Legal education's day of reckoning approaches

For decades, the legal profession has outsourced to law firms the early-years development of its newest members. Law firms, tiring of the deal, are in a position to demand an entirely new arrangement.

Illustration by Midjouurney

Law schools are deeply concerned with Generative AI. They understand that it augurs a potential revolution in how they teach their courses. They anticipate that students will use Large Language Models to write essays and complete other course work. They recognize that they need to think beyond their traditional understandings of what constitutes “cheating.” And in all events, they acknowledge that Gen AI is ushering in a new, much more challenging era for legal education.

I think they’re right on all these points, especially the last one — although maybe not in quite the way they expect. Generative AI’s impact has serious implications for the entire lawyer education, development, and licensing industry,

I use the term “industry” here deliberately. In 2021, there were 117,305 law students in the United States alone, each paying an average of about $73,000 a year for three years, for a total of $8.6 billion spent annually just on student tuition, books, board and expenses. And that’s just law school: The “non-profit” that designs the bar exam sits on $144 million in assets. The company that charges most law graduates to prepare for that exam pulls in about $60 million every year. There’s a long tail of other companies and organizations that also make money off the lawyer formation process.

Lawyer education and development is big business, in large part because it’s effectively operating as a monopoly. With very few exceptions, every law licensing authority in the US and Canada requires domestic licensure candidates to have earned an incredibly expensive three-year law degree from an accredited university. Law schools charge what they do in part because they can: The profession has appointed them its gatekeepers and has placed few restrictions on the manner, and none on the price, at which they keep that gate.

More to the point, everyone understands and accepts that despite its compulsory status, three-year term, and $200,000+ price tag, law school doesn’t and was never designed to make its graduates competent to practise law. Enter the bar exam and its associated prep industry, which don’t make lawyers competent either, but provide sufficiently plausible cover to allow bar admission authorities to issue law licenses to the lucky survivors of this gauntlet. Together, law schools, bar examiners, and bar admission authorities form a kind of licensing-industrial-complex that ensures a very small number of well-off (or deeply indebted) people become lawyers every year while generating a very tidy sum of money in the process.

As you might have guessed, I’m not a big fan of this arrangement. (I’ve written extensively about the crying need for lawyer licensing reform, as have others.) And so it’s quite possible that, in making the forecast below, I’m engaging in more than my usual amount of wishful thinking, hoping that maybe this complex will finally soon be broken. But I do think that there are forces in motion that will bring radical reform to lawyer formation over the next several years, and that AI will be the first pebble in that landslide.

Whenever I speak with law firms about Generative AI, one topic that comes up more often than you’d think is its implications for new lawyer training. Firm leaders recognize the likelihood that Gen AI will, within the next several years, shorten considerably the amount of time and effort new lawyers spend doing basic client tasks (legal research, document drafting and review, etc.)

Obviously, this would have an impact on hourly-billed revenue. But firms are also concerned because historically, by engaging in this work, new lawyers have familiarized themselves with the realities and practicalities of legal tasks and developed critical lawyer skills like evaluation, communication, and matter management. That’s not “training,” in any defensible sense of the word. But it is an immersive experience through which new lawyers intuitively learn by doing and piece together the bigger picture of client needs.

Now, even when well organized and overseen, this is a rough-and-ready, hit-and-miss approach to learning — and as you and I both know, it is often not well organized or overseen. The lack of proper supervised learning for new lawyers is a huge problem for our profession. But whatever its flaws, this approach does have its advantages — chief among them, that it’s better than nothing at all. And that’s exactly what’s got some law firm leaders concerned.

As part of their traditional business model, law firms assigned the most basic tasks to new lawyers, partly because they could bill that work to clients (although much less often than they used to), partly because that’s how they’ve always done it — but mostly, I think, because there’s almost nothing else new lawyers are really qualified to do.

This is maybe the number-one complaint lawyers and law firms have made over the years about the legal education and bar admission process — that it seems to accomplish so little. Three to four years of their lives, hundreds of thousands of dollars of their money, and yet new lawyers show up at the law office on Career Day One unready to do much more than write over-extended memos about the rule against perpetuities (which, yes, the bar exam is still testing). These people are bright, hard-working, and eager to please — but they are not competent to provide legal services, regardless of what their law license might indicate.

