AI & Law: Legal Argumentation

Exploring the role of AI in legal argumentation

by Dr. Lance B. Eliot

For a free podcast of this article, visit this link or find our AI & Law podcast series on Spotify, iTunes, iHeartRadio, plus on other audio services. For the latest trends about AI & Law, visit our website

Key briefing points about this article:

  • One might boldly suggest that humans are essentially biologically-based arguing machines


There seems to be something innate in humanity that we are able to argue. Indeed, it would seem that we oftentimes delight in arguing. As with most salient aspects of our various cognitive acts, there are problematic considerations when trying to distinctly ascertain whether our compulsion to argue is due to DNA, or whether it is a skill that we acquire during our lifetimes. In short, arguing might be imputed via nature or nurture, or perhaps an inseparable combination of the two.

There are many contexts in which arguing and the role of argumentation arise. In the case of the law, there are voluminous treatises about the form and formulation of legal argumentation. Scholars have attempted to dissect the very fabric upon which legal argumentation rests. Law schools stridently seek to guide budding attorneys toward the bright light of refined and highly seasoned legal argumentation.

Let it be said, if not overly obvious, that legal argumentation is crucial to our adversarial form of justice.

Cooked into the very stew of justice and ingrained indelibly throughout the judicial process, we use legal argumentation as a means to seek intrinsically optimal or sufficiently best-possible outcomes. Presumably, the more robust that competing legal arguments can go head-to-head, the greater the chances of arriving at a just conclusion. If one side proffers a weaker legal argument, the assumption is that it will fail to convince, and thus the stronger legal argument will prevail.

Where can computers and especially the advent of Artificial Intelligence (AI) come to play in this daily and ongoing battle utilizing legal argumentation?

This is an important question and bodes for the future of the law, given that the practice of law will inevitably imbue the application of AI-powered legal reasoning systems.

Legal Argumentation And The CARE Model

In my research on AI and legal argumentation, I’ve identified four keyways that legal argumentation is being augmented via advanced automation. These four essentials are coined via the convenient acronym of CARE, entailing Crafting, Assessing, Refining, and Engaging. These distinct actions represent the overall means by which LegalTech experts and legal scholars are making use of or attempting to infuse AI Legal Argumentation (AILA) capabilities into the practice of law and the furtherance of the legal profession.

Let’s briefly consider each of the four key actions.

The first of the noted actions or activities involves crafting a legal argument.

When a seasoned lawyer initially takes on a case, they tend to instinctively begin mulling over the legal position that they anticipate undertaking, thusly sifting through the elements of legal arguments that will underlie the case. This eventually blossoms and maturates into a larger overarching collection of legal arguments which are then coalesced into a cohesive whole.

For newbie lawyers, there is oftentimes a substantive chance that the legal arguments being brought together are unable to readily be fashioned into a unified whole (you might recall in law school struggling with composing your IRAC, describing the Issue, Rule, Application, Conclusion). At times, the novice lawyer cannot ferret out the forest for the trees. Also, they might have anchored themselves onto legal arguments that are potentially irrelevant to the case or that otherwise distract from the true core of the matter at hand. That’s not to suggest that even experienced lawyers cannot fall into the same mental traps, namely that they can, and do so, but there is usually a stronger probability that they will catch themselves before the fashioning gets too far along.

If possible, efforts to construct a robust legal argument are oftentimes bolstered by seeking the opinions and review of trusted colleagues. Presumably, by sharing a coalescing legal argument with fellow lawyers and partners, it is expected that any gaps, loopholes, or innate argumentation failings will be detected and resolved. The overarching goal is to produce a maximally vigorous legal argument, particularly since the opposing side will undoubtedly be aiming to find any otherwise undiscovered gotchas. Better to identify your own argumentation weaknesses beforehand, rather than having your head handed to you by the opposing side.

This is where the use of AI especially comes to the fore.

Various AI-enabled legal argumentation tools are being built and deployed to aid in serving as an over-the-shoulder aid when assessing the strength of a proposed legal argument.

Such AI attempts to ascertain whether there are omissions or inconsistencies in the legal argument logic. By flagging those potential problems, the human lawyer can be sparked into realizing any logic traps that they have inadvertently fallen into and seek to rectify those before proffering their legal argumentation to others.

