The reckoning of the billable hour, part four

Most lawyers are married to the billable hour. A few have at least flirted with alternative fee structures, daydreaming about the beach vacation and the piña coladas that some clever new arrangement might one day pay for. So it is worth asking: how will AI transform the fee arrangements between lawyer and client? Is this truly the end of the billable hour? Are we about to see the arrival of an entirely new fee structure? Or is it all just much ado about nothing?

Having run through the scenarios, my conclusion--for now--is this: the productivity boost and the increase in litigation activity that AI will bring to litigators are going to disrupt the current equilibrium of fee structures in the market. We can already see the direction of those disruptions. Whether the new equilibrium, months or years down the road, will look anything like the current one is anyone's guess. Here is how I think AI will affect the fee structures predominant among litigators today.

Pure contingency and success fees. Few arrangements align the upside and downside of litigation better than a pure contingency. The lawyer wins only when the client does and in proportion to the client's victory. What seems certain is that, in the immediate future, the supply of lawyers willing to take matters on contingency will grow. AI lets a lawyer dramatically reduce the hours needed to reach the same result, so even firms that have always shied away from contingency--wary of sinking thousands of unrecoverable hours into a matter that yields nothing--will start to dip their toes in. The natural effect of more lawyers willing to put their fees at risk is downward pressure on the contingency terms we see in the market. Expect more staggered contingency percentages, designed to be fairer to the client and to keep the lawyer from taking a slice of the recovery that is out of proportion to the work actually performed.

That sounds perfectly reasonable in the short term. But what if AI also turbocharges the adversaries these contingency lawyers face? What if those opponents become more vexatious because they, too, can now file more motions and make more havoc on the same budget? If that happens, contingency lawyers will recalibrate what they consider a fair percentage of the recovery. And then there is the lodestar. Judges are already skeptical of the fee applications that some firms submit, often observing that the hours claimed do not seem necessary. I expect those same judges to demand even more efficiency now that AI tools are available, and to grow more critical of firms that appear to be padding their hours for the fee application. If that comes to pass, it is good news for class members and other represented parties. In short: expect more lawyers offering contingency terms, downward pressure on the percentages clients will accept, and judges waking up to the fact that firms can--and should--do more with fewer hours given the tools at their disposal.

Flat fees and fee caps. Another form of "alternative" arrangement caps the hours or fees a firm will charge for a given quantum of work. At my firm, we often apply a cap when we do an initial case analysis, writing off the remainder as an investment in the relationship in case the client decides to hire us to take the matter all the way. Many flat-fee and fee-cap arrangements also carry some form of success fee on top to keep incentives aligned. More firms should be willing to offer these arrangements with AI at hand, for the obvious reason that they can get more done with fewer hours. But those same firms may be unpleasantly surprised to find that AI is just as much a force multiplier for their adversaries as it is for them--causing them to sink more hours than they planned, and tempering any sunny expectation that AI will make flat-fee arrangements more attractive across the board. The machine cuts both ways.

The billable hour. In light of all this, you will forgive me for observing that reports of the death of the billable hour are greatly exaggerated. That will disappoint technologists and clients alike, because few things in the legal profession are as hated as the billable hour. Everyone wants it dead. It keeps showing up to work. It seems oblivious to efficiency and arbitrary as a measure of work--after all, some lawyers accomplish far more in an hour than others, and how is a client to know whether the hours billed are a fair measure of the work done? And that is before we reach the way the billable hour misaligns incentives, rewarding the attorney for effort that may never produce a good outcome for the client. So why wouldn't AI spell the end of this dreaded, inefficient and misaligned system? Because, for all its faults, it is still better than many of the alternatives. There are at least three reasons.

First, there are matters for which the billable hour is the only logical arrangement. A case in which the client's goal is an injunction shutting down a competitor does not lend itself to a contingency. Neither does a representation whose desired outcome is some measure of equity that may never be liquidated, or preserving freedom from incarceration. Such matters may be better suited to a flat fee, but the flat fee carries the imperfections already described and may not present an acceptable formula to counsel.

Second--and this is something no one wants to hear--AI is as likely to increase the hourly rate counsel charges as it is to reduce the number of hours billed. Yes, firms will be able to do more with fewer billed hours per matter. But the hours that are freed up simply get spent representing other clients, so a firm ends up handling more clients and more cases for the same total hours--which increases its malpractice exposure. That exposure carries a real cost in insurance, and the only way to offset it is with a higher rate. Lawyers will also argue, naturally, that their greater productivity deserves a higher rate. And none of this counts the considerable cost of the AI products themselves--and of the other tools that now bake AI in and raise their vendor fees accordingly. It is impossible to predict with precision, but these are forces that could push the hourly rate up rather than down. AI may let firms replace some bookkeepers, clerks or administrative staff with automation--but it is just as likely that firms keep that staff and have them oversee more work, another cost that finds its way into the rate.

Third, for all the flaws of the billable hour, there is something inherently fair about paying someone for the hours they actually spend doing the work. Contingencies produce windfalls. Flat-fee caps can pit a client against their own lawyer, if the attorney feels constant pressure to write off hours and take a hit to their income. The billable hour, so long as it is kept in check by clients who follow the details, still seems fair in a way the others do not.

How much confidence do I have in these predictions? Not much. Treat them as observations and nothing more. The road ahead is untraveled. As Antonio Machado put it in his timeless Proverbios y cantares (XXIX) (1912), "caminante, no hay camino, se hace camino al andar" .... Translated: "wanderer, there is no path, you make the path by walking."  I have always loved that line. Perhaps that is why I took the risk of building a boutique from scratch, with very little idea how it would turn out.

In the fifth and final letter, I turn to the people who will have to live all of this: the making of a modern litigator, and what it will take to train one in the age of AI.