Last month one the largest American law firms, BakerHostetler, announced it had 'hired' a robot lawyer to assist in bankruptcy cases. The robot in question is named ROSS and is marketed as 'the world's first artificially intelligent attorney'. While ROSS may be the first, it will not be the last. AI is changing the way lawyers think, work, and interact and heralds a paradigm shift in how legal work is done. Legal professionals and law students should pay attention.
ROSS works by scouring millions of pages of case law and legal documents in seconds to offer answers to questions. Lawyers are able to upvote or downvote portions of text that are surfaced by ROSS based on program's interpretation of the question it is asked. Essentially, ROSS is a tool designed to help lawyers make better informed judgements. If ROSS continues to hold its own, it is quite likely that in the near future firms not using artificially intelligent systems will be seen as outdated, if not negligent.
What does ROSS mean for the future of the legal profession? For starters, there will be a lot less work done by human lawyers. Senior lawyers regularly express concern that most law students have little or no realistic hope of a career despite some amassing debts of up to £60 000. The same is more or less true for solicitors profession, which is often more procedural and bureaucratic in terms of the work that is done, just the type of work that programs like ROSS excel at.
The reality is that this crisis is not being improved by the law schools themselves. Law schools are after all great business for universities. It costs little to educate lawyers, and the influx of foreign students from common law countries such as Canada means that foreign student fees are far too lucrative to turn away. It is easy to conclude that the number of law school places should be limited, or indexed to the availability of jobs, but that ignores the value in studying law for a number of non-legal professions. Another solution might be to move towards the American system where once you complete your law degree and pass a bar exam there is no barrier to practising law. However, that is a recipe for ensuring a glut of bad lawyers, as there is no real substitute for practical experience in the shadow of a senior professional. Chambers and firms cannot simply increase the number of students they take as the legal profession is after all a market economy where less work to go around means fewer workers. So what can be done?
As a number of recent reports suggest, it is not just the legal profession that is threatened by new technologies. Profound technological change, particularly the use of AI and robotics, is set to catalyse sweeping economic changes and alter the nature and meaning of work. The legal profession will survive in some form, but it will not be, nor should it be, immune from change. One thing the legal profession should take away from the advent of systems like ROSS is how it educates future lawyers. For all intents and purposes legal education has not changed dramatically over the centuries. Law students still learn the same core subjects, write the same types of papers and exams, answering the same types of problem questions. As anyone who has gone through it will confess, it can be a bit mechanical. And what is mechanistic can be easily be automated.
If law schools are not going to cap the numbers they admit, at the very least they owe it to their students to educate them for the profession as it is now and as it will be, not as it was in the 19th century. At Oxford and my own institution, Cambridge, it is still compulsory for students to study Roman Law. While it is never easy to break with centuries of tradition, is there any justification for studying Roman Law over a course that covers the use of technology in the legal profession? As AI systems are increasingly adopted in professional practice, it is imperative that law schools adapt alongside them by providing students with a legal education that meets the demands of the modern legal marketplace and trains them with skills that are not yet so easily automated by AI lawyers like ROSS. This may well require a holistic reappraisal of how lawyers are educated such as the integration of the LLB with the professional practice skills of the BPTC/LPC. At the very least, the prospect of programs like ROSS should give aspiring lawyers reason to reflect on the economic realities of the legal profession, and question whether they are paying for an education that will prepare them for it.