15 Comments
User's avatar
Jeffrey Carr's avatar

Precisely what #DeliveredValue is all about — and that’s precisely what my team insisted upon and paid outside counsel for when I was a GC. What was truly amazing — no disconcerting — is that despite demonstrated effectiveness and results, true delivered value to the customer and increased profitability for those firms that delivered, the model, while praised and celebrated, never took hold more broadly. I blame the #LawLand guild member buyers for that disconnect — and indeed have argued that the failure to demand delivered value was tantamount to a breach of the in-house lawyer’s fiduciary responsibility to their customer. When the needs and the protection of the profession outweigh the needs of the customers, the profession mutates from one designed to protect customers to one dedicated to the preservation and protection of the prerogative, power and profitability of the guild-based intermediary. At the end of the day, like all intermediaries, they become a citadel for income redistribution — formed in the name of efficiency and protection of the customer, but that becomes a mere fig leaf to obscure the mutant and malignant tumor.

What you are arguing for and predicting is precisely what we deployed effectively. The tragedy of the past several decades when this model was freely available — and free — is that broad based-acceptance could have driven mutually advantageous results for firms and customers, but the customers failed to demand Delivered Value and the firms resisted this challenge to the prevailing business model tooth and nail. Like a bizarre combination of Diogenes and Sisyphus, I long argued that it was time for customers to vote with their fees and to demand delivered value. When they did this, the firms would be forced to change and deliver that value. The AI revolution simply provides the tools by which the customer can bypass the lawyer and the law firm for the vast majority of their legal needs — if only my team, and the very few GC fellow travelers focused on actual customer needs, had access to these tools years ago — the legal industry landscape would have been much different, much sooner. We didn’t and were patted on the head, compliments, but considered to be lunatics on the radical fringe.

But as is so often the case with your masterful delineation of the problem, you are once again correct — those needs where a lawyer’s judgment is actually necessary in the operation of the service delivery platform, as opposed to the design of that platform, is where the true delineation of legal needs must be discerned. What I find interesting is that most of the legal industry cognoscenti focus on what this means for lawyers and law firm — on the lawyer as the customer for AI enabled or enhanced services. It is not focused on what truly customer focused legal service delivery means and how the realization of those needs is what actually matters — to the customer — and therefore what the customer is willing to pay for.

In a world where 85% of legal needs are unmet because the prevailing delivery system is simply too expensive, that delivery system will be bypassed by those with the legal needs to be met and available tools to now meet those needs. For that latent unmet demand, this is not UPL — it is using a delivery platform that is pretty effective, recognizing that while it might be wrong from time to time, it is far better than simply going naked all the time.

Emma Elliott's avatar

I agree but I am left with more questions!

Is Sphere 1 addressing the access to justice imbalance? Will it also act as the Jevon's paradox by increasing access, leading to increases demand which could result in more Sphere 2 matters for firms and a great relationship foundation.....but most importantly how are the next generation of lawyers going to be trained, trusted, instilled with wisdom to be able to assist in Sphere 2? I hope Part 2 has these insights!

Jordan Furlong's avatar

Thanks very much, Emma! Some thoughts on each of these points below:

1. I view Gen AI as fundamentally empowering technology. It's going to put in people's hands the capacity to learn that they have a legal remedy to their problems and/or go some distance (maybe the whole way) towards obtaining that remedy. We'll see this in settings as diverse as a lower-income single parent learning about their legal entitlements to child support, and a Fortune 100 company insourcing and automating compliance activities that they once sent out to law firms.

In one sense, this will be very good for A2J. But in another, it will create more problems, because now there'll be hundreds of times as many claims and counter-claims to courts and requests to government agencies, and our public legal systems can barely handle the relative trickle of demand they now receive. (Jevon's Paradox with a vengeance.) So while AI will open the taps of demand, the systemic bottlenecks will remain and will get much worse. I've written more about AI as an access tool here (https://jordanfurlong.substack.com/p/level-the-playing-field-give-consumers) and AI as a systemic bottleneck reliever here (https://jordanfurlong.substack.com/p/the-end-of-courts).

2. The empowerment of demand enabled by AI will produce tons of Sphere 1 work that the AI will, in turn, be able to satisfy itself. But as you say, it should also create more Sphere 2 opportunities for lawyers. One easy example: a small business that couldn't get off the ground because it needed expensive legal help can finally launch using AI, creating not just more employment and business opportunities in its market, but also a thriving business entity capable of producing the higher-end, "journey-safeguarding" Sphere 2 work discussed above.

