Behind blue eyes: Shifting responsibility in the legal profession

I am referring to accountability, and the experience that underpins it. The legal profession is built around the premise of ‘selling’ a practitioner’s ability to be held accountable. Transactional lawyers ‘sign off’ on the legality of an arrangement and inherit part of the liability for when things go wrong. The value of this accountability is tied to the experience that informs it, which demarks senior from junior, and bolsters the authority of more seasoned practitioners’ sign-offs.

As a result, the profession currently moulds its future lawyers around the centrality of experience. Graduate lawyers are admitted under a condition of supervision, where more seasoned practitioners agree to guide them until they have ‘seen enough’ to practise on their own account. The assumption is that accumulating a certain volume of work is how one ‘earns the right’ to give their own legal advice.

This framework has survived for many years, but, now, machine output is beginning to replace the work that traditionally informed ‘experience’, and the graduate’s formation has radically changed.

The profession, however, still asserts that the advent of AI fits within its structural bounds. There is a response to the concerns that AI is doing what grads used to. Law firms have adopted the slogan that:

AI won’t replace lawyers; it will replace lawyers who can’t use AI.

This statement asserts that grads are still earning their stripes, but are simply doing so in a different manner. It is fundamentally based on a misunderstanding of what a graduate’s nine-to-five currently looks like. The assumption that underpins the statement is that AI is a tool rather than an agent, and that the challenge for grads has simply shifted; we now have to figure out the areas we can add value above and beyond the tool itself. Insert yourself into the limits of the machine’s capabilities, and figure out how you can steer its performance. There is a simple problem with this logic. The machine evolves every single week, and the limits keep shifting. It’s hard to build a career around a gap that is measured against a moving line.

VIEW ALL

The reality is that AI does not fit neatly into the current layout of legal professional development. The introduction of an in-house tool has completely changed the way my graduate program looks to those before me. Unlike the position that grads were in even one year ago, I can’t remember a task I’ve been given that I couldn’t get an AI platform to produce a first cut for. Where grads before me used to draft, it feels like I am now expected to supervise. The legal profession was designed around a process where juniors produced and seniors reviewed, but what changes when a third actor produces for juniors to review and seniors to then review again?

The experience upon which accountability is based starts to disappear. Hours of repetition and thousands of first drafts created a familiarity with processes that you can’t teach. Well, we now have an entire generation of junior lawyers who are watching something else perform these tasks for them and have begun stamping it with an approval that, unlike their predecessors, they haven’t earned.

For the ‘intellectually curious’ grad, this creates an economic problem. High-achieving university leavers don’t want to be given a novel task and ship it off to a third party. Anecdotally, it seems like many grads still perform first cuts themselves, to develop the familiarity they are surrounded by. Where a machine could produce something similar quicker, the grad’s instinct operates as an ‘apprenticeship tax’ on their employer.

This is because curious grads used to learn by producing a first cut themselves and billing it to the client. Now they learn by displacing the thing that could do it for them in seconds and billing tasks for longer periods than they otherwise could. This is an economic liability in hiding, and it feels like a matter of time until the profession catches up. Education can’t ride on production if a client isn’t willing to pay five hours for what a machine did in five minutes.

If law firms want to continue to hire ‘curious grads’ (as they all say they do), then they need to find a workaround to this problem. It could very well be that grads simply become a more expensive investment. Firms may pay for their grads’ learning rather than billing it to the client, and hope to reap the rewards when they mature into mid or senior lawyers. This works because mid-level lawyers and seniors offer value that AI can’t replicate. While an AI can review a share purchase agreement in seconds, it can’t build human relationships, make human judgements, or be held accountable when things go wrong, which is exactly what the average mid and senior does on a daily basis.

If the profession responds to the above problem by accepting that grads will (and need to) continue to do first cuts of work themselves, then they take a short-term economic hit in the hope of reaping a grad’s value when they turn into mid-level lawyers. They will also have to accept, however, that the investment they pour into a grad has a three- to five-year maturity date.

The problem for many top-tier firms is that this is the exact window in which lawyers look to move overseas or pivot in-house. Why invest in people who don’t want to stick around to pay off their debts? They could hire laterals, but this just moves the problem elsewhere. Will mid-tier firms take on bright university leavers, knowing they will likely be poached when they become hireable?

I am yet to hear a compelling answer to any of these questions.

Law is built on accountability, and accountability is based on experience. AI is removing the experience, and it’s time for the legal profession to acknowledge what this means for its structure.

This piece was penned by an anonymous Australian graduate lawyer.