In today’s column, I continue my ongoing identification and exploration of ways for budding startups to leverage generative AI as a potential product or service based on freshly uncovered business opportunities arising from new laws and regulations, see my akin coverage at the link here. I will present you with an exemplary brand-new use case that is ripe for the pickings and is based on a recent column posting in which I discussed the ABA’s (American Bar Association) release of a strict code of conduct for lawyers and law firm’s use of generative AI, see the link here.
First, allow me to explain the general concept of how new laws and regulations can open business opportunities for the fruitful and profitable application of generative AI. After doing so, I’ll dive into a particular use case to showcase the practicalities of this as a means for devising startup ideas.
The notion here is that this is a twofer discussion. The first part is about broadly identifying and exploring new business ideas. The second part is a viable and illustrative example.
Let’s get into it.
New Laws And Regulations Can Open New Business Avenues
One of the perhaps least known or rarely understood business ideation tricks is to scour laws and regulations for startup possibilities.
You might not have given this much thought, which is a shame because there are lots of potential fish to be fried via this method, so to speak.
The idea is that a given law or regulation might impose some form of statutory requirement that makes available the need for a new product or service to aid in fulfilling that requirement. When working with startups, I use a variation of blue-sky thinking that customarily involves thinking anew and outside the box, whereby we seek to identify and rule in or rule out prospective new opportunities due to recent changes in the laws.
It can be a wide-open field if the law or regulation is especially brand new.
You see, oftentimes a new law or regulation stipulates that this or that must henceforth be legally strictly done, even though no such common practice was taking place. Those businesses and entities that are now subject to the new law or regulation will need to scramble to comply with the latest stipulated requirement. Voila, a gap in the market has suddenly emerged.
Readers might recall my identifying and analyzing such an instance when for example New York City (NYC) imposed a new requirement associated with the use of automation for hiring and firing of employees, see the link here. This was a prime example of this principle and illustrated how generative AI could be leveraged as a solution to the business problem that had arisen out of thin air per newly devised and highly controversial legal stipulations.
A means of closing the gap or solving a newly imposed business problem is to apply generative AI as a mechanism for getting the job done. It could be that generative AI can be readily devised to fulfill the new law or regulation. Generative AI should always be in your thoughts. Happy face.
Word to the wise, do not try to apply generative AI as a hammer to all the world’s problems, such that you see everything as being a nail. Sad face. A screwdriver or other AI or non-AI tool might be a better choice. Be wise. Be judicious.
Use generative AI in the right way and you’ll have a stronger chance at success.
I’m also not saying that generative AI on its own will necessarily do. The odds are that you’ll need to encompass generative AI amidst some other surrounding components and might need to make software application bridges to existing software packages or utilities. Lots of assorted and cohesive pieces might be needed to solve the puzzle at hand.
The good news is that generative AI with its fluency and pattern matching can nowadays serve as a crucial core element that satisfies the crux of a new requirement. And, since the new requirement is legally imposed by the law or regulation, you have a customer base potentially eagerly and anxiously hoping to find a solution to their out-of-the-blue problem.
Boom, drop the mic, you have a solution for them.
Keep Your Eyes Open And Your Wits About You
I might mention that one downside here is that many of your prospective customers might not even realize that the new law or regulation exists.
You will possibly need to gently and compellingly educate them.
For the moment, they are heads-in-the-sand unaware that they dutifully need your product or service. You will likely need to convey to them the problem they face, whether they realize or not that the problem looms over them. Assuming you can open their eyes, and then after they get over the initial shock of having a problem that they blissfully knew nothing about, you are hopefully handing them a suitable solution on a silver platter.
As an aside, sometimes businesses get upset at the messenger.
Be wary of this. Be prepared for it. The logic is as follows. Executives weren’t aware they had a problem. They weren’t keeping up with new laws and regulations. You share with them that they do have a problem and if they don’t resolve it, they face various penalties for not complying with the law or regulation. Notably, this can involve not just their business getting dinged, but in addition, the penalties sometimes apply to transgressing individuals or leaders within a company too.
