“Law is probably the hardest of all the fields of knowledge work to transform with AI.”
That’s not a skeptic talking; it’s Richard Tromans, founder of Artificial Lawyer, who has spent more than a decade covering legal technology. And … he’s not wrong, is he?
As someone who studies and reports on the legal profession rather than practicing inside it, I find Tromans has a unique view of this industry that many of our readers aren’t privy to. He sees things from the outside looking in, unencumbered by the blind spots that come from being too close to the work.
That outside perspective is invaluable. So today, for the practitioners inside the firms and courtrooms, we’re spotlighting three 2026 AI Visionaries with unique “outsider” (in the very best way) perspectives: Professor Michael Legg, legal profession and education researcher; Oliver Roberts, legal AI educator; and Richard Tromans himself.
Let’s dive into what they’re seeing.
Legal Education Is Struggling to Keep Up
Oliver Roberts is straightforward about what drives him: he believes technology can fundamentally change who gets access to legal knowledge in the first place – expand it, democratize it, give people capabilities that were previously out of reach. That conviction shapes how he thinks about legal education.
But he also sees, up close, what happens when institutions try to address a fast-moving technology with slow-moving curriculum.
“There is a common misunderstanding that having one AI regulation course in the curriculum is sufficient to address legal AI education,” he says. “Legal AI education covers a wide range of areas, including practical use of AI, ethical considerations of AI, regulation of AI, and litigation involving AI.”
Legal education does unfortunately tend to be a slow-moving institution – a polar opposite to the whiplash the world experienced with the advent of AI. Rapid technological developments add significant workload for faculty, require curriculum overhauls, and demand a breadth of coverage that doesn't fit neatly into a single elective. It’s a tough challenge to solve.
However, Oliver is optimistic about what AI can do for legal education in the long run. And he also has a clear-eyed view of where the industry is heading structurally. As he noted in Bloomberg, for instance, big law firms could leverage their proprietary legal data to develop custom large language models – essentially productizing legal services.
“Economic pressures from AI efficiencies, alongside evolving client expectations, will catalyze that shift,” he says.
That move makes the education gap even more urgent: if legal services are heading toward that kind of transformation, the profession needs people who understand not just how to use AI, but how those tools are reshaping the business of law itself. Fortunately, many legal educators are on the case and eager to forge useful paths forward for their students.
What AI Can’t Do: The Case for Human Judgment
Professor Michael Legg has been carefully watching how AI reshapes legal education and what happens when humans and AI work together.
“Research has shown that human teams outperform individuals due to ‘collective intelligence’ and ‘team synergy,’” he says. “But AI and human teams combined outperformed a human team or AI alone.”
The implication for legal training is direct: if the winning combination requires a human in the loop, the question becomes what that human needs to bring.
Tech literacy is essential, yes. But Michael keeps returning to something that sometimes gets buried under simplistic “learn to prompt” types of conversations: “AI does not decide what real-world action to take in response to its output,” he says. “The human lawyer needs to be able to advise a client,” based on what insights the AI provides as well as her own inimitable expertise.
This means ethical judgment – the kind that naturally and necessarily draws on a lawyer’s humanity – remains a core professional skill. It also requires lawyers using AI to know what problem they’re trying to solve with it.
“An important issue is problem identification, not just problem solving. The client may have a view on what the problem is, but the lawyer needs to frame the issue to make sure they are addressing the real problem,” he says.
The Apprenticeship Pipeline May Soon Fundamentally Change
Fair warning: this is where the academic perspective may get pretty uncomfortable for anyone running a law firm.
Michael sees significant, systemic changes happening around junior lawyers’ development as AI takes over the laborious, detail-intensive work that has always served as immersive on-the-job training.
“With automation handling much of the routine work, junior lawyers may have fewer opportunities to learn by doing,” he warns. “Routine tasks often serve as a training ground for developing legal reasoning, attention to detail, and professional judgment.”
Document review, contract drafting, deep legal research – these were never glamorous jobs, but they were how lawyers learned to think. Firms may be absorbing the efficiency gains from AI without thoughtfully replacing the genuine good that can be lost to these types of operational transformation.
Michael’s prescription involves structured simulations, mentoring, and something he observes to be wholly underused: self-reflection as a professional practice.
“Junior lawyers, in particular, should be encouraged to self-reflect on two or three tasks they undertake during the year, or to even keep a journal of what they have learnt and what they would have liked to have learnt on an ongoing basis.”
It’s a small habit, but the effect can be huge.
From the Press Box: We’re Not in the Bottom of the Ninth Yet
Richard has covered legal tech since before most firms had an AI strategy. He’s watched what he calls Stage One of the legal AI revolution – NLP, early machine learning, the first wave of e-discovery tools – give way to Stage Two: the arrival of ChatGPT, rapid adoption, the emergence of AI-native firms.
The standard model of traditional law practice is still enormously profitable and still largely accepted by clients, so it isn’t going anywhere fast, predicts Richard. However, the writing is on the wall.
“AI will become so accurate, so universal, so effective, and so connected to all relevant legal knowledge, that every business in the ecosystem will have to fully embrace this transformation or be left behind. That is Stage Three,” he predicts. “Between now and then is basically a long, preparatory phase, where the legal world gets ready for the big leap to come.”
Take These Views from Outside the Firm as a Gift
Legal professionals building careers inside firms are, understandably, generally pretty focused on the immediate: clients, cases, deadlines. The work never seems to stop. Meanwhile, the academics and journalists covering this space have the luxury – and the obligation – of watching the longer arc.
What they’re seeing is a profession with real time to prepare, and a real risk of wasting it.
Oliver says it plainly: “Put in the work and dedication to follow AI updates and developments daily. Things change very fast in AI, and without a strong work ethic, it is very easy to fall behind. With that said, it also allows you to stand out and make an impact if you’re willing to put the work in.”
The legal profession has time. What are you going to do with it?
Graphics for this article were created by Kael Rose.
