Ad/Marketing Communication
This legal article/report forms part of my ongoing legal commentary on the use of artificial intelligence within the justice system. It supports my work in teaching, lecturing, and writing about AI and the law and is published to promote my practice. Not legal advice. Not Direct/Public Access. All instructions via clerks at Doughty Street Chambers. This legal article concerns AI Law.
Introduction
On researching a different topic, loosely ralated to AI in County Courts, I stumbled on a Justice Committee session where two senior figures in the civil justice system, the Master of the Rolls (Sir Geoffrey Vos) and the Deputy Head of Civil Justice (Lord Justice Colin Birss), gave evidence on how our County Courts in England and Wales are functioning.
Summary
There was a clear consensus that the County Court needs a robust overhaul of its paper-heavy processes. The speakers described how, in some courts, entire floors are taken up with files, and staff are forced to handle heaps of paper. While these staff members show great dedication, the delays and difficulties caused by paper-based systems hold back timely resolutions.
It was suggested that many of these problems could be remedied by a unified digital platform that would allow everyone, court administrators, judges, and litigants, to see case information and relevant documents promptly. Although much has already been spent on such reform, it seems that more targeted funding and carefully scoped improvements remain essential.
The conversation then turned to recruiting enough salaried judges, especially in London and the South East. The Master of the Rolls mentioned that heavy reliance on fee-paid (deputy) judges creates less consistency in case management. While fee-paid judges do sterling work, the courts function most efficiently when a stable, salaried bench is in place, building a stronger local relationship with court staff and the community. Providing more certainty in judicial postings, especially for individuals who do not want to relocate, was suggested as a way to boost recruitment. A strong, well-resourced, and digitally adept judiciary was presented as key to tackling backlogs and ensuring the courts serve local communities effectively.
On alternative dispute resolution, particularly mediation, they highlighted that many claims, can be resolved without a full hearing. Requiring parties to engage in an hour of telephone mediation can help people see where there’s room for settlement, rather than prolonging disputes out of frustration or misunderstanding. It was emphasised that settling before court is often cheaper, faster, and less stressful. Still, care must be taken to ensure that more vulnerable parties are not placed under unfair pressure.
Finally, and of particular interest to me, the Chair asked “How do you think artificial intelligence can affect the work of the County Court and improve it in the near future?”
How AI will affect the work of the County Court
The Deputy Head of Civil Justice (Lord Justice Colin Birss) responds that AI could be used as an assistance tool rather than a decision-maker, highlighting two examples:
- Summaries for Judges. Senior judges often receive short case summaries, but district and circuit judges do not, due to resource constraints. AI could generate brief overviews of each file, helping judges instantly see what a case involves. This would save them from having to open and sift through multiple files late in the day and could speed up preparation.
- Transcription. AI-driven transcription of oral hearings was cited as another potentially valuable tool, reducing the time and cost involved in getting accurate records.
The Master of the Rolls (Sir Geoffrey Vos) gives an example of a barrister using AI (e.g., ChatGPT) to produce chronologies and lists of parties. After verifying the AI’s output, she reportedly saves around five hours of work per week. This, he explains, is how AI should be seen: a tool to accelerate tasks and free professionals to focus on the essence of justice, rather than a mechanism for replacing human judgement. He emphasises that judges must follow guidance preventing them from relying on AI to write judgements in place of their own reasoning. They must remain the ultimate authors of each decision.
Finally, the Master of the Rolls stresses that it is important not to fear AI but to use it sensibly, as changing circumstances and modern technology have always reshaped society. He suggests that with careful oversight, AI can indeed help the courts deliver faster and more accessible justice.
Comment
These are further examples of very interesting observations by seniour judges on the benefits of AI in legal practice. I was particularly struck by the following observation by the Master of the Rolls:
“I gave a speech on AI at a Lawtech conference a month ago, and a young woman barrister practicing in the Employment Tribunal gave a speech about how she used AI in every single one of her cases to create chronologies and lists of participants for the court. She checked them over, they were done by AI but she then checked them over. She said, “I have set all this up on my computer because AI is not something that is done to you, AI is a tool that you have to use for the kind of work you do in an appropriate way,” and she said, “I am an employment barrister, I do a lot of cases that are 18th March 2025 28 quite similar. I have set up tools using ChatGPT and I save, every week of my life, five hours of my time by doing those things that give me a head start in my cases.” And that is not risky because she has checked everything, it is not dangerous, because she has checked everything, it is just time saving, and we would all use a hoover instead of sweeping the floor with a dustpan and brush, and so we should consider how we can use AI to save time and to make sure we get better, cheaper outcomes for the people we serve. We are not here to do justice for ourselves as judges, we are doing justice for the people, for the citizens, and for the business of this country and I think it is incredibly important with AI actually with everything that we have talked about to understand that.”