You are bringing bad news to their door. Ugh. Nobody usually heralds the bad news messenger. The messenger takes the heat. What, we didn’t know anything about this new law. We shouldn’t be subject to it. Who passed this crazy requirement? And so on it goes.
It might take a bit of thick skin to contend with this backlash. Your usual effort is to calmly and patiently clarify and emphasize that you had nothing to do with the new law or regulation landing on the books. Instead, you simply observed its existence, realized that a lot of pain would come to businesses, and you have kindly and out of the goodness of your heart put together a solution to this matter (well, plus to make a buck or two too).
Another point I often mention during my startup advisement in this sphere is that not all new laws or regulations necessarily bring forth stipulations that are big enough or compelling enough to be worth your while. Suppose for example that the stipulation only applies to a smattering of businesses. Perhaps the market opportunity is regrettably thin. Also, the cost of your solution has to befit the nature of the stipulation. Businesses might think that they’ll go ahead and take the risk of being non-compliant if the cost to be compliant seems out of proportion to the penalties or adverse outcomes of ignoring the new law or regulation.
All in all, this brief recap on the topic is my way of urging you to be mindful of business opportunities that can arise due to laws and regulations. It might be your cup of tea, or it might seem misaligned with your passions and interests. You need to decide that.
A few additional pieces of advice are important for me to mention. If this line of business opportunities interests you, there are two key strategies to employ.
First, proactively anticipate new laws and regulations. You want to be on top of laws and regulations while they are still being drafted and not wait until after they have hit the streets. The earlier you know of something coming along, the more time you will have to get things underway. If you wait until the law becomes law, it could be that others employing this same trick will already have beaten you to the market.
You snooze, you lose.
Second, you would be smart to have an attorney or two who are keeping watch of the laws and regulations on your behalf. Their legal beagle expertise might notice something coming up that could have business opportunities. Perhaps a layman would not see the openings. A lawyer might.
I’d like to say something which I assure you is not intended as a slam. The attorneys that you opt to have scanning various laws and regulations need to be fully cognizant that they aren’t particularly looking for legal loopholes or gotchas. That might be their default viewpoint. This makes sense since that’s what lawyers are usually paid to do.
For your purposes, they are aiming to find business opportunities embedded within the laws and regulations. This might have nothing to do with legally problematic issues. A requirement for example that businesses must begin reporting on certain types of efforts doesn’t need to have a legal problem afoot. The requirement alone is your business opportunity.
Notably, if they do find legal issues during their scanning this too could be another path toward business opportunities, but of a different ilk and likely entails a different type of pursuit and not of the kind I’m referring to here.
The biggest bonanza for the finding of business opportunities in new laws and regulations is when entirely new lines of business arise from those laws and regulations. In those rarer instances, you aren’t just going to provide a narrow problem-solving solution. You might find yourself faced with the breathtaking whale of an opportunity. Perhaps a system solution from whole cloth is needed and within which generative AI could be a differentiating factor.
You might just get lucky and be at the beginning of something completely new. It is enough to keep you dreaming and hoping.
Generative AI And Going Fishing For Business Opportunities
Before I showcase an example of identifying and exploring these kinds of business opportunities, I will let you in on a little secret.
Are you ready?
I’ve been using generative AI to aid my concerted efforts in finding new laws and regulations that contain potential business opportunities. I share these around to the startups I am mentoring. As I mentioned earlier herein, always make sure to have generative AI on your mind. This is proof in the pudding, as they say.
Here’s the nitty-gritty. I have set up generative AI to automatically examine a select few databases of new laws and regulations that I have access to. The prompts that I’ve given to generative AI are regarding the ins and outs of finding business opportunities. This took a bit of back-and-forth for me to refine. You don’t want to be deluged by false positives and false negatives.
