Meet Lord John Thomas, former Lord Chief Justice, whose career has spanned the Commercial Bar, the judiciary and the digital transformation of courts in the UK and internationally. From helping drive major investment in court technology to exploring the role of AI across commercial courts worldwide, he has seen first-hand how technology can reshape access to justice.
We caught up with Lord Thomas to discuss the realities of implementing AI in courts and arbitration, what legal tech founders need to understand about how judges and arbitrators actually work, and why building reliable, easy-to-use AI tools could be key to improving access to justice for non-lawyers.
Photo: Lord John Thomas
Can you tell us a bit about yourself and your background?
I practised as a barrister at the Commercial bar for 24 years becoming a Judge of the Commercial Cour in 1996. It became clear we had to get the then IT system replaced. In 2002 I worked with practitioners and Richard Susskind to get this underway before I became a Judge of the Court of Appeal and held various leadership posts and got involved in wider digital projects. When I was Lord Chief Justice I got underway through Michael Gove, then then Lord Chancellor, the biggest investment that the Government has made in IT. Since my retirement in 2017 I have been President of the Qatar International Court and until earlier this year I chaired the Steering Committee of the Standing International Committee of Commercial Courts (SIFoCC); there are 70 member jurisdictions including China, India, the US and most of the G20. SIFoCC’s major current project is AI as the same issues face courts worldwide.
The theory of AI in justice is one thing; implementing it in real courtrooms and arbitration is another. What's the messiest, most challenging part you've encountered? Where do people's best intentions collide with reality?
The list could be much longer long, but I have frequently encountered each of the following:
- Money: Getting the UK Government to provide funds
- Selection: Ensuring the right system is picked. This requires massive input from knowledgeable judges/arbitrators; an analysis of the requirements of the system and a proper comparison between available systems, even if the supplier of some may be small; then ensuring the contract is properly negotiated and drafted which is far from easy to get done.
- Delivery: Keeping tight judicial control over the delivery by the supplier and rollout to the courts for the entire length of the process. This is often very difficult to achieve.
- Use: Ensuring that the users see the benefits as quicky as possible with proper technical help always at hand.
If you sat down with a legal tech founder building the next tool for international arbitration or courts, what's the one thing you'd tell them they're getting wrong about how arbitrators and courts actually work?
- Making sure that the tool only produces legal materials from a reliable database.
- Assisting lawyers in writing concise and focussed submissions, witness statements and reports; the availability of a greater quantity of material can very easily lead to greater prolixity and length.
- Providing accurate references that justify/explain statements made so that the statement can be easily checked where necessary – cross checking is very time consuming when writing an award or judgment but is essential as it is so easy to make a submission that the words actually used may not fully justify.
- Where the case involves the use of two or more languages, facilitating accurate and easily accessible translations.
After discussing these challenges with the Legal Geek community in October, what's one concrete thing you'd like practitioners, judges, or legal tech builders to actually do differently in the next year?
Concentrating on reliable and easy to use tools to provide for access to justice for non lawyers. Legal Geek was of great help in 2017 in organising a hackathon to this end, but progress has been far too slow. The need in the UK and across the world is immense, as lawyers are unaffordable for most. Although the monetary reward may not be great, the achievement and the results are of far greater value.
If this succeeds, then good AI to give real assistance to judges will be needed to deal swiftly with the use that will then be made of the courts.
