41148
post-template-default,single,single-post,postid-41148,single-format-standard,stockholm-core-2.4,qodef-qi--no-touch,qi-addons-for-elementor-1.8.3,select-theme-ver-9.5,ajax_fade,page_not_loaded,,qode_menu_,wpb-js-composer js-comp-ver-7.9,vc_responsive,elementor-default,elementor-kit-38031
Title Image

The Future of AI Lawyers is Still Far-Off

The Future of AI Lawyers is Still Far-Off

Artificial Intelligence proponents have frequently spoken about a future where the huge workforce of highly-paid attorneys is replaced with AI, which can conduct the legal research attorneys usually devote hours to in a much faster manner.[1] A recent ruling within the U.S. District Court for the District of Delaware is a huge blow to this vision of the legal field.[2]

Thomson Reuters owns Westlaw, an online database of judicial decisions that lawyers use to research legal problems.[3] When decisions are given by courts, Westlaw digitizes and uploads them to its database, along with headnotes, which are summaries of the main legal points, and Westlaw’s Key Number System, which is a system that enables lawyers to find cases pertaining to specific legal topics.[4] AI-start-up Ross Intelligence asked Westlaw to use its content, its collection of decisions as well as its headnotes and Key Number System, to train its legal research model.[5] Ross’ tool would essentially replace the need for Westlaw, and minimize the time needed to conduct research, thereby eliminating the number of lawyers needed.[6] Westlaw of course, said no.[7]

Ross then commissioned 25,000 memos from a third-party legal research company, LegalEase Solutions, to serve as this training content.[8] However, LegalEase’s content closely resembled the language of Westlaw’s headnotes and Key Number System.[9] Upon discovering this alleged similarity, Thomson Reuters filed suit against Ross alleging copyright infringement.[10]

The court found that both Westlaw’s headnotes and the Key Number system were entitled to copyright protection, and that there was striking similarity and sufficient access to find copying in fact, and substantial similarity of copyrightable matter to find copying in law.[11] Ross, however, put forward a fair use defense.[12] The court rejected this defense, finding for Thomson Reuters and through this, sent a message towards future and pending cases debating AI’s fair use argument.[13] Fair use is the legal doctrine in copyright law where limited copying of a work is allowed depending on the characteristics of the use of the material, even without the permission of the work’s owner.[14] If AI start-ups were allowed to claim fair use towards copyrighted works like Westlaw’s content, these start-ups could train their tools without having to license these works from their owners.[15] Courts’ acceptance of the fair use defense for AI training has inevitably become the lynchpin to AI’s continued progress; without it, AI would develop slower and be more costly.[16]

The court found that Ross’ use was commercial and under Andy Warhol Foundation for the Visual Arts, Inc. v. Goldsmith, was not sufficiently transformative.[17] Warhol established that use being transformative is not enough to qualify for fair use – there must be a certain degree of transformative-ness, so that two uses with the same or similar purpose would weigh against a finding of fair use.[18]

The use was also for exactly the same purpose as Westlaw’s use, therefore failing the first (and frequently deciding) fair use factor.[19] The market effect of Ross’ tool was emphasized by the court, because Ross’ aim was to create a competing product to Westlaw. [20] The aim of Ross’ tool is indisputably to act as a substitution for Westlaw, a fact pattern that is similar to many other AI tools trying to claim this fair use defense.[21] After the fair use factor analysis, the court granted summary judgment to Thomson Reuters. [22]

This decision is the latest addition to the many cases that are shaping the laws around AI-development, and its potential to replace certain members of our human workforce.[23]While AI defenders might put forward other legal theories in court, fair use was certainly a “key defense” and a difficult loss for the development of AI.[24] If more courts follow suit, the future that AI-companies promised seems far-off, if not unlikely.

Footnotes[+]

Isabella Herrera

Isabella Herrera is a second-year J.D. candidate at Fordham University School of Law and a staff member of the Intellectual Property, Media & Entertainment Law Journal. She holds a B.A. in Communication and Media Studies from Vanderbilt University.