Advances in technology mean the end of the legal profession is nigh. Eventually. Maybe.
For 20 years, the academic Prof Richard Susskind has been predicting the end of legal practice as we know it, and warning lawyers to wake up to the reality of advancements in IT.
Susskind’s 1996 book The Future of Law argued that IT is reshaping the provision of legal services: if lawyers are to survive, they must concentrate on the more specialist tasks that cannot be performed by others with the assistance of technology. Since he made that prediction, the number of solicitors has nearly doubled from 87,000 to over 160,000, which suggests he was rather off the mark.
However, his most recent book, The Future of the Professions, returns to this theme. In it, he argues that we are in a transition phase from a print-based to a ‘technology-based internet society’. Although traditional lawyers are still needed to interpret information, this will change in the near future as new technologies will carry out this work.
Some technology-driven innovation has been happening in legal services over recent years, but the economic crisis and cuts in state spending on legal services, is acting as a catalyst to increase the rate of change. The Royal Courts of Justice Advice Bureau’s CourtNav is an example of a combination of information and a form-assembly program to assist members of the public. It is also useful for practitioners, as it eliminates keying errors on court forms. CourtNav was developed as part of the response to the increase in litigants in person in family cases, caused by the withdrawal of legal aid.
The Dutch Rechtwijzer 2.0 family law internet service is the most commonly cited example of the likely future of online legal services. The system is based on the same program that deals with 60m disputes a year on eBay. Rechtwijzer fits with the zeitgeist of austerity, as it appears to offer a cheaper alternative to a full lawyer service by unbundling cases. It does have high setup costs and, like all such programs, needs constant updating. When it is fully launched, it will be interesting to see if it can recover its costs, thought to be around €2m, from private paying clients.
Since Prof Susskind began predicting the end of the legal profession as we know it, the number of practising solicitors has nearly doubled.
On a somewhat more modest budget, LAG is developing a housing law advice program using similar programming techniques to Rechtwijzer. We believe that we will eventually produce a useful product to support lawyers and advisers, but we recognise it will have its limitations. The nuances of law do not lend themselves easily to binary code. Legal advice software is largely dependent on the experts behind it designing decision trees to take the user along a path to the correct answer. Sometimes an answer needs to be qualified with an explanation – and advice to seek specialist help.
However, Susskind and others argue that advancements in technology are such that we are nearly at the stage where lawyers and advisers will be replaced by artificial intelligence. The IBM AI system, known as Watson, can scan massive amounts of written material and is used to support cancer diagnosis and to recommend treatment plans. Susskind believes it will soon be possible for precedents and different combinations of facts to be fed into such a program to enable advice to be given in individual cases.
Law is not science, though. LAG can envisage situations in which the government and others might not be so willing to disclose their data. The legal press, for example, recently carried the story of the volunteer law students at Bristol Law Centre who enjoyed a 95 per cent success rate in ‘fit to work’ welfare benefits tribunal cases. Presumably, a computer program drawing on this data would always find in favour of claimants? Similarly, whiplash claims in road traffic accidents could be automated in a computer program, but insurance companies might not want to disclose data, if it were perceived to be in favour of claimants’ cases.
If AI programs are to develop, the process of accurate data collection is crucial. Due to costs and institutional reluctance, commercial organisations like insurance companies and government departments might be reluctant to give full disclosure of facts and other information, making data about large numbers of cases difficult to obtain.
Programs would also need to be regulated, as liability would have to be established for when things went wrong.
Assuming these issues are addressed, LAG can envisage more widespread computerised AI advice. Susskind and others may be right that change will eventually come but, like the religious prophet predicting the end of the world, it’s knowing when this will happen that is the difficult bit. ■