Generative AI like ChatGPT could revolutionise the way that lawyers deal with costs, rendering detailed assessment largely unnecessary, a High Court judge suggested last week.
But Mrs Justice Joanna Smith told costs lawyers that their skills would still be needed in this technology enabled world.
Speaking to the Association of Costs Lawyers’ London conference on Friday about the impact of AI on the justice system, she said many of the tasks they undertook – crunching and transposing data, and drafting bills of costs and costs budgets – “are exactly the sort of tasks that could be undertaken using a specialist AI, appropriately trained on rules and precedents in the costs domain”.
She continued: “Indeed the arena of costs management over the course of litigation would appear to be obviously suited to the use of new technology.
“Commentators who know a great deal more about this subject than I do have suggested that costs management systems could be embedded into case management systems, making accurate real time costs management an achievable objective and ensuring that legal teams are alerted to cost overruns almost immediately.
“It has also been suggested that, properly instructed, AI could suggest changes to budgets as they are drafted, informed by the mining of information about cases with similar characteristics.”
This could even render detailed assessment “unnecessary save in respect of discrete points, perhaps requiring evidence or detailed arguments about the retainer”, the judge went on.
“Questions of reasonableness and proportionality might even be capable of being dealt with by an algorithm, at least in simple cases.”
But Dame Joanna argued that, while this would not deprive costs lawyers of work, there would be “a real imperative to find new ways to add value, including, perhaps, learning new IT skills”.
Costs lawyers would still be needed to supervise the AI, she said, while their role in delivering strategic advice to clients would become “increasingly important”.
Costs lawyers “can clearly add significant value to clients in the development of high-level litigation strategy, including by the application of a cost-benefit type analysis at an early stage in the litigation process”.
“Additionally, there is plainly scope for costs lawyers to become involved in negotiations – particularly where costs are a potential obstacle to settlement. This is likely to involve offering effective ADR, including mediation in disputes over costs.”
More broadly, the judge said the “increasing understanding and use of specialist legal AI tools, together with the development of appropriate regulations and safeguards, is likely, in my view, to herald greater efficiencies in litigation and the very substantial reduction in costs, leading, it is to be hoped, to much improved access to justice”.
Dame Joanna said the areas where AI could cut costs were disclosure, organising documents, carrying out research – “although perhaps not, at the moment, complex legal research requiring accuracy and certainty” – inter partes correspondence, producing simple pleadings “and perhaps advice”, and predicting outcomes.
Taking inter partes correspondence and pleadings as an example, she said: “A specially trained AI should have the ability quickly to adopt a firm’s preferred style and to produce chronologies and standard paragraphs for letters and simple pleadings at the touch of a button.
“It can verify information included in such documents and check for inconsistencies with existing evidence, advice or previous correspondence.
“Its ability to predict case outcomes – always assuming it has a sufficiently broad data set on which to draw – may prevent the commencement of unmeritorious cases or at least provide substantial assistance in connection with strategic advice around settlement.”
At the same time, she stressed that AI was “a tool to be used to enhance human reasoning capabilities rather than replacing them”.