The Law Society of Scotland
Annual Conference
Sir Geoffrey Vos, Master of the Rolls
Thursday 29 October 2026
Opening Keynote Address
Are human lawyers a thing of the past?
Introduction
1. I am so sorry that I cannot be with you at the Edinburgh International Conference Centre this morning. But I do think I have a good excuse. As Rachel Steer has said, I am retiring as the 98th Master of the Rolls (since 1286) on 31 October, and this afternoon, the Lady Chief Justice of England and Wales is holding a valedictory hearing for me here at the Royal Courts of Justice. I guess I ought to be there!
2. When Rachel Steer asked me to speak, she asked me to address the impact that AI will have on the skills of the legal profession, including the “the ongoing role of the human element, critical thinking, and human judgment v. technological capability, alongside AI governance, risks and opportunities for access to justice”. She asked me to be positive about AI. She probably did not need to say that!
3. I have spoken extensively recently about the use of AI in judicial decision-making. It is something that, in my opinion, is of momentous importance. In September 2026 in Seoul, I made a plea to the Conference of Chief Justices of Asia and the Pacific that they should get engaged in an international debate about the acceptable uses of AI within the judicial process.
4. That plea was prompted by my visit a few days before to the Shenzhen Intermediate Peoples’ Court. There, an AI-powered platform has been developed that allows AI to assist the judge
by distilling the important points to be derived from the evidence, listing the issues that the machine thinks need to be decided in order to determine the case, and providing drafts of the reasons and order that the judge might want to provide to explain and encapsulate their decision. Whilst that last part of the process is based on the judge having decided those issues first, one judge told us that she could, using the system, do 6 cases in the time she had previously taken to do one.
5. Without reprising what I said in my lectures to both Legal Geek and the International Chamber of Commerce on 14 October 2026, this utilisation of AI raises: (i) questions about how far we are justified in telling countries with significant court backlogs that they should not be using AI to expedite the judicial process, and (ii) questions about where the lines are to be drawn between acceptable use of AI to decide small disputes with the consent of the parties and situations in which it might be said that constitutional, and even human rights, barriers would be crossed if human judicial decision-making were to be abrogated.
6. That is not what I want to talk about this morning, but it is an important foundation for the debate, because the use of AI by lawyers is always going to be informed, even shaped, by the rules about AI made by judges and courts, and by the use that those judges and courts themselves make of AI.
7. I want to consider this morning the question of whether and how the legal community can ensure that legal skills, the process of critical legal reasoning and human legal judgment can be preserved in future generations. The title of this talk: “Are human lawyers a thing of the past?” is slightly tongue in cheek, since I guess you already know that I think the answer is a resounding “No”. But that does not actually make the question any less important.
The obvious premise
8. The obvious premise for the discussion is the reality that every litigant or business with a legal problem in fact consults
ChatGPT, Gemini, Claude or some other LLM to seek the answer as their very first port of call.
9. Moreover, most lawyers also use LLMs, whether publicly available ones or dedicated legal ones such as Harvey and Legora, in researching legal problems.
10. In this state of affairs, we will first need to persuade prospective law students that it is worthwhile studying law, when the answers can be obtained in two clicks from AI. I know that many Law Schools are very concerned about this problem, and about how they are going to be able to examine law when all students have unlimited access to AI.
11. Against that background, I still wonder whether we are being unduly pessimistic when we forecast, as some in all seriousness have done, the “end of lawyers”.
The outline of my cause for optimism
12. There are several reasons why I would reject the pessimism.
13. First, it is beyond doubt that law students, law teachers, professional lawyers, litigants in person, mediators, arbitrators and judges will all use AI extensively in their work – they, as I have said, already are.
14. Secondly, it is one thing to be told what the law is. It is quite another to accept it or believe it. Legal advice and judicial judgments are not automatically of value or accepted by the parties just because they are there. The acceptance of the legal advice by the parties receiving it is actually entirely based upon trust and confidence rather than just technical competence. In fact, something very similar can be said of judgments too. The question is not just whether the advice or judgment of the machine is reliable. It is whether the parties receiving it actually trust it and have confidence in it. That means that simply obtaining even an objectively reliable legal opinion from an AI tool is only the start of the process.
15. Thirdly, even if a human client obtains reliable legal advice from a machine, and accepts it, if that advice is negative, we know that humans take a long time to come to terms with an adverse position, whether it relates to their financial, property or private lives.
16. Fourthly, the law and the legal position are often actually only a relatively small part of solving a legal problem. Most legal disputes have underlying human issues that cannot be resolved simply by having a reliable opinion as to the law, whether that opinion is obtained from a human or a machine.
17. Fifthly, humans tend to value human advice and judgment very highly indeed. They are prepared even to accept idiosyncratic or, worse, wrong, advice or decisions if they think they have been heard. I often tell stories of irascible judges that I appeared in front of when I was a young barrister in the early 1980s. The clients were warned about how difficult the judge was. They observed that for themselves and accepted the outcome – just wishing they had been luckier and had drawn the more measured judge in the next door court. I emphasise, of course, that nothing of this kind can ever happen today …
18. Sixthly, life in the age of AI is hugely more difficult and entangled than it ever was. People and businesses have many more legal problems and issues caused by the complexity of modern life. This means that, even if some legal problems are likely to be resolved by AI, there will be many more intractable ones coming along that will still require human lawyers’ input.
