Our nation’s laws are a mess.
Boom, drop the mic.
Yes, I regret to inform you that whether you know it or not, you are at this very moment and each day ahead incessantly faced with an imperiling game of law-breaking dodgeball.
The gambit goes like this. There might be an obscure law or an abundance of Byzantine laws on the books that you manage to innocently breach and had absolutely no idea or even a general conception that such laws somehow existed. Akin to a wholly frantic and bruising form of dodgeball, laws of all shapes and sizes are out there zipping around and waiting to strike this person or that person.
Some of the laws make sense, some don’t. Some are applied suitably; some are applied arbitrarily. Laws can come from just about any direction and catch you completely by surprise. You likely won’t see it coming at you and will undoubtedly be shocked, dismayed, and in great legal peril when struck by one.
Supreme Court Justice Neil Gorsuch’s newly released book entitled “Over Ruled: The Human Toll Of Too Much Law” vividly describes and decries the incontrovertible fact that we have a tsunami of laws and that on a daily basis additional laws are piling up on the beaches. The number of laws has grown to epic proportions. They are said to be here, there, and everywhere. Furthermore, our laws are often written on the fly or in the vaguest of ways, allowing for loosey-goosey interpretations, and providing ample imprudent opportunity to get you caught on a technicality. You are left to scratch your head that the seemingly innocuous or ambiguous wording applied to your circumstances and cannot imagine that your purported transgression was really a charge-worthy above-the-law imperilment.
What are we to do about this troubling trend?
I’ve got an answer that I’d ask you to consider.
A potential means of solving this legal morass could be the advent of modern-day generative AI and large language models (LLMs).
Surprised?
Perhaps so, but when I walk you through the judiciously astute logic involved, you might find yourself leaning into the hope that AI can aid humankind in overcoming the staggering legal conundrum that we have on hand. It is relatively straightforward as to how this can work. Humans have a problem that entails words and the wording of words. Humans can use state-of-the-art AI to solve the human-devised wording problem. Perfectly sensible and altogether plausible.
The devil is in the details, which I’ll be getting into momentarily.
For my ongoing readers and new readers, in today’s column I am continuing my ongoing and extensive analysis of AI and the law, see my comprehensive overview of the AI & Law topic at the link here.
My latest pieces on AI & Law have included ironclad reasons that lawyers, law firms, and law partners ought to be adopting generative AI for their legal efforts, see the link here, plus, I recently examined closely the American Bar Association (ABA) proclamation known as Formal Opinion 512 that was newly released and governs how lawyers are expected to use and not misuse AI in their law practices, see the link here.
An added twist of worthy importance is that generative AI can be used to guide lawyers toward proper uses of generative AI, as I explain at the link here. That’s right, AI can be a mechanism for keeping users of AI, such as lawyers, on a suitably straight and rational track.
Okay, on with today’s show.
What’s The Deal About Our Myriad And Confounding Laws
Let’s take a look at some key excerpts from the fascinating and vital new book “Over Ruled: The Human Toll Of Too Much Law” by Neil Gorsuch and Janie Nitze, Harper Publishing, August 6, 2024. While sharing with you the essentials, I will interweave my own take on what the concerns are and how we might deal with the problems afoot. After setting the stage, I will subsequently bring generative AI and LLMs into the big picture here.
First, the notable book indicates these high-level qualms about the status of our laws:
- “What does it mean for our nation’s promise of equal treatment when our laws are so numerous and so complex that only an affluent or connected few can navigate their way?” (ibid).
- “And what happens to the respect for law itself when the law no longer just reflects commonsense norms but includes unpredictable traps for the unwary?” (ibid).
- “This isn’t an academic work or a legal brief. It is a book of stories – stories about real people, their struggles to make their way in a world awash with law, and the toll on their lives and families.” (ibid).
The gist is that we are confronted with a double whammy.
Allow me to point out the whammies.