Since I don’t expect this to necessarily find a gem in real-time that requires an instantaneous response by me, I have generative AI send me a nicely formatted summary each day. For tips on prompt engineering approaches that enable these kinds of setups, see my comprehensive overview at the link here. I take a quick look to see whether my fishing expedition might have landed a worthy opportunity. I am somewhat peering in the rear-view mirror since the scanning right now is recently passed laws, but some interesting opportunities show up anyway.
For the ones that seem like desirable catches, I then have generative AI do an online search to see whether other existing solutions would seem to fit that opportunity. If the market is already well satiated and those are going to easily be stretched or adjusted to meet the opportunity, it probably isn’t worth the startup angle. An exception might be to devise something that the prevailing market solutions might want to buy or rent, though that’s a different path on this journey.
Generative AI is acting as my daily fisherman and continually casting a line to see what interesting and profitable fish might be in the legal opportunity’s sea.
Wish me and the startups I mentor some uplifting good luck in the fishing adventure, thanks.
Generative AI And An Exemplar Of These Considerations
Speaking of generative AI, I want to shift now and give you a handy example of the various ways in which new laws and regulations can be leveraged for business opportunities, especially opportunities where generative AI is a reasonable and befitting solution.
This example arises from a recent posting I did about the ABA promulgating Formal Opinion 512, a new guidance document for lawyers. See my in-depth analysis at the link here.
The guidance specifies numerous significant and at times stringent facets of how lawyers and law firms are to conduct themselves when it comes to using generative AI in their law practice. Here’s what the ABA had to say about the Formal Opinion 512 as released on July 29, 2024 (quoted from the ABA website):
- “In its first guidance on the use of artificial intelligence, the American Bar Association Standing Committee on Ethics and Professional Responsibility has released a formal opinion urging lawyers to be mindful of several model rules, including ones related to competency, communications, confidentiality, and fees.” (per the ABA website).
- “This opinion identifies some ethical issues involving the use of GAI (generative artificial intelligence) tools and offers general guidance for lawyers attempting to navigate this emerging landscape.” (per the ABA website).
The newly released Formal Opinion 512 “Generative Artificial Intelligence Tools” guidance document is fifteen pages in length and includes the kind of customary footnotes that you would find in any bona fide legal narrative. The document is jampacked with key insights, tips, restrictions, suggestions, and rules of the road when it comes to lawyers using generative AI.
The keystone ingredients are stipulations on a lawyer’s use of generative AI in these six major buckets:
- (1) Competence
- (2) Confidentiality
- (3) Communication
- (4) Meritorious Claims and Contentions and Candor Toward the Tribunal
- (5) Supervisory Responsibilities
- (6) Fees
I’d like to share with you a few snippets from Formal Opinion 512 so that you’ll have a general grasp of what the code of conduct stipulates.
An example in the Competence subsection includes these salient points (excerpts):
- “Before lawyers input information relating to the representation of a client into a GAI tool, they must evaluate the risks that the information will be disclosed to or accessed by others outside the firm.” (source: ABA Formal Opinion 512).
- “Lawyers must also evaluate the risk that the information will be disclosed to or accessed by others inside the firm who will not adequately protect the information from improper disclosure or use because, for example, they are unaware of the source of the information and that it originated with a client of the firm.” (ibid).
I assume you can readily accept the above premise by the ABA that attorneys and their respective law practices must be cautious about entering client info into generative AI.
The main reason is that the off-the-shelf generative AI apps such as ChatGPT, GPT-4, Gemini, Claude, and others will usually contain in their licensing agreements something that says the AI maker and their developers can access the prompts that have been entered by users. Furthermore, they often indicate they can reuse the entered info for ongoing data training of their generative AI app. There are ways to deal with these potentially unnerving and disconcerting privacy and confidentiality issues of generative AI, see my coverage at the link here.
If you’d like to dig deeper, I have examined the risks of disturbing the vaunted attorney-client privilege due to attorneys and their clients making use of generative AI, see my analysis at the link here. This is something that can happen at the hands of the attorney or the hands of the client.
Lawyers and law firms need to be thinking mindfully about these sobering matters.
