We often assume that even as genAI steadily takes on more ‘process level’ legal tasks, that the ceiling of complexity will keep rising to accommodate this change. But will it? What if legal work ‘tops out’ and it is only volume and scale that keep increasing?
Topping Out
What is meant by topping out in this context? In short, it means that lawyers are expected to conduct certain tasks for a client, e.g. negotiate a contract’s terms, manage aspects of the legal work around a deal, provide direct advice, provide updates on legal changes that matter to them, advise and represent in disputes, and more. But, those tasks are not infinitely expansive. Eventually ‘a lawyer’s tasks’ reach a professional limit.
Now, genAI is set to absorb more and more basic tasks, (and some complex ones too), and the argument goes that this is OK, because legal work will always ‘get more complex’ and ‘new things we cannot think of yet, will need to be done’ – but is that true?
Overall legal demand probably will keep rising. The world itself doesn’t seem to want to call a halt to new regulations and new laws. But……the key point is whether the core tasks lawyers need to apply to this constant flood of new laws and new regulations, and ever-growing demand for help, actually changes the game – or whether the game stays exactly the same, there is just ‘more game’? The same applies to the ever-expanding length and density of contracts, and the volume of ediscovery materials. More scale, more volume, yes – but does the job really change? Does this increase push lawyers up and out into new territory that AI cannot impact? Right now, the answer seems to be: no.
One possible future is what the legal world does ‘manually’ just tops out, i.e. reaches a limit for the human input that is expected under the auspices of being ‘a lawyer’. There is nowhere else to go in terms of ‘vertical complexity’. GenAI eats up the process work around multiple tasks that junior associates would do, but the entire legal work environment doesn’t suddenly compensate by becoming a lot more novel in ways that demand new skills and new esoteric legal abilities.
Volume and Scale
Regardless of AI’s impact, what seems likely is that: overall corporate need for legal input will grow – because the ‘legalification’ of the world just keeps increasing; also that contracts will keep getting longer; and if prices drop for some outputs then demand for legal services will increase globally (i.e. Jevon’s Paradox).
But….and it’s a centrally important but, that doesn’t mean that the vertical scale of what lawyers do goes up in terms of becoming way more complex on an intellectual level.
I.e. will there be more legal work out there? Yes, it’s hard to see any real decrease given the current direction of how societies and their governments and regulators operate. But, that work itself doesn’t necessarily mean we need more and more legal super-brains to handle it. It’s just more verbiage. More hurdles. More data. More risk. More….just….more.
Therefore, junior associates whose ‘basic work’ is eaten by genAI don’t immediately find a huge new land of more complex work to instantly move into (not that they could do that more complex work immediately in any case, as they’ll still need to be trained to do it).
Likewise, partners at the top of their game don’t suddenly find that clients will now pay them for a whole new range of skill offerings. The clients won’t come to lawyers for psychological counselling, foot rubs, or chiropractic support. Lawyers will still be asked to be lawyers and to do the tasks that lawyers carry out. This legal world is not infinitely elastic.
Elastic Lawyers
Good lawyers are quite elastic. Even in these days of super-specialisation, lawyers should be able to handle more than one task, or one very narrow aspect of client need. An M&A lawyer during a recession can help out with restructuring, for example. A property lawyer who mostly does commercial property should be able to handle complex residential property occasionally.
But those are horizontal movements.
A law firm should also be able to handle both small flows and high-volume flows of work. A top client wants your brightest minds on a very important employment contract – just one document, but for a super-essential hire. Another client wants a due diligence project that will cover 100,000 documents. Again, these are horizontal movements for the most part.
Now, the employment contract really gets into a lot of areas of concern. It becomes a very, very long contract. But, each clause is still a negotiated set of terms between two parties framed in the context of the law, it doesn’t mutate into something entirely different. I.e. is it a long contract? Yes. Is it a lot to keep in your mind at one time? Yes. But, it’s not moving into a whole new realm, even if it involves some very smart legal work.
Meanwhile, the due diligence job is a large one, but if the 100,000 document collection moves to 500,000 it’s still the same set of tasks. There is just more volume of documents to check.
I.e. although we expect lawyers and law firms to have some elasticity, some ability to stretch their skills and to scale up and down in terms of work needs, those work needs don’t necessarily morph into something wildly outside any prior experience, or beyond the limits of ‘what lawyers do’.
What This Means
If total ‘gross volume’ of work goes up and the scale of work projects increases, e.g. contracts just keep getting longer and with more parts to them, but lawyers carry on being lawyers, while genAI absorbs more and more of the process level end of this work…..what happens?
At the broadest level it means that ironically enough commercial law firms are assured a bright and prosperous future, as more work means more money overall. Yet, internally the personnel structure needed, the teams needed, to do that work must surely change.
If in 2024 you need an associate to spend X amount of time on Y process level task, will that be the same in 2034? That seems very unlikely. The bit that cannot be commoditised, the bit where partners really provide a lifetime of experience-based advice and oversight, remains the same as ever. But equally, that partner-level input doesn’t rise up into a whole new world of esoteric activity previously unimagined by the legal world. What clients want from lawyers remains in the legal realm.
So, more matters, less time needed to handle the volume parts of each of them, hence, maybe we also need smaller teams for each matter as well.
Law firms become busier and the legal engine just needs to run more efficiently to handle this. Of course, genAI is there to both support this and is also driving this increase in volume. It’s a self-reinforcing cycle where AI soaks up drudgery, reduces the total amount of human labour (and time) needed, but also helps the firms to handle the increasing volume of work that will stem from our ever-evolving legalification of life.
Conclusion
Now, we could get into the economics of it, and there is a lot to unpack there, but suffice it to say the pyramid will stay in place. But the lower levels of the structure will need to be broken down into much smaller teams. Yet, those teams will take on much more volume of work than in the past.
The challenge for most law firms, as ever, is making that structural-economic shift, while clients become increasingly doubtful about paying for time taken at the lower levels of the pyramid. I.e. law firms have to handle growing commoditisation spliced together with selling high-level input as part of each matter, just as volume increases as well.
But, overall we get: more work, less time taken for each matter, smaller teams, and more AI (both driving this change and supporting the new legal economic model). It is a complex recipe for law firm management teams to handle.
Yet, the legal sector will adapt, as it always has. And the first step will be recognising the industry as we know it really has no choice but to change.
–
Richard Tromans, Founder, Artificial Lawyer, Nov 2024