That's where much of the urgency with which I address these topics comes from: Soon enough, almost the *only work* that will come to lawyers will be the more intricate, sensitive, challenging Sphere 2 requests -- but lawyers aren't developed and trained to meet that demand. Law school, bar admission, and maybe the first year or two of practice in a law firm are mostly devoted to Sphere 1 work, on the premise that that multi-year immersion in a law practice environment and repeated exposure to building-block legal assignments, ideally combined with supervision and feedback, will eventually enable the lawyer to develop the judgment, insight, and experience necessary to tackle Sphere 2 retainers. That has worked, more or less, up until now. But I fear AI is going to wreck that haphazard contraption by automating or mechanizing the billable work that once sustained early-career lawyer training in law firms.

3. That brings me to your third question, which I agree is the most important: How do we reconfigure our lawyer formation system to accelerate the development of the skills and capacities needed to tackle Sphere 2 work? That's the challenge I've set myself in the book I'm currently working on, and while there are potential solutions that I can see, most of them are highly disruptive to our current approach and some of them are truly radical, at least when considered from the perspective of the traditional legal profession.

I can see one way forward in which large law firms "spin off" their internal training systems into separate businesses that train lawyers of all kinds and accelerate their development into Sphere 2-ready lawyers. I can see another in which the profession collectively sponsors and oversees "teaching law firms" where new lawyers are exposed to street-level legal services in highly structured supervised-practice environments, like maxed-out law school clinics. I can see a third route in which traditional CLE providers are repurposed to become transitional training academies that bridge the gap from school to practice. Each has its pros and cons. (You'll notice I don't suggest reconfiguring law schools themselves, because I think that's simply not a practical option to explore.)

I've been coming at this issue in a number of posts here at the Substack over the past three years, but I haven't been fully satisfied with any of them yet. I hope that writing this part of the book will help me sharpen my ideas to a useful point, because I really don't think we have a lot of time left to make this transition.

Kelli Dunaway's avatar

I wrote an article for the National Law Journal in 2013 arguing for a medical residency type program in Big Law (it was when we were starting to see clients refuse to pay for 1st year associates on their bills and wondering how in the heck we were going to have great midlevels without training excellent junior associates). It might be time to dust off this idea as a solution to the AI revolution.

Taylor Logan's avatar

As a law student, I have been trying to anticipate the ways in which the legal landscape is going to change with respect to AI. I think similar to what you articulated here lawyers could step into more of a consulting/ advisory role. With that I can fathom a legal AI oversight position, checking the soundness of arguments/clauses and catching any misinterpretations before filing or sending to opposing parties. Even then, it seems the man power needed is going to decrease drastically. As you stated, with new technology and new fields comes new challenges. There may be entire fields of law that emerge. I would like to see AI’s automation of Sphere1 allow for a creation of a Sphere 3. Something akin to an enlightenment period in the law where lawyers can tackle systemic issues or wrestle with more abstract ideas.

Jordan Furlong's avatar

Taylor, you're absolutely right to be tracking AI's trajectory and trying to figure out where to meet it during the course of your legal career. It's extremely challenging, of course, given the speed at which it's changing and growing, but important to do nonetheless.

Two quick thoughts on your observations.

1. I agree that there will be opportunities to perform legal AI oversight, especially in these next few years as we all collectively try to determine the limits of this technology's accuracy and effectiveness. I've written previously about the growing importance of "quality control" as a basic lawyer competence: https://jordanfurlong.substack.com/p/the-newest-lawyer-competence-quality.

That having been said, I don't think I'd want that to become the primary role that lawyers perform in future. There's very little room for creativity in that kind of work, not to mention the prospect of rarely if ever dealing with an actual human being! One of the supposed promises of AI is that it will free lawyers from the "drudgery" of basic legal work. I think there are developmental reasons for us to value the "drudgery" more than we do, but I also wouldn't want to substitute one kind for another. full-time That wouldn't be much of a future for you and your colleagues to look forward to.

For new and aspiring lawyers in particular, I've written about the post-AI opportunities that can be found in:

- Sole practice (https://jordanfurlong.substack.com/p/youll-never-be-solo-again)

- Niche practices (https://jordanfurlong.substack.com/p/ai-and-the-rise-of-the-niche-lawyer)

- Relationship-based practice (https://jordanfurlong.substack.com/p/new-lawyers-start-building-your-client)

2. I really like your idea of "Sphere 3" work -- that's exactly the kind of growth mindset and imagination that we'll need in this profession as AI continues to terraform the legal sector. Back in 2023, inspired (maybe unnerved is a better word) by AI's burgeoning capacities, I wrote three articles that talked about these kinds of "systemic justice" and "Renaissance Era" opportunities that are now within our grasp; I've shared them below in case they might have some value for your own thinking in this regard.