Moving on, consider these selected snippets about the importance of communicating with a client concerning the use of generative AI by the law firm or lawyers (excerpts):
- “The facts of each case will determine whether Model Rule 1.4 requires lawyers to disclose their GAI practices to clients or obtain their informed consent to use a particular GAI tool. Depending on the circumstances, client disclosure may be unnecessary.” (ibid).
- “Of course, lawyers must disclose their GAI practices if asked by a client how they conducted their work, or whether GAI technologies were employed in doing so, or if the client expressly requires disclosure under the terms of the engagement agreement or the client’s outside counsel guidelines.” (ibid).
- “There are also situations where Model Rule 1.4 requires lawyers to discuss their use of GAI tools unprompted by the client.” (ibid).
The focus in the Communications subsection is that a lawyer and a law firm might need to inform their client that generative AI is being used to support their legal case. You can see that the rules have lots of twists and turns. I will come back to that in a moment.
The essence is that very little of this code of conduct is reducible to simplistically distinct steps or ironclad codifiable rules. To a certain extent, the code of conduct is more qualitative than it is rote.
Hang in there, I made that point for a notable reason that will become apparent momentarily.
Finally, as another snippet, here’s a portion from the Supervisory Responsibilities subsection (excerpts):
- “Managerial lawyers must establish clear policies regarding the law firm’s permissible use of GAI, and supervisory lawyers must make reasonable efforts to ensure that the firm’s lawyers and nonlawyers comply with their professional obligations when using GAI tools.” (ibid).
- “Supervisory obligations also include ensuring that subordinate lawyers and nonlawyers are trained, including in the ethical and practical use of the GAI tools relevant to their work as well as on risks associated with relevant GAI use.” (ibid).
The proposition is that the law partners or other managerial personnel are supposed to establish policies about generative AI use in their law practice. Those policies need to be clear about what is proper and what is improper usage.
Indeed, the law firm will be held accountable for establishing such policies and they must also make sure that the lawyers and even nonlawyers of the firm abide by the stated policies. The takeaway is that not only are attorneys at risk for not using generative AI properly, but the law firm has its feet in the fire too.
For a now famous example of how two attorneys at a law firm got themselves and their law firm in trouble by having carelessly relied on generative AI to produce content for their legal brief (which encompassed AI hallucinations or made-up falsehoods), see the link here.
That completes the background on this example, and I am ready now to share you with a mind-bending reveal.
Using Generative AI To Aid In The Proper Use Of Generative AI
Let’s add up things.
The ABA has come out with a new code of conduct.
The stipulations require lawyers and law firms to undertake various efforts. If they ignore those stipulations, they do so at the peril of being dinged as to their legal privileges and could find themselves in the doghouse as to being able to practice law. Practicing law is of course their said-to-be reason to exist. This needs to be protected dearly, nearly at all costs.
I realize a smarmy person would say that all they need to do is avoid generative AI. By doing so, they would seem to be outside the bounds of any requirements regarding the use of generative AI. Ergo, if they don’t use generative AI, they escape the realm of punishments for wrongly or inappropriately using generative AI.
That seems like impeccable logic. The problem is that inch by inch, lawyers and law firms are going to be needing to use generative AI. It is going to become not simply a competitive advantage; it is a competitive necessity. Take a look at my lengthy depiction of why this is going to happen, see the link here and the link here.
For the time being, yes, some lawyers and law firms can steer clear of generative AI. I don’t think they will be able to do so for very long. They will find themselves unable to attract good lawyers who expect and want to use generative AI. They will find themselves facing clients who ask whether the firm is using generative AI, and if not, the clients will go elsewhere. Etc.
Wait and see, if you dare.
Getting back to the matter at hand, you can consider the ABA Formal Opinion 512 as though it is a newly passed law or regulation. Don’t interpret that literally. It is a code of conduct, which is not the same as a law or regulation. Still, it has the semblance of importance and possible bite that a law or regulation has, at least in the narrow and particular realm of practicing law.
Mull this over.