The number one whammy is that there are lots and lots of laws sitting out there. Checkmark. Besides an overwhelming number of laws, the second whammy is that those laws are often written in a manner that defies common sense. They are at times a ticking legal timebomb, just at any moment to be interpreted in some unexpected way and applied because the moment in time allowed the possibility, not because that was what the law was designed to do.
Coining this as unpredictable traps for the unwary stridently hits the nail on the head, as it were.
You might have cleverly deciphered that the title of the book phrases the situation with the concern that we are being overly ruled (the book title states “Over Rule” as two distinct words), a quip of a turn for the conventional use of the word overrule. Maybe a bit cheeky. Nonetheless, it emphasizes that the abundance and ambiguous laws open widely Pandora’s door to overly zealous uses and abuses of our nation’s laws.
The authors at length focus on detailed examples of how these legal endangerments have ensnared everyday people. I mention this point because you might at first assume that the expressed worries are merely a theoretical consideration. In theory, the vast, intersecting, and confounding laws could ensnare someone, but does that truly happen?
Yes, it does. And the odds are it is all going to get worse as more and more people fall into or are driven toward the same legal traps, coincident with more and more laws getting added to the books. A kind of legal law-emmeshing death spiral, of sorts. More laws, confusing and uncertain laws, will traipse along and swallow more everyday people into its unexpected and unrealized path. Handwringing will occur. Likely little will be done to upright the situation. The cruel world marches along.
Whew, disconcerting and quite disturbing.
You might be mulling this quandary and thinking to yourself that the people entangled are those who flout the law anyway. Those who live by the sword are destined to be legally nicked or possibly run through by the sword. Sorry to say, that’s not the big picture. Please don’t let your mind or society try to convince you that these issues are narrowly scoped and only deal with legal scofflaws.
We are all vulnerable and, frankly, sitting ducks.
Per these excerpts, consider the widespread predicament involved:
- “Often enough, men and women going about their lives with no intention of harming anyone are getting thwacked unexpectedly and at times haphazardly, by our multitude of statutes, rules, regulations, orders, edicts, and decrees.” (ibid).
- “Law has proliferated at the local, state, and federal levels alike.” (ibid).
- “Much of it now comes from administrative sources largely unresponsive to democratic elections. But plenty also comes directly from our elected representatives.” (ibid).
A crucial consideration is that laws arise in a multitude of forums.
You must keep one eye on federal laws. You need to keep your other eye on state laws. You must furtively glance all around to ensure that you also see the local laws that surround you. This is entirely happening in 3D, such that you are fully immersed and can’t just stare straight ahead. Looking to your left and right, and behind you, is needed too.
Laws aren’t the only source of legal bombardment. There are executive orders. There are regulations. On and on it all flows. You would need more than two eyes, perhaps dozens of eyes on your head, to swivel back and forth to detect and discern all the laws and regulations covering your day-to-day activities.
Real Numbers Tell The Real Story
If you want quantification and hard numbers regarding the bursting-at-the-seams going on, consider these excerpts:
- “At the most basic level, law in our country has simply exploded.” (ibid).
- “By 2018, the U.S. Code encompassed 54 volumes and approximately 60,000 pages.” (ibid).
- “Over the last decade, Congress has adopted an average of 344 new pieces of legislation each session. That amounts to about 2 to 3 million words of new federal law each year.” (ibid).
- “Even the length of bills has grown – from an average of around 2 pages in the 1950s to 18 today. That’s just the average. Nowadays, it’s not unusual for new laws to span hundreds of pages.” (ibid).
- “Meanwhile, by 2021, the Code of Federal Regulations spanned about 200 volumes and over 180,000 pages.” (ibid).
I’d like to proffer an additional perspective on these surging and bursting waters.
It might seem cynical but to some degree, these laws arise not solely due to a heroic desire to stop bad acts. One issue is that oftentimes lawmakers and regulators aim to tout their grand successes by pointing out how many laws or regulations they manage to pass during their tenure. It is a numbers game.
