In early December 2025, Corsino San Miguel wrote an article for Scottish Legal News in which, drawing on selected examples from law firms and certain law schools in England, he advanced the claim that Scottish law schools are lagging behind in their engagement with generative artificial intelligence. The article is framed by the broader assertion that AI represents a transformative development to which legal education must urgently respond. San Miguel underpins this claim with an analogy with the historical introduction of electric lighting into the law faculties of the Universities of Glasgow and Edinburgh. On this basis he appears to argue that meaningful progress in the adoption of generative AI requires substantial financial investment in bespoke software solutions, particularly those being developed by emerging legal technology companies, which are primarily oriented towards commercial legal research and the working practices of law firms. Unfortunately, San Miguel did not speak to any of the Heads of Scottish Law Schools when preparing his article, so he gives a partial and misleading picture which we want to correct.

It needs to be pointed out that the work of legal tech start-ups is not exactly unknown in Scottish Law Schools. Edinburgh Law School hosted a regional final of the 2025 edition of the Global Legal Tech Venture Days in February 2025, with the first prize carried off by Amender (at that time, Edinburgh-based), who later won the intercontinental final in Madrid. Part of the well-attended event, bringing together colleagues from many different Scottish law schools, comprised a roundtable on Legal Tech in Scotland, offering insights on ‘smart’ and ‘good’ governance of technology in the legal sector. But it would be premature or even inappropriate to move towards giving students, even students on the diploma in professional legal practice, practical instruction on using specific legal AI tools: there are many legal tech startups at the moment and it is impossible to say right now which of their products, if any, will become widely adopted (or even which of them will still exist by the time the students enter practice). Law Schools work closely with those in legal practice in Scotland, not least through the work of the Joint Standing Committee on Legal Education in Scotland, where AI is a standing item on the agenda. We know that AI is potentially transformative for ‘law work’. We are not hiding from this. Legal technology is widely taught on the diploma and elsewhere in the legal curriculum. However, we think that rather than investing in tools which might be outdated before students come to legal practice, we would do better to make use of the work of bodies such as the Law Society of Scotland, which has produced a guide for use of generative AI, which is one product of the important LawscotTech initiative, or to listen to and collaborate with other bodies with training responsibilities in the legal sector, such as the Faculty of Advocates and the Judicial Institute.

Writing in a personal capacity, Craig Anderson of Stirling Law School addressed some of the general points about the nature of GenAI made by San Miguel, and its relationship to legal education, in a response also published in Scottish Legal News. He focused on the limitations of GenAI and the wider scope of the ‘critical thinking’ goals of legal education compared to that which GenAI tools currently can provide. It is important always to focus on the distinctive contribution of law schools as partners within legal education and training more generally. It is widely argued that GenAI, by imposing a certain type of rationality on the world, can disable our capacity for critical thinking and therefore for conceiving of and delivering change. Thus, our concerns about the role of GenAI in the legal education process go well beyond the fear that the misuse of GenAI in assessments represents a form of academic malpractice although that is certain a worry. However, it is one which universities are obviously aware of and seek to find ways of managing, either through detection or by encouraging the use of forms of assessment which focus on the individual or “group” thinking capacity of students, including their capacity to use legal tech ethically.

The problems with GenAI tools are manifold, when we apply either Scots law lenses or generic ‘legal education lenses’. The products are often inaccurate, especially when it comes to Scots law; in that sense, they are currently often useless although they might get better in the same way that voice recognition software has got better at decoding different accents and languages or specialist terminology. Second, passing the work of a ‘chat bot’ as your own is academic malpractice and dishonesty and such dishonesty could raise questions for someone’s future practice ambitions in the legal profession (and of course there are also plenty of well publicised examples of AI usage in legal professional practice to discuss with students to help them understand the point). Some courses can and do encourage or even require GenAI use, but this needs to be acknowledged. In a core first year module at Dundee Law School, students are asked to use an AI tool themselves and critique the results. At Abertay University, a member of staff uses ChatGPT to teach critical thinking by requiring students to criticise a ChatGPT answer to a legal philosophy question. A similar approach is adopted at the University of Stirling in a module on arbitration, where teachers encourage students to interrogate the potential discriminatory outputs of AI reflecting input bias. Such an exercise is also embedded within a range of modules from Contemporary Issues in Criminal Justice to a commercially-focused skills module.

So, of course, we don’t simply say to students ‘don’t use AI’. We engage them, as far as we can, in thinking about these limitations and problems, and encourage them to develop the skills to know what, at present, AI is useful for, and where it is not going to be helpful. For example, in Edinburgh Napier University’s compulsory Law and Technology module, students look at generative AI both for the societal and regulatory issues that it raises as well as its potential and actual use in the legal profession, the judiciary, and the broader legal system. Students benefit from a guest lecture from the co-heads of Shepherd and Wedderburn’s Smarter Working team where the students (and the module leader) get a constantly updated picture of how law firms are approaching AI adoption in the workplace. Dundee Law School runs law and technology modules which educate students on the use of AI and legal tech generally, in legal practice. This has been a strategic priority for several years, and the School is working with partners in the city to expand provision in a purpose-designed space. At Glasgow Caledonian University, sessions on AI literacy have been added to core classes Introduction to Legal Systems and Study and to a new module in Advanced Legal Studies and Skills, running for the first time in 2026-27.