19. In summary, I think that it is because of the human need to accept and understand and have trust and confidence in legal advice, even advice delivered by machines, I feel certain that human lawyers will survive.
20. So, I am actually quite optimistic about the future of lawyers, but I think they will be doing their job rather differently and they will need to adapt quickly to make the best of the a new AI-enabled legal practice.
If human lawyers are not a thing of the past, what do they need to do now?
21. I think that the legal profession will even benefit from the advent of AI if it embraces its advantages and learns to use it for the benefit of their clients rather than fearing and demonising the entire technology.
22. The real question, then, is what are lawyers going to have to do now to embrace the advantages of AI?
23. First, lawyers must learn to use it effectively to enhance their efficiency and effectiveness, to save time for their clients and ultimately to reduce the costs of legal services.
24. Secondly, I think, lawyers must recognise the massive potential for AI to increase access to justice, and to allow many more disputes to be resolved more quickly and at far lower cost than was ever possible before. This is a good thing. Dispute is crippling for the economy, and unresolved dispute even more so. Increased access to justice is a great prize.
25. Thirdly, lawyers must begin to recognise that it is unhelpful simply to focus on their fears that AI will take their jobs. The better approach is to learn how AI can help and where its usage may have risks, and to learn how to avoid and mitigate those risks.
26. Fourthly, I think lawyers should stop pretending that AI may go away and stop saying endlessly that AI hallucinates and is too dangerous to use. Every guideline issued by legal professions and judiciaries emphasises that lawyers and judges can only use the work product of a machine if they have checked it thoroughly and satisfied themselves that what they are producing to their client or the court or the parties is accurate and true. In short, the lawyers and the judges remain wholly responsible for their work, whatever AI they may have used.
How can lawyers and judges preserve the ability to undertake human legal reasoning?
27. There is, in my view at least, some confused thinking on this subject. The assumption is that, because Harvey or Legora can give plausible legal advice, lawyers will stop needing to evaluate that advice. The assumption is that, if judicial decisions can in some cases be provided by a machine, judges will stop needing to evaluate, check or deal with appeals from those decisions.
28. I do not think those assumptions are correct. Of course, the logical conclusion of machines being able to provide legal advice and decision-making seems to be that everyone will utilise them for that purpose to the exclusion of humans. But for the reasons I have already given, I think that is incredibly unlikely. Humans, as I have now said repeatedly, need to trust and have confidence in the legal advice and decision-making that they receive.
29. Clients will only have that trust if their experience demonstrates that it is justified. We know that eBay decides 60 million cases a year, 90% of them without human intervention and apparently without much push back, but these are very small claims. We should not assume that clients will necessarily extend that trust to more significant legal problems affecting their lives – like major commercial cases or cases involving losing their homes, their jobs or their children.
30. In my view, lawyers will still need to be trained to be able to evaluate the work products of dedicated AI tools.
31. I can return to the problem I mentioned earlier, which is whether we can persuade our young adults to continue to sign up to study law when they can seemingly get the answers on ChatGPT. That, I think, will depend on how relevant we make our law courses to the lives of a generation brought up with technology, the internet and AI.
32. I am sure we can do so, but we must not under-estimate the scale of the task. I heard from a fellow senior judge yesterday
that she had had the opportunity to address 4,000 young people online just this week. That is exactly what senior lawyers and judges should be doing to get the message across that careers in law are still valuable to their communities and to provide what I have described as the great prize of access to justice.
Conclusions
33. So, let me try to sum up my conclusions on whether and how the legal community can ensure that legal skills, the process of critical legal reasoning and human legal judgment can be preserved in future generations?
34. I suppose the answer is that this can and should be achieved by embracing AI – by understanding what it is good for and what it is not good for, so that lawyers can make their legal advice services, and judges can make their justice systems relevant and effective for the new generations of citizens who are being brought up with unlimited access to Gemini, ChatGPT, Copilot and Claude.
35. That does not mean forfeiting or even compromising our principles. It does not mean giving up on human legal advice and human judicial decision-making. But it does mean embracing the opportunities that AI provides to expedite the processes by which legal advice is delivered and legal decision-making is undertaken, and to reduce the costs that are associated with legal process.
36. As I have tried to explain, the law is all about trust and confidence. In many situations humans need other humans to give them that trust and confidence, but in some they do not. Lawyers should be realistic about what can properly be done by machines. They should avoid protectionism. If they do so, both the process of critical legal reasoning and the legal profession will have a bright future. And human lawyers will certainly not become a thing of the past.
37. Thank you for listening.
The post Speech by the Master of the Rolls: Are human lawyers a thing of the past? appeared first on Courts and Tribunals Judiciary.