If as Philip Larkin wrote, sex began in 1963, then it can be asserted that artificial intelligence (AI) began in November 2022 with the release of ChatGPT for general use on the internet. Although dismissed by the unknowing or uncaring as just another fad, I would suggest AI stands poised to become a technology, that is potentially transformative of our society. It is likely to become transformative of legal practice, including those areas of work undertaken by the Bar and particularly the way which barristers charge for their work.
Barristers undertake advisory, drafting and advocacy work for clients. A barrister can charge a professional or lay client for the services they agree to provide, on any basis agreed with that client. The agreement can be on a privately paid basis or some form of contingency fee arrangement, such as a conditional fee agreement or damages-based agreement. The amount of the fees payable may be agreed by way of reference to an hourly rate, or by reference to a fixed lump sum fee. If a conditional fee agreement is made, the barrister can charge a success fee, if the client will agree to pay it.
Many barristers will charge for advisory or drafting work on the basis of an hourly rate, and a lump sum brief fee for advocacy work. But they need not. Some clients prefer to agree a fixed lump sum for any element of work a barrister does, and some prefer to pay hourly rates all work undertaken by a barrister.
Irrespective of the charging structure adopted, the key element reflected in the price is always that of time: how long will it take to perform a particular piece of work and how is that to be priced, whether in hours or by a daily rate? And it is because of the factor of time, if the use of AI tools can facilitate a barrister doing a piece of work far more quickly, than would hitherto have been the case, it will not only be possible but necessary to charge a client on the basis of the value of the work that is done, rather than the time that it took to undertake that work, as by using AI tools, time spent on many traditional activities will shrink dramatically.
A few examples will suffice. Take advisory work on disclosure, or evidence, or latterly trial preparation, where a huge amount of material may need to be considered, analysed and where relevant utilised, and where irrelevant discarded. Reading lever arch bundles would have been the staple activity underpinning these tasks. But AI is increasingly deployed in document review and E-disclosure: AI, particularly machine learning and natural language processing can sort review and analyse large volumes of documents to identify relevant information for the purposes of giving advice far faster than any human can do so.
Another area of activity is legal research. All lawyers are now familiar with electronic databases and digital research. Yet it was not so very long ago, that legal research in a paper based legal system, required half a day in a law library, with a copy of “Is it in force?” AI systems should be a step beyond the experience of many users of electronic legal databases, as the interface should be more sophisticated, easier to direct and through iterative enquiries save large amounts of time.
Perhaps the most interesting use is the role that AI can have in content creation. Many documents drafted by barristers, such as advices or statements of case, are often heavily based on precedents, templates or standard forms. AI can be used now, through voice prompts to create legal opinions or memoranda, far faster than any conventional dictation software can work, and far faster than the typical barrister typist, even one familiar with “cut and paste”. The amount of time paperwork takes, will significantly decrease.
On a traditional economic analysis this could be described as a shift from labour to capital, and analogies can be drawn going back to the first Industrial Revolution and the adopting of the spinning jenny. But if an advice on evidence or a set of particulars of claim can be done, through simple voice commands, effectively even removing the need to make “a click of a button”, then barrister to capture the added value they are providing, will have to charge not by time spent, but on a “value basis”. What clients will pay for is the analysis, exercise of judgment, and “answer” rather than the time spent getting to the “answer”.
I predict that a shift to exclusively value-based fees (fixed fees) will be undertaken. As AI becomes more widespread and facilitates more tasks, reshaping workflows, I would suggest there will be an ever-increasing drive to value-based fees (or fixed fees as they are otherwise known) and a marked decrease in the use of the hourly rate: if a 20 page opinion is required but can be generated in 30 minutes, rather than the 8 hours it might formerly have taken, the business imperative will be to shift the basis for charging accordingly.
AI will accelerate and bring to its logical conclusion in the provision of legal advice and drafting services, the concept of “near zero marginal cost” a phenomenon of the digital age, where it costs virtually nothing to provide additional products at minimal, or even zero cost.
AI will also have an effect on the way that costs are claimed and recovered between the parties. It should be noted that although notionally a barrister and his client decide how to structure a barrister’s remuneration, this dialogue is inevitably shaped by questions of recoverability between the parties, as a receiving party to an award of costs, will wish to maximise the recovery of their own legal spend, if they are successful, in getting an award of costs against their opponent.
Increasing automation will make a nonsense of the notion that time claimed in fee notes is a reliable indicator of the reasonableness of the fees claimed: a large claim for time may in years to come simply be an indicator that the counsel concerned is a “dinosaur” who does not invest in the technology to run their business with the benefit of AI grounded efficiency savings.
The court’s approach to the assessment of contentious costs is governed by rule 44.4 CPR containing the seven (now eight) pillars of wisdom. Of these factors only one, factor (f) specifically enjoins the court to have regard to the amount of time spent on the case.
If time can be saved by increasing the power of the programs and platforms used for legal research, or standard forms, skeletons or pleadings can be produced more quickly through AI that will benefit both barristers and their clients. It will in fact mean a greater transparency on fees negotiated on a value basis: it will also mean that barristers and their clerks will need to think more deeply about questions of value and how they discern it, and how they present a case on value when negotiating a barrister’s remuneration with clients, for the old world is slipping away.
A version of this article first appeared in “Counsel” magazine.