What have I been saying so far about finding business opportunities when new laws or regulations arise?
If you find them, consider what you can do to leverage or exploit them, and especially whether generative AI can be a notable element for a business solution.
Aha, I think we have a winner-winner, chicken dinner.
We could put generative AI up to the task of data learning about the Formal Opinion 512, and then use that data-trained generative AI as an advisor for law firms and lawyers when they opt to use generative AI.
This is a sensible place to use generative AI since the code of conduct is rather qualitative and not pinned down to a precise set of rules. If it was based on cleanly specified rules, I would suggest using some other type of AI rather than generative AI. Turns out that the situation in this instance is ripe for generative AI due to the code of conduct being reliant on fluency and being open-ended. It is imprecise or loosey-goosey.
We like that for making use of generative AI.
I went ahead and did data training of generic generative AI with the Formal Opinion 512, using a commonly accepted technique known as RAG (Retrieval-Augmented Generation), see my explanation of RAG and in-context modeling at the link here. If there is reader interest, I’ll do a follow-up piece to show you the use of augmented generative AI that I performed. Let’s stay for now at the 30,000-foot level.
I’d say for a rough draft and off-the-cuff experiment, it turned out pretty good.
The overall sense of things is that while using generative AI for lawyering, the same generative AI can be advising a lawyer about abiding by Formal Opinion 512. This is done at the right time and in the right way. I say this because you could cheat things by simply displaying a warning to a lawyer that they need to be careful when using generative AI.
That kind of bland alert won’t get you very far. Some lawyers might decide to not use generative AI because they aren’t sure what the warning entails. That’s not a good outcome. Others might see the warning and summarily disregard it. That’s also a not good outcome.
The beauty of having generative AI become data trained on the code of conduct is that the AI can at the right time and right place notify the lawyer when they are seemingly veering out of bounds. This allows a lawyer to use generative AI, remain immersed, and not have to keep in their head the byzantine set of rules that say what they can and cannot do.
Generative AI is helping with that.
As furtherance of this approach, I also opted to compose a short version of a law firm’s set of policies about generative AI (I adapted one I had previously used in the real world). I included that into additional data training using RAG again.
This same approach is what should happen at a law firm, namely the code of conduct needs to be married up with the specifics of the particular law firm. You can’t leave out the specifics. Those are the rules of the road for what a law firm has presumably mindfully decided is the best way they believe that generative AI is to be used by their lawyers and nonlawyers (within the boundaries and stipulations set forth by the ABA).
Exemplar Path And Implementation Insights
Consider the numerous ways that generative AI can be used to adopt the ABA Formal Opinion 512.
Generative AI comes into use for each of the six major categories:
- (1) Competence. Generative AI can assist lawyers in learning how to properly and best make use of generative AI for legal purposes, raising their competence in such usage, doing so in a ratcheting manner of starting easy and guiding them toward advanced legal uses.
- (2) Confidentiality. Generative AI can actively caution when a lawyer enters seemingly confidential data and opt to guide how to avoid doing so or otherwise assist in what to do.
- (3) Communication. Generative AI can advise on whether communicating with clients about the use of generative AI is needed, plus can keep track of the usage and provide summaries or reporting accordingly.
- (4) Meritorious Claims and Contentions and Candor Toward the Tribunal. Generative AI can warn about potential AI hallucinations, including AI actively double-checking and emphasizing the need for lawyers to by-hand double-check the legal work produced too.
- (5) Supervisory Responsibilities. Generative AI can aid in devising a firm-specific set of policies and practices associated with generative AI usage, plus share that while lawyers are using generative AI, and detect if out-of-policy uses have occurred.
- (6) Fees. Generative AI can advise on various fee and billing arrangements associated with the use of generative AI and keep track of usage for reporting and invoicing purposes.
Those are just some of the relevant uses.
One crucial element to consider is whether the generative AI is generic or whether it has been tailored for the domain at hand, in this case, the legal domain.
Here’s what that means.
