The questions keep coming up: if genAI will increasingly be able to do process-level work, what happens to the most junior associates? What work will they do? Who will train them? Do we even need them?
At several legal tech events recently it’s been clear the question of how to embrace genAI’s productivity gains while operating a pyramidal, chronological advancement-based law firm model is testing plenty of lawyers’ minds.
The first question that comes up is always: who will train them in the future and who will pay for it? Well, the reality is that most businesses have to train their junior staff whether the clients directly pay for it or not.
At present the intake level of the pyramid (i.e. the ground floor) is both a ‘lawyer nursery’ and a way to make money via the billable hours needed for the young advocates to do their work, albeit probably not the most complex work at the firm.
They learn, the clients pay, work gets done, partners make money….it’s a winning formula (for firms). But, then add in genAI and eventually a big chunk of that work will be absorbed. Now, even if total volume of work goes up, e.g. because of Jevon’s Paradox, the lower-level work will still be affected.
So, what happens next? Here’s some scenarios and issues to consider:
- Can you operate a law firm without the most junior lawyers (or at least not many)? Well, the answer is: yes, of course you can…in theory. The problem is that law firms are based on a leverage model (at least for now). Moreover, those junior lawyers are the future. If you fail to hire this young talent where do the senior lawyers of the future come from? (N.B. software engineers are increasingly saying they don’t need to hire junior coders as genAI does most of that basic level work already…..)
- What if you hire mostly only 3rd and 4th years, and let someone else train the younger ones? Possible, but then you may lose the best talent who will stay with the firms that support them from day one.
- What if you hire just a very small handful of juniors each year? It could work, but attrition is massive at law firms. People leave, people are asked to leave, people who you want to stay decide they don’t want to do fee-earning work. The way around that is to hire a lot of people and expect many not to progress. It’s wasteful, but that’s how law firms work. So, if you reduce your intake then you need to do more prep ahead of them joining. Do they really want to be a lawyer? Do they want to progress….? But, how can they know until they’ve tried it? Hmmm….it’s tricky.
- Then we get to the economics. Clients pay for junior lawyer training via their lower level work product. But, if genAI does most of that, then what? Well, law firms will have to do that training directly, i.e. outside of any billable work projects, which will add to their running costs – and they may need plenty of external trainers and/or genAI training systems to help with this. Of course, that cost will eventually be picked up by the client, as firms will raise their rates once again to pass on the expense of this unbilled training.
- And some may say: don’t worry, these junior lawyers will just do much more complex work from day one………But….will they? Do you want a totally inexperienced junior lawyer, who because of genAI has not done much real-world work, getting pushed into a really complex piece of work that only more senior people would normally do? Nope. You want them trained first. So, there is no dodging the training part.
There are a lot of unanswered issues here. But, it seems that in a future where genAI does more and more of what the ground floor lawyers do within the leverage pyramid then we will see:
- Overall there will be less demand for junior intake per leverage group of a firm, i.e. partner, mid-level, and junior team members. (But, because of multiple reasons: Jevon’s Paradox, equality of arms increasing activity, and generally increased affordability across the legal market, the total number of juniors needed may not decrease across an entire country as the market for legal services will expand.)
- More lateral hiring of trained associates – which will no doubt cause tension between firms, as firm A throws a lot of money at upper-junior to mid-level associates to take them away from firm B, thereby reducing their own training costs.
- More training will be done ‘off the clock’ i.e. it won’t be billable, as that work will have been absorbed already by genAI and one cannot just throw a trainee lawyer into very complex work. That training cost will be passed to the client one way or another. So, there’s no escaping that. The cost will just be delayed a bit.
- Law firms may ask law schools and other providers to do more for junior lawyers to ensure additional training speeds up their development – and genAI itself e.g. legal simulators, may well be part of this solution. And there will be more external trainers and larger internal training teams inside law firms (or at least the firms that have the scale to offer this.)
Conclusion
The end of junior lawyers…? Nope. But, the environment will change for them.
The end of the pyramid model….? Nope. But, the bottom levels will also change.
In fact, what seems likely is that as on-the-job billable learning becomes scarcer at the ground floor level, the need for good old-fashioned education and training will increase to prepare juniors for the real action.
One last point…..the billable hour model, will that survive? Well, it’s another nail in the time-based system’s coffin, at least for the most process-level work. If you can’t leverage the juniors on this work to train them, or create billable time from that training, then time-based approaches at this part of the production line don’t make sense.
(To be continued, for sure…)
Richard Tromans, Founder, Artificial Lawyer, Oct 2024