AI & Law: Legal Simplification

Image for post
Image for post
Using AI to simplify the law, but the outcome might not be desirable

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:

  • An ongoing drumbeat is whether the laws are overly convoluted and need to be simplified
  • Simplifying the laws might provide upsized benefits for the public and the practice of law
  • Not everyone is convinced that law simplification is an answer and might worsen matters
  • AI has the potential for providing a type of legal simplification engine
  • There are qualms that the AI simplifying machinations could bode for legal difficulties


There is an ongoing tension between the law as necessarily being complex and unwieldy, perhaps innately so, versus the belief that the law can inexorably be greatly simplified and streamlined. The public at large would presumably be more amenable to observing the laws if the arcane morass of the laws were straight-and-narrow rather than existent as bloated and altogether confounding. In theory, the courts too would be better off, and the practice of law might be less obtuse were it not for the existing oblique nature of our laws.

Not everyone though might hold that same view of seeking heightened simplicity. Recall the famous words of Montesquieu (1748) as uttered in De l’Esprit des Lois: “Thus when a man takes on absolute power, he first thinks of simplifying the law. In such a state one begins to be more affected by technicalities than by the freedom of the people, about which one no longer cares at all.”

For details on this and other AI and law topics, including sample code excerpts and salient explanations, see my textbook entitled “AI and Legal Reasoning Essentials” at this link here:

In the quote by Montesquieu, there is a raised concern of an authoritarian state taking on absolute power and opting to subsequently simplify the law. This simplification, in turn, might lead to law that no longer affords the freedoms that we enjoy and could become stagnant and stifling in their effects.

Depending upon how one wishes to interpret the matter, it is plausible to argue that the pursuit of simplicity in of itself is not necessarily a best-fit aspiration and nor always proffers a beneficial outcome. That’s not to denigrate the value of simplicity and only to offer food for thought that we cannot assume axiomatically that if laws were simpler it ergo portends that we would be better off. What could presumably be said, at the very least, is that the law is simpler than it once was.

There is also the matter of the chicken versus the egg in this conundrum.

Perhaps Montesquieu is merely pointing out that those wishing to ultimately take control will tend toward simplifying the law, greasing their devious and strident efforts to overtake society and summarily dictate the nature of our laws. The notion of simplicity is inadvertently drawn into this maneuvering of power tectonic shifts and otherwise is a standalone topic that unfairly is being dragged through the mud. In a sense, simplicity has nothing or little to do with the power dynamics per se, neither promoting it and nor inhibiting it and merely an unsuspecting tool for (in this instance) untoward ends.

Questions also arise at the endpoint of this quest for simplicity. In essence, can one continue to split the atom indefinitely, or is there a stopping point at which the reach of simplicity can go no further? In the use case of the law, the simplest linguistic utterance might be elusive and indeterminate. Endless debates about trying to squeeze more blood from a rock, aiming to make the laws simpler and simpler, could be distracting and unwieldy of themselves.

Practitioners might find this discourse on the simplicity versus the complexity of the law as rather philosophical and less so applied or valued in the actual practice of law.

AI And The Simplification Engine

Let’s shift gears and toss the advent of Artificial Intelligence (AI) and the law into the mix to see if we can get some grounded traction on the matter.

One belief is that the use of AI will inevitably simplify the law.

Here’s how.

It is customarily assumed that to achieve any true semblance of AI and the law, the use of AI techniques and technologies such as Machine Learning (ML) and Deep Learning (DL) would need to scan the text of our laws and pattern-match at the lowest level of tokenization. This would likely need to be augmented by human lawyers that aid in the labeling of these minimalist linguistic participials and provide by hand guidance for the ML/DL during those AI “learning” efforts.

Inexorably, this mechanization would produce simpler and simpler laws or at least AI-embodied laws that were transmuted from our everyday present laws.

If we then began to utilize the AI to provide legal reasoning and aid in judicial matters, a type of slippery slope will then have been intrinsically traversed. The now inner coded simpler variants of the laws by the AI would become the de facto set of laws. Eventually, there would seem to be little need for and nor reliance upon the laws on the books, as it were, and instead the AI-based and derived simpler laws would rue the day.

Thus, we fed into the AI the existent arcane and complicated laws, and out pops the simpler codified set. Voila, while trying to achieve artificially intelligent systems for use as automated and autonomous lawyers, judges, and the like, we just so happened to hit two birds with one stone, having generated simpler laws in the same breath.

Okay, based on that scenario, we will ultimately then land upon the shores of simpler laws.

Is that a good thing or a bad thing?

Well, for those asserting that simpler is always better, this must be a good thing. We got lucky, some would say, achieving a twofer by hence having AI for autonomous legal reasoning and simultaneously transformed the byzantine present-day laws into a simpler collective that is available on-line at any time from any location on the planet.

What’s not to like about this, they would exhort.

Consider some of the counter-arguments.

If the laws of today were divided and subdivided into simpler pieces, being churned through in a sausage factory manner, the voluminous nature would worsen and make it even harder for lawyers to practice law. Presumably, human attorneys would not seemingly be able to keep mentally attune to the millions upon millions of such simplified laws that might be so derived.

In that sense, whereas it might appear that the law is being simplified, the reality would be that it is becoming increasingly complex by a classic forest-for-the-trees kind of difficulty (more and more trees obscuring the overarching shape of the forest).

Imagine too the permutations and explosive combinations that could arise.

Each of the itsy-bitsy laws is undoubtedly going to be related to many others of the same ilk. A particular simplified law might find itself interconnected with dozens of others of these simplifications, or more likely to hundreds, thousands, or millions of them. As a crude analogy, it is said that the human brain has about 100 billion neurons, which is certainly a large number. That being said, the estimates are that there are perhaps 100 trillion interconnections. The number of interconnections is many times that of the actual number of neurons, and indeed it is generally believed that the basis for our mental prowess is as much due to the interconnections as much as it is due to the neurons themselves.

All told, the viewpoint about AI and the law by some is that if indeed this simpler mantra is going to be the core foundation for how autonomous AI-based legal reasoning will be achieved, we might be opening up a Pandora’s box and not even realize that we are doing so.

Per any topic that falls within the realm of the law, please know that there are cantankerously extensive debates on this topic and numerous points and counterpoints of the prophesized Dystopian result. For example, it could be the exact opposite of the aforementioned computational rendering, such that the AI embodiment converts the laws into inherently complex magnums and does not generate the simplifications that some assume will arise. In that case, if you believe that complexity is our friend, we would be thankful for the AI providing such a mechanism.


Yet another perspective is that it won’t matter whether the AI serves as a grist mill and turns laws into being simpler or more complex since we humans won’t care. We’ll treat the AI as a veritable black box, having it undertake all of our legal wrangling's. No more human lawyers or human judges. Just the AI. We’ll be satisfied with whatever internal makes-the-hamburgers kind of intricacies exists, not being interested in looking under-the-hood.

That seems a bit farfetched and indubitably would spark Montesquieu, if alive today, toward worrying that the AI might undertake absolute power and, as we all know, lead to absolute corruption (per the wise words of Lord John Emerich Edward Dalberg-Acton).

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.

Written by

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