Look at what we accomplished, umpteen more laws or regulations. You ought to commend us for doing our job. Our job is to presumably punch out as many laws and regulations as we can come up with. The more, the merrier. This has little to say about the quality of the laws and regulations. The quality aspects are not as readily identified nor cited. The easiest metric is volume.
I’m sure that those skeptical remarks will be irksome to some. There indubitably are lawmakers and regulators working their hearts out to aid and protect society. One means to achieve this entails putting together and enacting more laws. We can be thankful for their tireless endeavors.
At the same time, there is truth to the old line that you don’t want to see what happens behind the scenes when sausage is being made. To pass some of these decrees, all manner of negotiations and compromises take place. The language used is shifted around, reworded, replaced, edited, and can get snarled and convoluted. The pace to compose laws and regulations is often extremely hectic. If you saw it happen, you’d probably think twice about ever eating sausage again, and likewise, be more dubious of the laws and regulations that end up in the books.
On top of this, the time-based pressure to get these catch-all rules out the door is immense. This tends to dispense with an abundance of quality control, omitting any sense of repeated double-checks and triple-checks. Questions that deserve rapt attention are given short-shrift. Is the language of the law or regulation tightly composed? Does it stay within its lanes? Are there other unintended ways to interpret the language such that the design is no longer being retained? Etc.
Put a mental pin on that topic since I’ll come back to this litany of concerns when we move into the AI-related considerations.
No Expiration Date Despite Laws Getting Stinky And Out Of Order
Continuing ahead, the insightful book gives some handy jabs about the lack of common sense underlying some of the laws that we face:
- “As you might imagine, much in this growing mountain of law isn’t exactly intuitive, either.” (ibid).
- “Did you know that it’s a federal crime to enter into a post office while intoxicated?” (ibid).
- “Or to sell a mattress without a warning label?” (ibid).
- “And if you’re a budding pasta entrepreneur, take note: by federal decree, macaroni must have a diameter between .11 and .27 inches, while vermicelli must not be more than .06 inches in diameter. Both may contain egg whites – but those egg whites cannot contribute more than two percent of the weight of the finished product.” (ibid).
This brings up something else about laws and regulations that we need to put on the table.
In a sense, laws don’t customarily have an announced and astutely pre-determined expiration date in the way that food at the grocery store does. You buy some packaged fruit and a warning on the label tells you that by such-and-such date the fruit is not any good anymore. Fair and square.
The same doesn’t especially happen with our laws, disquietingly so.
Laws often are put into place that are aimed at a point-in-time concern, which at that particular time is perhaps needed and justifiable. Meanwhile, the outside world marches on. The laws that pertained earlier are now still sitting on the shelf, gradually becoming outdated. Their unstated and unconsidered expiration date has come and gone. Toss out the expired and ungainly contrivance.
There is little incentive to go through the laws and purge ones that are seemingly out-of-date. The problem is that it takes time and effort (but with little or no reward), and second, it could be like trying to pull out a precarious domino in an obtusely devised stack of dominos.
You might genuinely be worried that if you pull out one seemingly no longer pertinent domino all the rest might come crashing down. It can be the same with our laws. If we remove this one outdated law, there could be other still current laws that by precedent or other basis rely upon that law. Might as well leave the law alone and be safer to assume it still needs to be kept in place.
Voila, the never-ending piling up of more and more laws. No garbage collection takes place. It is seen as risky and not rewarding. Lawmakers and regulators are more likely to be rewarded by increasing counts of passage, rather than decreasing counts (oops, cynicism again). The laws are like a molehill that becomes a mountain, which becomes a landscape filled with mountains.
Stacking up of our laws will pretty much occur eternally.
Zombie Laws That Keep Coming Back To Life
The assumption usually is that old laws are a no-harm no-foul artifact.
Follow me on this twisty road.
Historically, maybe people were doing this or that and had to be given a law to curtail those adverse actions. The world changes. The law is rarely ever invoked anymore because the world has changed and the precursors or assumptions underlying the law no longer generally occur. Okay, good, just leave the laws there since they aren’t needed per se, nor do the laws need to be expunged.
