Artificial intelligence (AI) is set to change the way lawyers and judges work, the Master of the Rolls said last week, predicting it will be used to take some – “at first, very minor” – judicial decisions.
Sir Geoffrey Vos warned too that “businesses will ultimately not want to pay for things that are available free… and that applies as much to legal services and dispute resolution as to anything else”.
Delivering the McNair Lecture at Lincoln’s Inn on the future of London as a pre-eminent dispute resolution centre, Sir Geoffrey expressed confidence that the capital could retain this status if it adapted to the impact of technology.
“GPT-4 and other advanced machine learning is likely to transform the work that lawyers need to do and possibly even, in the slightly longer term, the business of judging,” he said.
He noted how the previous iteration of ChatGPT scored in the bottom 10% when it took the US Bar exam, but the most recent version was in the top 10%.
“This demonstrates the speed at which generative AI is developing. It perhaps makes the point that there is a real possibility that AI may become more intelligent and capable than humans. It is obvious that these advances will affect the legal world as much as any other part of our society.”
Sir Geoffrey has been forecasting the impact of blockchain technology on the law for some time.
Although its adoption has not yet gone at the speed that had been predicted – trading methods were not changing quickly, he observed – the speed of change in relation to AI was different. AI was “already used by all of us all the time”.
He cautioned: “Businesses will ultimately not want to pay for things that are available free. And that applies as much to legal services and dispute resolution as to anything else. This eternal economic reality will be a significant challenge for dispute resolution.
“Just as we now see far fewer lawyers earning air miles than we did before Covid, legal services will need to add value, and dispute resolution will need to make as much use of generative AI as is consistent with user confidence.
“Digitisation will speed up transacting in all sectors, and faster transacting will… lead to a requirement for faster dispute resolution.”
This was why the UK Jurisdiction Taskforce Sir Geoffrey chairs issued its streamlined digital dispute resolution rules two years ago.
He suggested that judicial decisions could in time be taken by machines. “I think that AI will be used within digital justice systems and may, at some stage, be used to take some – at first, very minor – decisions.
“The controls that will be required are (a) for the parties to know what decisions are taken by judges and what by machines, and (b) for there always to be the option of an appeal to a human judge.”
The “limiting feature” was likely to be the need for users to have confidence in the judicial system. “There are some decisions – like, for example, intensely personal decisions relating to the welfare of children – that humans are unlikely ever to accept being decided by machines.”
Litigation and arbitration procedures needed to change “a lot” for London to remain a leading dispute resolution hub, Sir Geoffrey added.
He has been championing the development of online dispute resolution for small civil, family and tribunal cases and said it would have to extend, in time, to commercial litigation and arbitration.
“Cases in arbitration and in the Business and Property Courts are not taking any less time to resolve now than they did 20 years ago.
“That is despite active case management, restrictions on disclosure and on the length of witness statements and other streamlining techniques… We need to re-think the process for the digital age.”
Pleadings, experts report and witness statements – “analogue concepts rooted in an analogue age” – all needed to be “considered afresh”, the Master of the Rolls said.
He continued: “The central element of any dispute resolution process is to identify the issue or issues that divide the parties. That issue, even in a complex case, can often, once identified, be simple. The difficulty is getting to it quickly and early enough to avoid massive cost.
“It is here that generative AI may be able to help. It may be that the power of AI could identify, from a mass of complex facts and transactions, the real issues that divide the parties and that require resolution.
“If that could be done, the actual resolution process itself could become shorter and less costly, particularly if on-chain recording meant that the scope for factual disputtoes was much reduced.”