- https://jordanfurlong.substack.com/p/law-unlimited-welcome-to-the-re-envisioned

- https://jordanfurlong.substack.com/p/its-time-for-a-legal-moonshot

- https://jordanfurlong.substack.com/p/what-do-you-want-the-future-to-be

Taylor Logan's avatar

I will read up on supplemental articles! Interesting discussions to be having and anticipating!

Jon Van Loo's avatar

This is a great observation and it seems to hold a lot of promise for clients, the customers who buy legal services. But how long is it going to take until the technology works its way into viable solutions for consumers of legal services? The speed of the technology is remarkable, but working its way into the marketplace might take many years.

Jordan Furlong's avatar

Jon, that's a great question, and my honest answer is that I don't know. The raw capacity of Generative AI is increasing at a rate that outstrips our institutional and entrepreneurial ability to respond and adapt to it. Even the AI companies themselves aren't sure what they've got on their hands and how or when it will transform sectors like law (see, e.g., https://www.youtube.com/watch?v=N5JDzS9MQYI&t=1214s). We're trying to fit a firehose onto an ever-increasing torrent.

The primary obstacle to widespread public adoption of AI tools for solving legal problems, it seems to me, is the absence of freely available and searchable authoritative legal databases. In most countries, this data is proprietary, owned by legal publishing companies or sometimes by courts themselves. In places where this data is available, such as Canada, there are early signs that a frontier AI application like Claude Cowork could potentially provide everyday users with access to reliable legal knowledge and analysis (e.g., https://www.nationalmagazine.ca/en-ca/articles/law/opinion/2026/a_breakthrough_in_legal_research, and also, https://www.slaw.ca/2026/04/03/what-if-legal-ai-doesnt-need-legal-data/).

In a way, this is kind of like the old anecdote about the unstoppable force meeting the immovable object. Generative AI sure looks like the former, and the legal profession has definitely established itself as the latter. AI is pulling all of us into the future at an accelerating rate, while the law is digging in its heels to hold us back. But I suspect that, if consumer-facing AI products can someday provide people with real legal assistance, there is very little the profession or its regulators will be able to do to stop it.

So to me, this is a question of when, not if -- but I can't even make an educated guess about when the "when" might be. Richard Susskind used to ask his audiences: "Do you think the practice of law will be radically different in 100 years?" And his audience would all say, "Of course." Then he'd ask: "Do you think it will be radically different in one year?" And his audience would say, "Of course not." "Great," he would say, "now we have our time frame of change. Let's narrow it down." I don't think it will take 100 years for us to have viable consumer-facing legal AI products. I'd be surprised if we didn't have it in ten. But no one knows for sure.

Jon Van Loo's avatar

The proprietary data point is spot on. Companies like Bloomberg and Lexis are obviously going to fight to protect their data but I already find, much to my amazement, the additional help provided by Bloomberg with their rich store of data to be only marginal compared to a ChatGPT deep research search, which has the advantage of being far easier and faster.

Debbi Mack's avatar

I've asserted for a long time that more people should understand the law. And that too many people view legal services as a privilege for the rich.

I'm very interested in any thoughts on how to change the legal business model. And how to make sure the right people know about it.

I'm also a little concerned about using AI for anything other than the most administrative aspects of providing legal advice.

Kelli Dunaway's avatar

I'm excited to see what your talent framework looks like. This is the thing that keeps me up at night. How do we train Sphere 2 judgment if we eliminate the training ground in Sphere 1?

Jordan Furlong's avatar

Kelli, Part 2 will talk about the talent framework of these new enterprises in broad strokes, but it won't go into detail on the training of new lawyers -- it's just too big a topic to squeeze into these shorter articles. My response to Emma's comment (above) provides an outline of three potential ways in which we can reinvent lawyer formation, and I'll be writing more extensively about them in future.

Charlie J Writes's avatar

Strongly agree, I believe the next capability shift to achieve this will be how lawyers learn to sell this vision of their role (and price it). From a consumer perspective, the additional often unspoken benefit of a lawyer is having someone who is not emotionally attached to the matter or case taking carriage of it. Sort of like the “keep angry emails in your outbox for 24-hours” advice but on steroids.

Chris Hanson's avatar

I agree. If Ai is made available as libraries were once designed, it opens the door for lawyers to fully investigate and analyze the facts in unprecedented ways. It offers a return to real and demonstrable fact finding and truth. We've strayed to far from that as lawyers and culturally. Ai offers a path to fix that? Do people really want that though?? If they understood where we're headed without it, I'm sure they would.

Enjoying your work. Discovered u on a Florida cle. I've got a great story about leveraging ai to level the playing field in crazy litigation against dozens of powerful lawyers and institutions. Would love to share it with you sometime if interested?