They are out of sight, out of mind. Nobody cares that they are still residing on the books. No harm, no foul.
That kind of works, sometimes, except for the clever or at times insidious resurrection of those dormant or overlooked laws. One crucial element of laws is that they are semantically ambiguous. This is AI-related in the sense that laws are based on words, and words are not precise. Words can have all kinds of meanings and interpretations. More on this shortly.
The crux here is that a law once aiming to have a said purpose of R, can be interpreted differently at a later time T, and given a new meaning that says it now connotes S. The old law gets retooled, but without changing a single word or letter. It is how we interpret the law that changes.
Leaving laws lying around on the books provides ample opportunity for legal tomfoolery by reimagining what the law means and when it should be applied.
I recently discussed the current emergence of lawfare, see the link here, namely using laws in ways not especially anticipated and that are used to target specific people. It is the classic line of show me the person and I will show you the lawbreaking crime they have committed. Our laws aren’t supposed to be used that way.
But they can be.
And, lamentedly, they are.
A final quick quote from the book that demonstrates the realization of our forefathers that laws can be used in dour and sour ways mentions James Madison’s famous comments from The Federalist Papers:
- “At the dawn of our republic, James Madison contemplated the dangers to individual freedom, equal treatment of persons, and respect for law itself when a nation’s laws are allowed to grow “so voluminous that they cannot be read, or so incoherent that they cannot be understood… or undergo such incessant changes that no man, who knows what the law is today, can guess what it will be tomorrow.” (ibid).
I believe this adequately introduces the issues at hand and we are ready to shift into AI mode.
Generative AI And The Means Of Solving Our Laws Problems
Kudos for slogging through the needed background on the status of our laws. I trust that you are of a mind that we do have a problem concerning our laws. Good, thanks.
What next?
Fundamental principles of problem-solving say that you must first acknowledge a problem if you have much hope of solving it. The problem has been stated and we are primed to find a solution.
My proposed solution entails generative AI and large language models (LLMs).
I will begin by setting the record straight about what today’s AI consists of. There are myths that need to be busted. I aim to show you that the solution of using AI is neither farfetched nor far off in the future. It is here and now.
Let me first set the record straight on one very significant facet. None of today’s AI is sentient. Sorry to break the jarring news to you. I mention this since there are lots of headlines that seem to proclaim or suggest otherwise. The AI that I will be discussing doesn’t have sentience (none do at this time, see my discussion about the angst over anointing AI with legal personhood, at the link here), and we don’t need sentience in our AI for the problem at hand to be solved.
AI is a mathematical and computational construct or mechanization that just so happens to often seem to act or respond in human-like ways. Be very careful when comparing AI to the nature of human capabilities, which for example I delicately cover and differentiate in my recent discussion about inductive and deductive reasoning associated with AI versus that of humans, at the link here. Another handy example is my recent coverage of AI that purportedly has “shared imagination” with other AI, see the link here.
The gist is that there is way too much anthropomorphizing of AI going on.
You might have also heard of the two lawyers who got themselves in hot water with a court when they submitted a legal brief containing falsehoods or fictitious cited legal cases that they obtained by using generative AI, see my coverage and discussion at the link here and the reactions by judges and the courts at the link here. This reflects a current structural weakness in how AI is devised and has been referred to as exhibiting AI hallucinations. I disfavor the terminology since the word “hallucinations” has human-like connotations. AI hallucinations are more aptly coined as AI fabrications, and there is a tremendous amount of AI research occurring to reduce or mitigate these vexing and serious AI issues, see my explanation at the link here.
I want to next bring up the overall topic of generative AI and large language models (LLMs). I’m sure you’ve heard of generative AI, the darling of the tech field these days.
Perhaps you’ve used a generative AI app, such as the popular ones of ChatGPT, GPT-4o, Gemini, Bard, Claude, etc. The crux is that generative AI can take input from your text-entered prompts and produce or generate a response that seems quite fluent. This is a vast overturning of the old-time natural language processing (NLP) that used to be stilted and awkward to use, which has been shifted into a new version of NLP fluency of an at times startling or amazing caliber.
