Dr Corsino San Miguel

Generative AI has occupied a similar half-lit status. For three years, deans and faculty tried to keep it at arm’s length – tolerated in policy papers, banned in coursework, whispered about in plagiarism meetings. With the arrival of Harvey AI in Oxford, The University of Law – and their American counterparts – and Luminance, Legora and Lucio Ai academic programmes in King’s, that phase is over. The power has been switched on. Law schools are no longer debating whether to let AI into the building; they are arguing, sometimes reluctantly, over how to re-wire the curriculum without burning the place down.

Three years at arm’s length

For most of the past three years, law schools have treated generative AI less like a library and more like a leak in the roof. When ChatGPT arrived in late 2022, the instinct was not to explore but to contain. Admissions offices rewrote forms to forbid AI-written personal statements. Honour codes were hastily amended. AI was added to the same paragraph as plagiarism and collusion: another way of cheating, not another way of thinking.

The language gave the game away. Students were told not to “use” AI, as if it were a performance-enhancing drug, while staff were quietly advised to “be vigilant” for machine-generated prose. Bar bodies followed suit, making applicants certify that every word in a pupillage form was their “own unaided work,” even as those same applicants knew that large commercial firms were already experimenting with AI-assisted drafting.

Inside the law schools, AI appeared mainly in two places: risk registers and tech-law electives. It was treated as an object of regulation, not as part of the craft. For a brief period, legal education tried to hold the line at arm’s length – to let the profession experiment with AI in its innovation labs while keeping the classroom safely inside its analogue comfort zone.

When AI becomes part of the craft

If the last few years were about keeping AI out of the exam hall, this year is about letting it into the workshop. The arrival of Harvey, Luminace, Legora and others is the clearest signal that the experiment has moved from the innovation lab to the curriculum. In the US, over 25 leading schools have signed up to give students and faculty free access to Harvey AI and co-design AI-infused materials for the 2025–26 academic year.

Across the Atlantic, the UK announcement feels like the moment the lights have finally come on. Oxford’s Faculty of Law is starting with Harvey as a research tool, giving academics the chance to push its boundaries before it takes root in the curriculum. Meanwhile, King’s Dickson Poon School has jumped further ahead: launching an AI literacy programme that blends hands-on access, workshops, and curriculum embeds. This move signals that competence in AI tools is now a core part of the law graduate’s toolkit. The University of Law, for its part, is embedding AI directly into professional training – the final frontier of legal education, where drafting, research, and procedure are drilled into muscle memory.

What links these moves together is a quiet, decisive shift. AI is no longer just something to regulate, to warn against, or to fear – it’s something students need to learn to wield. A tool of the trade, to be picked up, tested, and, where necessary, distrusted. From Oxford to NYU, the question has shifted from “What rules should govern this?” to “what kind of lawyer are we training if they never learn to work with it?”

Scotland remains conspicuously absent from these lists for now. But it’s not too late for Scottish universities to act – the window is closing fast.

Beyond the classroom: upskilling the profession

As universities start to make AI a core part of the legal curriculum, the real challenge is upskilling those who’ve already left the lecture hall. While future lawyers will be trained with AI tools, the legal sector faces a pressing need to equip those already in practice: trainees, junior lawyers, senior solicitors, and professionals in smaller firms, who are often behind their big-firm counterparts.

Large firms are already leading the charge, integrating AI into their workflows and heavily investing in upskilling their staff. Freshfields, for example, has partnered with King’s College London to offer a fully-funded LLM in law & technology to future trainees. Lawyers on the cutting edge of practice now navigate not only statutes and case law, but also sophisticated AI tools that assist with drafting, legal research, and compliance. For those in smaller firms or just starting their careers, however, the learning curve is steep – and without the same resources, many risk falling behind.

It’s time for legal education to expand beyond the walls of the university. This isn’t just a matter of providing future lawyers with the tools they’ll need. It’s about making sure the professionals who are already in practice – who weren’t trained in the age of AI – are given the opportunity to catch up. Training programmes, workshops, and courses designed for the legal workforce should be as much a part of the upskilling landscape as university modules. Without these opportunities, the gap between large firms and smaller practices will only widen, leaving a significant portion of the profession out of step with the future of law.

Time to rewire or get left behind

The lights are on, and AI is here to stay. Law schools are integrating it into the curriculum, but the real question is whether the profession will keep pace. Large firms are already embedding AI in their workflows, while legal professionals in small and medium-sized firms risk being left in the dark. Scotland’s law schools have yet to step up, but it’s not too late to make the choice: adapt now, or get left behind.

Dr Corsino San Miguel is a member of the AI Research Group and the Public Sector AI Task Force at the Scottish Government Legal Directorate. The views expressed here are personal.