One of the benefits of this blog is speaking to many lawyers in numerous jurisdictions about how they are using AI in practice and, similar to the example here, finding that it is greatly improving efficiency and cutting down on repetitive tasks. Many report that AI-driven drafting and research tools not only reduce hours spent on the mechanical aspects of legal work, but also help them tackle heavier caseloads and release more time for deeper client engagement. Over the longer term, such gains can lower the cost of legal services and increase access to justice.
However, this experience is not shared universally. Discussions include the following issues:
- Some practitioners remain wary of the potential for unintended errors, pointing out that AI sometimes hallucinates facts or misinterprets legal nuances. Thats a valid concern, but with careful supervision, and full awareness of these potential pitfalls, this may be surmounted.
- Some practitioners report that, rather than saving time, using AI can actually increase it. If AI provides no clear efficiencies, it may not serve the client’s best interests unless another identifiable benefit can be derived. Several practitioners say they have found that the time spent checking the accuracy of AI-generated outputs, verifying references, and rewriting or correcting errors can equal or exceed the time it would have taken to do the work themselves. Although AI can produce an initial draft more quickly, the final review and editing needed to ensure precision often makes the process more time-consuming overall. Practitioners sometimes compare this to training a pupil barrister or trainee solicitor, given the similar level of oversight required to ensure correctness before sending the work to a client.
- Confidentiality concerns. This area is particularly troubling, and my position is not yet settled on how best to resolve it. Lawyers must ensure that any client information fed into AI systems remains securely protected, especially since many AI models rely on data centres or cloud services. I’m told a number of firms have imposed strict guidelines on how these tools should be used, insisting on regular human checks and robust data security protocols. Whether such measures are fully sufficient is a question for a more detailed discussion in a future post. Nevertheless, it must remain at the forefront of every lawyer’s mind.
- Training the next generation of legal professionals. While experienced practitioners can gradually integrate AI into their routine, that process may present particular challenges for future lawyers. Experienced lawyers, already classically trained, can spot how AI usefully supplements their established practice; but what about new lawyers whose first exposure to legal research and drafting may be shaped by these digital tools? They need a grounding not only in AI’s capabilities but also in traditional methods, so they learn how to spot errors or omissions that a piece of software might miss. It’s important, too, that they develop strong legal judgment, balancing the need for efficiency with careful professional oversight. If students rely too heavily on AI at the outset, they risk under-developing analytical skills and a thorough understanding of key legal principles.
I am going to add two of my own thoughts, which I hope will contribute meaningfully to the ongoing discussion in this area.
Firstly, the AI tools we use today are not the same tools we were using in 2023. That may seem obvious, but it is important. The technology is evolving extremely rapidly, and the benefits and concerns I highlight here are unlikely to remain the same by 2026. As a legal profession, we are currently playing catch-up, we tend to be reactive rather than proactive, largely because the pace of technological change makes it nearly impossible to anticipate effectively. It may well be that AI development slows sufficiently in the coming years, giving us, as a profession, the time to fully digest and strategically implement these tools. However, I fear this scenario is unlikely. The more probable outcome is that AI technology evolves exponentially, and at that stage, the pressing question may shift from “How can we effectively use AI tools in practice?” to “What value can we, as legal professionals, add to an AI-driven system?” I appreciate that this is a daunting prospect for many, nevertheless, it is something we may soon need to seriously consider.
Secondly, we may benefit from carefully considering the challenges we could encounter if, for any reason, these AI tools become temporarily unavailable, particularly as practitioners and judges increasingly come to rely upon them.
As AI tools become integral to legal decision-making and evidential analysis, courts and legal practitioners risk unintentionally losing, or significantly impairing, their ability and confidence to operate effectively without them. Consider, for example, scenarios involving technical failures, cybersecurity breaches, licensing disputes, or unexpected regulatory restrictions on AI use. If the courts become dependent on AI-generated insights, even temporarily losing access to these tools could seriously disrupt proceedings, causing delays or compromising the quality of decision-making.
Ultimately, while AI represents an extraordinary opportunity for enhancing legal practice, we must proceed cautiously. Maintaining independence and resilience within the profession is essential. Please continue to reach out to discuss these important issues. These issues are things we must address collectively.