The customary means of achieving modern generative AI involves using a large language model or LLM as the key underpinning.
In brief, a computer-based model of human language is established that in the large has a large-scale data structure and does massive-scale pattern-matching via a large volume of data used for initial data training. The data is typically found by extensively scanning the Internet for lots and lots of essays, blogs, poems, narratives, and the like. The mathematical and computational pattern-matching homes in on how humans write, and then henceforth generates responses to posed questions by leveraging those identified patterns. It is said to be mimicking the writing of humans.
I think that is sufficient for the moment as a quickie backgrounder. Take a look at my extensive coverage of the technical underpinnings of generative AI and LLMs at the link here and the link here, just to name a few.
You are now sufficiently up-to-speed overall on AI that we can dive into the matter.
There are three key paths to solving the prevailing issues about our laws:
- (1) Writing of laws. Using generative AI to aid in composing laws, including doing initial composition, double-checking, refinement, etc.
- (2) Rewriting of laws. Using generative AI to assist in rewriting laws, such as clarifications, simplifications, closing loopholes, and so on.
- (3) Interpretation of laws. Using generative AI to perform interpretations of laws to identify applicability, especially as a 24×7 real-time always-available resource for everyday use.
I will be borrowing content from my extensive prior work and research on these AI & Law topics, such as at the link here, the link here, the link here, my book at the link here, and so on. For purposes of this herein discussion, I have space requirements that constrain the size of the discussion, so I urge interested readers to see my other works accordingly.
Let’s briefly unpack each of the three major paths.
Generative AI And The Writing Of Laws
I’ll start by discussing the writing of laws.
An axiomatic pledge by lawyers should presumably be that all new laws when handcrafted shall “first, do no harm” (a variation of the Hippocratic oath). I mean to say that the law should cover what it is supposed to cover. It shouldn’t be sloppy. It should not have gaping holes. The standard-bearer notion for the revered rule of law necessitates laws be understandable, well-devised, well-composed, and meet their requirements. It should be clear-cut as to what the scope is and intentions are.
The design should be transparent. Imagine that we are building a bridge when composing a new law. You certainly don’t want to have a bridge that collapses upon first use. Engineers are versed in the dynamics and physics involved. Likewise, in theory, lawmakers ought to be composing laws in a carefully calculated engineering manner. Step by step, mindfully, systematically, and dutifully.
Turns out, the handcrafted reality right now is perhaps aspirational more so than what seems to be attained.
I already mentioned that laws are typically written in ways akin to the making of sausage. The end product will often be mired in all kinds of problematic issues, including legal gotchas, legal loopholes, legal omissions, legal conflicts, and so on. Regrettably, those thorny and at times devastating legal issues aren’t always discernable at the front end, or in haste they are overlooked. Only downstream, after the law is enacted and put into place, does the angst of what was stated or not stated become more readily apparent.
The aim ought to be to do the best lawmaking at the front end or while upstream, thus reducing or possibly eliminating the flood of heightened legal nightmares that arise downstream, see my detailed discussion at the link here.
Lawmakers will often compose news laws by reusing content from existing laws and then adjusting or augmenting the being-devised law as needed. Another avenue would be to write a new law entirely from scratch, perhaps relying primarily upon legal training and law-making experience of how laws are structured and composed to pen the new law. The use of word processing is certainly leveraged, though mainly in somewhat pedestrian ways such as spellchecking, tracking of document changes, and the like.
Generative AI can be used to actively write new laws.
First, I am not saying that we should allow AI to write the law without any human intervention. There are plenty of sci-fi tales that scarily portray a future consisting of AI writing our laws. Lo and behold, the AI then surreptitiously crafts laws that ultimately enslaves humankind and give AI the right to stand as our overlord. You’ve seen these or read such tall tale stories, I’m sure.
For now, having the human in the loop is the prudent approach.
