GA; and Brooke R. Watson, Charlotte, NC.
Nugget Quote
“Editorial headnote annotations provide a public-facing bridge between complex judicial opinions and the communities who must understand and apply them: judges, practitioners, students, journalists, and pro se litigants. … By distilling controlling holdings, locating issues within an established taxonomy, and articulating rules in accessible prose, headnotes reduce the time, cost, and error associated with legal research. This lowers barriers to entry for smaller firms and self-represented parties, promotes more uniform application of the law, and
enhances courts’ efficiency by directing them to the precise passages that matter most.”
Robert Ambrogi Blog