You can make use of generic generative AI such as the off-the-shelf ChatGPT, GPT-4, etc., and the AI is not especially versed in the legal domain. It is considered a jack-of-all-trades but a master of none. Another rising approach involves domain-specific data training a generative or LLM (large language model) from scratch. This might be a legal realm, medical realm, financial realm, and so on, see my coverage at the link here.
LegalTech providers that are either employing generic generative AI or domain-specific generative AI ought to be giving serious thought and action to devising the above-indicated use of their generative AI for implementing the Formal Opinion 512. If the generative AI is generic, perhaps clients will have lower expectations about this alignment. Domain-specific generative AI data trained in the legal domain is going to almost certainly be perceived as a must-have to contain this added capability.
If a LegalTech vendor ignores the idea of implementing Formal Opinion 512, the lawyers and law firms that have adopted the said LegalTech will likely proceed blindly ahead and not realize that there is a code of conduct for their generative AI use. Imagine their rightful dismay upon finding out elsewhere and being quite perturbed that the LegalTech vendor had not notified them of the significant matter.
Suppose a client knows of the code of conduct, and they are actively using a LegalTech product that includes generative AI, but the LegalTech vendor has not taken any action associated with implementing the code of conduct. A determined client might potentially resort to manual methods. This is an onerous path, bound to be done half-baked, and will be out of step with the generative AI usage since there will be almost undoubtedly out-of-sync written guides separate and out-of-mind from the hands-on usage. In addition, this forces all the clients into reinventing the wheel, each of them independently having to figure out how to map the two together.
I trust that this lays out some of the parameters associated with this business opportunity. It is an example of a business opportunity arising from a form of legal stipulation.
Admittedly, this instance is quite narrow, and the market size currently is still in its early stages. Lawyers and law firms are gradually adopting generative AI, though it is relatively sporadic and not yet widespread. These are the early days of the use of generative AI in the legal realm.
Keep that chin up. Predictions are that the existing global legal service market of around $800 billion will eventually be extensively using generative AI, see my analysis at the link here. The point is that entry now will require a longer-term perspective and a setting of expectations that the best days of generative AI usage in the law are up ahead.
It’s a long play and substantively attractive due to the market size all told. I’ve repeatedly indicated that we will see a massive transformation in legal services and that startups getting in now have the enviable but also a tough road of being on the ground at the launch pad for what will come.
Conclusion
Let’s get to a few concluding remarks.
I have previously discussed business opportunities based on new laws and regulations that involved the use of generative AI for shoring up gaps and implementing new requirements. Those instances did not involve having generative AI already in the mix. Generative AI was to be newly adopted.
This time, I decided to take you through a somewhat more intriguing ride.
We added generative AI to already making use of generative AI.
Nice.
When I tell people that it is possible and businesswise sensible to use generative AI as an augmentation to generative AI usage, their eyebrows raise, and a stunned look of momentary bewilderment typically happens. This use case is now my handy example to point this out as a viable opportunity.
As you saw, lawyers and law firms using generative AI can further leverage generative AI to abide by the ABA code of conduct for generative AI usage. Seems like one of those circular ideas. Maybe it is reminiscent of the movie Inception. The idea here though is more straightforward than the movie, I’d suggest.
And you can expect to see a lot more of this.
If you want another brain teaser, sit down for this. You can use other generative AI in conjunction with whichever generative AI that you are already using. Some refer to this as compound generative AI. Simply stated, you connect multiple generative AI apps together.
I think you can see how this would work with the example of the lawyers using generative AI. They might say be using ChatGPT and a code of conduct adviser about generative AI usage might be devised in say, Claude. You could then have ChatGPT and Claude be connected via APIs, see my discussion at the link here.
That about covers today’s discussion.
For those of you who are entrepreneurs and thinking about doing a startup, I like to quote Thomas Edison for his inspiring line: “The value of an idea lies in the using of it.”
I’ll know I’ve done something useful here if it gave you an idea that you then act upon and turn it into something amazing. I’m sure you will.