Second, the reason that generative AI can be a useful aid in writing laws is due to the incredible fluency apparatus involved. The reuse of prior laws is really a kind of pattern-matching exercise. What portions are useful to carry over? What portions are best to be left out? That’s something the mathematical and computational capabilities of generative AI and LLMs are good at doing.
Third, generative AI can be directed to keep the law as simply stated as possible, doing so without inadvertently leaving things out or simplifying to the degree of only having mush. I have covered the advantages of using generative AI to translate legalese into more comprehensive language, see the link here. This can be done without having to necessarily sacrifice legal aptness or lawful accuracy.
Fourth, the use of generative AI can accommodate the need for quickness in putting together new laws. Some might insist that it will slow things down and get in the way of writing laws. False. Those trepidations are often voiced by those who aren’t familiar with modern-day AI in the legal domain or are stuck in the past of the earlier days of applying AI to the law.
I want to clarify that this does not mean that generative AI is a silver bullet toward the writing of new laws.
One issue that I earlier noted is the difficulty of encountering AI hallucinations or AI fabrications. The horror story in the legal realm goes this way. A law is written by AI. The law is enacted. Oops, the law contains made-up junk that opens a legal can of worms. Darn, we ought to ban the use of generative AI in the writing of laws.
The kneejerk reaction to the proposed use of generative AI for writing news laws is almost always along the same lines of bemoaning the emergence of AI hallucinations. This becomes a rallying cry to stop the matter at the get-go. It is easy pickings. You don’t have to know one iota about AI to scream to the rooftops that AI hallucinations make this a proclaimed non-starter proposition.
Well, let’s get real, shall we?
Please observe that I openly declared that we should keep humans in the loop.
In that sense, the expectation would be that humans would review the AI-drafted law and aim to see that it is sensible and doesn’t contain AI-based fabrications. The usual retort is that if humans have to review it, why not just have humans write it, but that’s yet another distracting ploy. The time saved and preciseness due to the AI is likely a huge benefit at the lower cost and effort of having humans review the generated content.
Another retort is that suppose the lawmakers do review the AI-generated content but fail to catch the AI fabrications. Yikes, this could easily happen since the lawmakers might mentally be lulled into believing or assuming that the AI must be right. I hear you. We could have some form of reviewing process that is tracked and guided to ensure the human reviews occur with proper and documented due diligence. Admittedly, blemishes might still get through.
Ergo, an added approach to the review process would be to use generative AI to review the generated new law that some other generative AI had drafted. This is known as compound AI, see my description at the link here. You can easily have one generative AI, such as ChatGPT, be reviewed by putting a different generative AI, such as Claude, on the case. If you like, you can do this in a double-checking triple-checking series, namely AI that checks an AI that checks an AI. You get the picture.
The gist is that with some thoughtful and meaningful measures of quality control, encompassing multiple layers of review, the new laws would seem to be at least as good as what we have today (recall the tendency of how things happen now). And, auspiciously, there is a solid chance the laws would be better written, noticeably so.
Generative AI And The Rewriting Of Laws
We can breeze through the use of generative AI for rewriting laws since the same considerations generally occur with those of using generative AI to write new laws.
The idea is that if an existing law is going to be revised, you can make use of generative AI to assist in doing so. Once again, this is assistance. There ought to still be a human in the loop. There ought to be multiple levels and types of reviews.
Soak and repeat.
An important point that comes up when rewriting laws is something generally hidden from view when talking about the writing of new laws. I didn’t mention this hidden factor in my above rendition, knowing that I was going to cover it here.
The deal is this.
An existing law almost for sure has entanglements with other laws. I had noted that point earlier when discussing the domino theory of having laws crash down when removing one domino. The same logic is that if you change a domino, the crashing can also happen.
Revising a law needs to be done in the context of what else is dependent upon that law. How will the revisions impact those dependencies? You might think you are doing a wonderful revision that resolves a long-standing gotcha. Happy face. Turns out that if you put that revision into place, it causes other dependent laws to suddenly have all manner of different interpretations and implications. Sad face. You managed to shoot your own foot when making the revision.