This is nothing new. Law firms have never liked it, but they’ve put up with it — and when they could still get clients to pay for new lawyers’ on-the-job training, there was at least some recompense for them. But clients — especially the large clients of large law firms that employ the lion’s share of new lawyers — aren’t willing to do that anymore. And now Gen AI is coming along to eliminate that basic work altogether, so that regardless of who might theoretically pay new lawyers to do it, that’s not how it’s getting done in future.

And so a difficult question is fast approaching: What will law firms’ newest hires do when they get there? Law firms are going to answer that question in ways that suit their interests, and I suspect neither law schools nor bar admission authorities are going to like those answers very much.

One option, which I very recently advised law firms to consider, is to develop their own structured learning programs for the early years of their lawyers’ careers. Rather than relying on high-stress, hit-and-miss, experiential immersion, build a multi-year supervised training system that accelerates lawyers’ professional growth, putting old heads on young shoulders and turning new recruits into highly productive, business-developing veterans years ahead of schedule.

I think this is a great idea. I also think that only a tiny handful of the largest and richest law firms in the world have the money, bandwidth, and long-term vision to actually do that. For almost every other law firm, those with 90 lawyers, or 25, or 6, this is simply not an option, and they won’t pursue it.

What else might they do? Well, they could try outsourcing this training process to a qualified third party that will do all this accelerated supervised teaching for them. Except that there’s hardly anyone out there offering those services, it would still cost a lot of money, and it would take new lawyers out of the office (if they’re even in the office) during their most formative first months.

And besides, those firms might add: What, now we need to pay a third-party institution to teach our new lawyers how to be lawyers? Then what the hell is law school for?

And that brings us to the two likeliest responses by law firms to the spectre of low-skilled new talent arriving in workplaces where there’s no low-skilled work for them to do.

One is that they will simply stop hiring new lawyers, or will hire them in drastically fewer numbers. They will instead poach more experienced talent from other firms, letting someone else do all the heavy lifting on early-years training. There’s some indication that this is already underway in the American legal market, and that lawyer development is rapidly devolving into a primarily late-career free-agency play.

There’s an obvious tragedy-of-the-commons element here, of course: If nobody is hiring and training new lawyers, where will all the mid-level and upper-year lawyers come from? That’s a real problem for the legal profession. But it’s not a problem for individual law firms, which will pursue their own interests, use technology to increase profitability with fewer lawyers, and ensure that it’s their competitors, not them, that are the ultimate losers of this new phase in the talent war.

The other option is that law firms — both individually and collectively, through formal associations and informal concords — will finally tire of paying large salaries to sub-competent new lawyers while law schools cheerfully pocket huge amounts of cash and bar admission authorities perpetuate a bar examination system that does little to inculcate or assess basic competence to practise law and serve clients. They’ll tire of hearing “Let the law firms take care of it” from the people who ought to be taking care of it, but who are very much not.

What could fed-up law firms do? Lobby their state bars and state Supreme Courts (and in Canada, law societies), using both financial and peer pressure, to shift the burden of lawyer development away from the private sector and back towards the regulatory authorities, where it belongs.

They could insist that putting law firms in charge of ex post facto competence assurance in a dereliction of regulatory duty — that it is inexcusable for regulators to give someone a license to practise law, knowing full well that that license is telling a lie to the person who owns it and the clients that person will serve. And they could make it clear that spending money to bring new lawyers up to a state of minimum competence after they’ve been licensed is not a job law firms asked for, want, or are willing to perform any longer.

If law firms do stop hiring and developing new lawyers, the legal profession will have a massive succession problem and law schools will have a massive revenue problem. If law firms decide to throw around their considerable weight and upend the entire legal education and regulatory system, that might have even more dramatic consequences. I don’t know if law firms will do either of these things, or if they do, when they might get started. They’ve put up with this nonsense to this point; maybe they’ll decide to put up with it longer.

But if I’m part of the lawyer licensing-industrial-complex, then either one of those possibilities, combined with the rapid strengthening every day of the forces driving them forward, would be keeping me awake at night — tense, worried, and constantly listening for the rumble of an oncoming avalanche.