How long before we’ll all be getting our legal advice from AI?

WORDS: Brendan Nyst PHOTOGRAPHY Brian Usher -

Good question.

It should come as no surprise to hear that many lawyers – and businesspeople in general – are already accessing some level of AI as a useful tool to help them maximise their efficiency and effectiveness in delivering services to their clientele. But at the moment, we’re all just taking our first baby steps on the AI journey, and where those steps eventually may lead us is anybody’s guess. Will robots eventually replace lawyers? I doubt it, but who knows?

All that can be said for certain is that ChatGPT-type technology, in its current state, is often already an effective legal research tool to help make an initial assessment of the meets and bounds of many legal problems. All the written statute law and legal principle is out there in cyberspace, along with records of innumerable previously decided cases, just waiting for AI to dig it up, distil it, and then spit out a potted treatise on the issue. That means, instead of lawyers combing endlessly through dusty law books and comparable cases, theoretically at least they can now get a useful summary on the subject matter with the push of a button.

One should be cautioned, though, even in that context, that AI may not always be entirely on the money. In a recent case in the USA, an American attorney found himself with more than a little egg on his face when his written submissions to a court were found to contain citations from supposedly earlier-decided cases, which were in fact entirely fictional. It turned out the lawyer had used AI to assist him in his research and didn’t think to double-check the information he was given. For some reason known only to the AI (but, then again, maybe not), the AI simply decided to create a bunch of bogus case citations out of nothing but thin air.

That kind of glitch, it seems, is always possible at the moment, but all signs suggest it will very soon be a thing of the past. Undoubtedly, as the technology continues to advance, and the product becomes more error-free, we will see more and more reliance on AI in the delivery of legal services. I recently had lunch with the Dean of a prestigious local university law school, who explained to me the ways in which academia is already looking to adapt curricula to anticipate the prevalence and importance of AI in legal practice. They hope to actively teach students how to reliably and effectively incorporate such technology into their work, to improve the quality of legal services.

But, of course, the law isn’t just about what’s written down in books or decided in old cases. It’s much, much more than that. Law is essentially about the interaction of often flawed and always complex human beings with each other in every unique circumstance conceivable, with all the attendant nuance, subtlety and substance one would inevitably expect.

Whilst AI will no doubt eventually surpass what any flesh-and-blood lawyer can offer in terms of the analysis of data, it’s hard to imagine how a mere machine will ever truly grasp the essential elements of the human condition, experience and sentiment that will ultimately dictate what we see as justice. Maybe one day it will but, for now at least, it’s probably fair to say that, while AI can help identify the problem, it can’t serve up the solution. That takes a smart, insightful and hardworking, real-live lawyer. For now.