Generative AI can be devised to look for legal dependencies and take those dependencies into account.
Besides offering revisions that meet numerous constraints, the AI can be directed to explain the basis for the suggested revisions. This is known generally as XAI or explainable AI, see my coverage at the link here. In that manner, a human reviewer can be alerted or notified as to why the revisions are being recommended.
The hidden rabbit-in-the-hat trick is that this same legal analysis should be taking place when devising new laws. I didn’t emphasize that point previously and waited to spring it on you now. Human lawmakers might not invest the effort to ascertain what other laws are going to be impacted by making a new law. The same goes for revising laws. We assume they would. We hope they do.
Generative AI provides an added tool to do so.
Generative AI And Interpreting Laws
There are lots of twists and turns in using generative AI to interpret laws, see for example my analyses at the link here and the link here.
Let me share a 30,000-foot level example use case that pertains to the everyday person carrying out their everyday lives.
Go with me on a journey.
The landscape reality is that even if we somehow waved a magic wand and started using generative AI to aid in the writing of our new laws and the revising of our existing laws, the legacy laws are still going to be around for a very long time. The better-devised new laws and revised laws will be a mere drop in the bucket. The bucket mainly will continue to consist of the existing laws. The specter of doom and gloom concerning poorly written prevailing laws will still be hovering mightily and ominously over our heads.
Unpredictable legal traps for the unwary are still solidly ready and primed for use.
What can be done about this?
One angle would be to use generative AI for what I refer to as just-in-time law-checking and interpretability-applicability.
The scenario is as follows. A person has a smartphone that contains a generative AI app established for law-checking and interpretability-applicability. This is beyond the generic generative AI that you are using for doing homework or fun-filled essay writing. I would advise you to not use generic generative AI for this type of legal purpose. Indeed, if you look at the licensing agreement for most of the major generative AI apps, they usually say outrightly that you aren’t to use their AI for seeking legal advice, see my coverage at the link here.
Using a customized law-checking specifically built generative AI app, a person indicates to the AI that they intend to do this or that. Generative AI provides indications about the potential legal implications, noting laws that might be applicable. There is certainly grayness in this. The AI will be interpreting laws since the prevailing laws have their gotchas, loopholes, and numerous variable interpretations.
A notable facet is that the person can access this generative AI app at any time of the day, 24×7, and do so from just about anywhere. We are gradually going to see that generative AI apps do not necessarily need an Internet connection and will operate on smartphones directly, a trend underlying the Small Language Models (SLMs), see my predictions at the link here.
The person can decide what they want to do as a result of the generative AI providing them with the law-checking interpretation-applicability. The AI can be ignored by the person. The AI might stir the person to change what they were planning on doing. The AI might spur the person to contact a lawyer, which I mention for reasons I’ll get to in a moment.
A comment that I get when I give talks and presentations on this topic is that people are not going to laboriously enter into their smartphones every action or planned step they intend to take. The effort would be exhausting and impractical. In that sense, the use of this kind of generative AI would seem rather narrow, only being leveraged in instances where someone already suspected that maybe there are laws encompassing what they are about to do.
The irony there, in this context, is that the overarching notion expressed by Justice Gorsuch consists of unpredictable traps for the unwary. If only the mindfully wary are using this kind of generative AI, only a tiny fraction of those under potential legal risk are going to be helped. The vast majority won’t be.
Aha, the reply or answer is that we are moving rapidly toward multi-modal generative AI, see my coverage at the link here. You might have seen the recent spate of e-wearables which are smart rings, necklaces, eyeglasses, pins, and other wearable devices that contain generative AI. Generative AI can continuously keep up with where you are, and what you are doing, and otherwise be relatively on top of the context of your actions. In that manner of functioning, you would not laboriously need to enter queries into the smartphone. The generative AI could be working proactively on your behalf and alert you when appropriate.
I believe that provides a high-level sketch of this particular use case.