Third, and above all, there are the problems of cognitive offloading: using AI as a tool to replace your thinking means that you are not developing the critical skills which are precisely those which led humans to conceptualise and then develop large language models in the first place. As educators, we are alive to the confidence issues that arise when students think that AI is smarter than they are. We undoubtedly have a duty to support students in the ethical and effective use of artificial intelligence, rather than just punishing them if they do it ‘wrongly’.

These are not issues which are confined to law schools or legal education. In each of our higher education institutions, we work closely with academic and professional services colleagues to preserve the integrity of assessment. Staff receive training in the quality assurance aspects of AI usage. So, for example, at the University of Glasgow, the School of Law has developed a policy document advising colleagues in relation to the use of AI in student assessments. This aligns with University and College of Social Science policy and requires that students are informed about whether AI use is completely banned from any particular assignment or not. There is scope to allow it to be used in various ways, again with a specific instruction to the students about this. Colleagues are also advised about what to do if they suspect unauthorised usage.

Along with what some might suggest is the retrograde step of a return to more in-person unseen written examinations, there has also been an explosion of interest in practical skills assessment, portfolios and oral presentations, not least because we are encouraged by the skills framework provided by the Law Society of Scotland for the qualifying law degree to think deeply about these matters and to provide appropriate diversity to test different skills. Of course, these different approaches to assessment can themselves present challenges in terms of the fair marking of assessments and ensuring that teachers are both trusted in their professional skills but also held accountable within the framework of the external examiner system and other quality assurance benchmarks. On a related matter of student support, the University of West of Scotland is already experimenting with the use of chatbots, to give routine informational support to students and this will become more widespread.

In our teaching and research, often collaborating with scholars in other disciplines such as data scientists who help us to understand the frontiers of the computational issues which underpin GenAI as a technology, we try to stay at the forefront of the discipline. For example, Dundee Law School works with the Leverhulme Research Centre for Forensic Science, whose staff deliver an optional module in digital evidence. As legal education specialists, our work is closely intertwined with that of philosophers and ethicists. Several Law Schools across Scotland have colleagues participating in the delivery of a large UK Research and Innovation investment, including in postgraduate research scholarships, on responsible AI use. Together we can explore regulatory possibilities (and limitations) in such a fast moving arena. There are numerous colleagues in Scottish law schools publishing excellent research on these topics, including as regards the challenges that AI poses specifically for law schools. What are the interactions, for example, between AI and copyright? Overall, GenAI use is undoubtedly increasing and universities are responding to this. At the University of Edinburgh, all staff and students at the university have access to ELM, the Edinburgh Language Model. This is designed to encourage using GenAI within a more controlled framework where we are not constantly uploading our personal data or professional insights onto the servers of technology companies. Across Scotland, environmental law scholars have been keen to highlight the potential environmental impact of the adoption of AI tools. They argue that we therefore have a responsibility to think carefully about if and how AI tools are embedded into educational practices.

At Edinburgh Napier University, students organised a conference on AI in the legal workplace in 2024, with the support of staff. The conference had speakers from Shepherd and Wedderburn and Avail (a legal tech startup) as well as a contribution from a law and technology expert in another Law School, Professor Burkhard Schafer, who gave the keynote address. At Edinburgh Law School, there is a long history of teaching about innovation and intellectual property law, including issues of information technology, especially at masters and honours level. This has been complemented, more recently, by an Introduction to Law and Digital Technology course, initially for students on the Global Law LLB and offered in the first two years of the degree as foundational for legal thinking skills. Strathclyde University has an LLM in Law, Technology & Innovation. The University of Aberdeen now offers the LLB with options in Computer Science. Robert Gordon University will launch a new LLB degree in 2027 entitled Law with AI, and in common with other institutions has a Legal Technology module on the Diploma. Glasgow University School of Law is developing a short, practitioner-oriented course – a form of executive education – on Tech Law and AI for delivery online and in London, in partnership with the British Institute for International and Comparative Law. The school also offers a number of masters level courses in the area.

This article was written using evidence compiled from an informal survey of the heads of the Scottish Law Schools. It is evident that there is now quite a lot happening in this space, and it has been happening for many years with several schools having 10+ years of experience. However, it is not tied to the use of particular commercial products. It is hoped that the approach taken by Scottish law schools is not only ethical but also environmentally and financially sustainable. Ironically, as pointed out by one of the respondents to the informal survey, even before this text was compiled and – once online – probably swallowed up into various large language models, a simple AI-enabled internet search actually produced a lot of the same data that also came out of the more discursive and perhaps less scientific approach of just asking people what they are doing. It also offered the following and quite revealing headline to the search results:

The following Universities are represented in the Committee of Heads of Scottish Law Schools, normally through their head of school or head of law: Abertay University, Edinburgh Napier University, Glasgow Caledonian University, University of Aberdeen, University of Dundee, University of Edinburgh, University of Glasgow, University of Stirling, University of Strathclyde and University of West of Scotland. The convenor, in 2025-2026, is Professor Adelyn Wilson of the University of Strathclyde and the drafting lead for this article was taken by Professor Jo Shaw of the University of Edinburgh.