Even more advanced is the capability of the AI for refining legal arguments.

Rather than simply identifying or assessing where legal arguments might be deficient, the refinement capability provides hints or suggestions of what ought to be improved. In fact, if extensively advanced, the AI offers outright indications of the best possible refinements to be made.

Lastly, in terms of the CARE aspects, there is the engaging facet, consisting of an interactive dialogue of the AI in conjunction with the human lawyer, allowing for real-time Natural Language Processing (NLP) dialogue and debate about the legal argumentation. In a sense, this can act as a practice session, allowing the lawyer to tryout the legal arguments and see what responses the opposing side might attempt, though doing so in a simulated environment and without the risks associated with toe-to-toe opponent skirmishes.

For my in-depth research paper on this topic, see the paper entitled “AI And Legal Argumentation: Aligning The Autonomous Levels Of AI Legal Reasoning” at this link here:

Today’s Reality Of AI-Powered Legal Argumentation

Before you whisk out your credit card to purchase any such AI-enabled legal argumentation toolset, be aware that most of these efforts are still quite exploratory and tend to be exceedingly research-oriented. Let it be known that creating a good legal argument is indubitably hard work and getting AI to do so is devilishly difficult. Assessing a legal argument is also onerous, being quite problematic for today’s AI capabilities. The same can be said for AI that attempts to refine legal arguments. Perhaps especially difficult is carrying on a fluid and seemingly informed dialogue or debate, the engaging activity, as it entails direct AI-to-human interaction and, as might be evident from today’s Siri and Alexa, there is still a long way to go before NLP is truly proficient and fluent.

The application of AI to legal argumentation will gradually be improved (there seems little doubt about that trend). Indeed, some are worried that eventually, AI will take over legal argumentation entirely, as though we might end-up with AI arguing against AI in our courtrooms, rather than human lawyers arguing with fellow human lawyers. This is a rather futuristic prophecy and you can stop losing sleep at night worrying about it, for now.

On a realistic bent, there is a much greater chance that we’ll sooner see the use of AI Legal Argumentation as a handy tool for lawyers, boosting their legal argumentation skills and readying them for trial and the deployment of their legal arguments. Thus, in the foreseeable future, AI for legal argumentation will be relatively mechanistic and serve as a spading tool for lawyers. This involves aspects such as storing catalogs and snippets of legal arguments, enabling their online access. To some degree, this also will encompass indications of which legal arguments tend to fit with others, and which are considered counterpoints.

Via the use of AI capabilities encompassing Machine Learning and Deep Learning, drafted legal arguments will be classified into categorizations indicating their potency and relevancy, along with aiding in predicting which logic-based attacks might be successfully employed against those arguments. Likewise, this will assist in finding serialized attack points, referred to as attack vectors, regarding the legal arguments of the opposing side.


Shifting gears, there’s an intriguing additional perspective regarding why it makes abundant sense to blend AI and argumentation.

If arguing is indeed a fundamental tenet of humans, you could assert that argumentation is crucial to human intelligence.

Keep in mind that the goal of those in the AI field consists of trying to make computers that can exhibit human intelligence, and therefore one can logically deduce that infusing argumentation capabilities into AI might demonstrably aid in pushing the technology closer and closer toward displaying human-like intelligent capacities. Interestingly, getting AI to be argumentative could be the missing link toward achieving AI’s ultimate aspirations. Furthermore, legal argumentation, in particular, turns out to be one of the richest sources of argumentation and ergo a focus of AI not solely due to the interest in law, but because legal argumentation raises the art form and to some degree the science of arguing.

Perhaps that might make you feel better the next time you make a misstep in a legal argument for a case, knowing that nonetheless you are perceived as a stellar argument-making and argument-professing human, for which AI is “envious” and aiming to catch up with you.

May all your legal arguments be fruitful and unassailable.

For the latest trends about AI & Law, visit our website

Additional writings by Dr. Lance Eliot:

And to follow Dr. Lance Eliot (@LanceEliot) on Twitter use:

Copyright © 2020 Dr. Lance Eliot. All Rights Reserved.

Dr. Lance B. Eliot is a renowned global expert on AI, Stanford Fellow at Stanford University, was a professor at USC, headed an AI Lab, top exec at a major VC.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store