The Ramifications For Lawyers And Budding Legal Beagles
You might have noted that I mentioned that generative AI could spur someone to contact a lawyer. I brought up that point because many lawyers instantly get riled and upset that this kind of generative AI seems to intrude and be aimed at taking over their jobs.
The reality for this kind of generative AI that does law-checking is that it might fortuitously drive more people toward seeking the advice of lawyers. Say what? Yes, people who realize that they might face a legal question are probably more likely to reach out to a lawyer for legal advice.
Currently, many people, perhaps most, don’t even realize when they ought to be contacting a lawyer.
I submit to you that the use of AI in this fashion will provide a gusher of new business for lawyers and law firms. There is a tremendous amount of untapped demand for legal services. People usually have little awareness that they should be consulting with a lawyer. This tool will exponentially increase that awareness. Boon times for lawyers will ensue.
Notice too that the AI is seemingly not proffering legal advice (well, maybe?).
Take a deep breath because the question is a hornet’s nest of a heated topic that can get lawyers up in arms.
Is the act of warning about a potential legal transgression the same as giving legal advice?
Some insist yes. Some say no. Some see muddied water. It is assuredly a question of a hotly contested nature.
A quick heads-up might help those of you unfamiliar with the trauma and drama at play. You might vaguely know that in the United States, there are rules about who can practice law. I’ve time and again discussed the topic of the Unauthorized Practice of Law (UPL), see my in-depth analysis at the link here. I assure you that the topic of UPL is going to soon land between a rock and a hard place due to modern-day AI.
For those lawyers and law students who are yelling and screaming at the screen right now, as you read about generative AI doing these kinds of legal-related endeavors, I suggest you take a serious and sobering look at my writing about where UPL might be headed. Best to keep your eyes wide open and be informed about what the future might portend.
Conclusion
Legendary English jurist John Seldon made this now-classic remark: “Ignorance of the law excuses no one.”
The stated rule is bendable, sometimes. Lawyers can argue that their client was unaware of a law and therefore should be excused or at least given some less onerous treatment for the legal transgression. I wouldn’t bet on it for myself, you are welcome to do as you please.
Generative AI can aid in everyday people not transgressing laws that are especially unpredictable traps for the unwary. One supposes that the same AI could demonstrably aid those who indeed know of predictable traps. The AI warning might be sufficient to change their minds about violating a law that they already knew about.
That’s a bit of a twofer for using generative AI in this manner.
I customarily during my talks to audiences of lawyers and law students provide some additional grist for the mill. Here it is. Have fun.
If a person decides to use this type of generative AI, and it does warn them, but they proceed anyway, would this make their legal situation even worse? Their chances of arguing they were ignorant of the law seem to be diminished. You can imagine the subpoenas going out to the smartphone service provider and the AI maker requesting all interactions that took place by the allegedly transgressing person.
Another twist comes to mind, which it turns out I have a barrel of them, would be that a person uses this type of generative AI, it fails to warn them, they proceed believing they are good to go, and somehow get nabbed by the law for an alleged legal transgression. You know that the person will go after the AI maker and seek damages for not having suitably warned them. They might logically also try to use the lack of an AI-based warning at trial, asserting that they did everything they could do to beforehand seek a green light for their actions.
I hope that got your legally minded neurons stirring.
On a final remark, I think it is safe to say this about these meaty matters. We are faced today with a vast array of laws that are confounding and contain untold numbers of legal traps. In addition, at the existing pace and societal norms, it is going to worsen. You would be hard-pressed to cogently argue otherwise.
What are we going to do?
We can sit around and let the law pound away at us. That’s the head-in-the-sand approach. Not advisable. A viable alternative would be to employ generative AI, doing so in the right way and with proper sensibility and diligence. I’ve given you an overview of what that consists of. The future on these matters is here and now.
That’s surely plain advice, decidedly not legal advice, just to make it abundantly clear to everyone reading this, thanks. Consider keeping up with my column, see the link here, if you want to see how all this progresses. It is going to be one heck of a wild ride.
