As artificial intelligence or AI has started to feature more regularly in both mainstream and legallyspecific media outlets (3Kites’ Paul Longhurst was on the panel at a recent AI event), we thought it would be useful to provide a short paper outlining what it means for the legal sector. This paper is by no means definitive, but it should provide you with sufficient context to be able to sort the myths and the misconceptions that surround AI from the practical applications. If you would like further information on why this area is gaining traction in the sector and how 3Kites Consulting can help your firm, please contact us using the following details:
firstname.lastname@example.org +44 (0)7785 254909 www.3kites.com
What is AI, and why should you care about it ?
The term Artificial intelligence (AI) was first used in the 1950s by John McCarthy at MIT to describe the area of computer science that develops systems which are capable of exhibiting intelligent or human behaviours. Today it is often associated with robotics and self-driving cars, although you are more likely to encounter it as systems that interact with us through natural speech.
The realisation of AI as a practical tool which businesses can use is largely down to the increase in raw computing power which makes possible the number crunching necessary to complete huge numbers of near simultaneous decisions which, in turn, give the appearance of intelligent thought. This computing power has been combined with sophisticated cameras, microphones, speech synthesizers and other such devices to improve the ease of system operation. A great example of this is the speech recognition available on many mobile phones such as Apple Siri and Microsoft Cortana. These tools help us to interact with our systems without the need for keyboards and mice, which is akin to the way we communicate with each other, ergo displaying human behaviours.
There is a deeper, philosophical debate about whether or not computers are or will be able truly to think for themselves, learning and making real judgements rather than just applying rules to known situations. Google and other large organisations are investing heavily in developing systems that can make connections and interpretations from raw information in order to make predictions and take decisions. Google’s driverless car works by having crunched the huge amount of information gained by watching the road and judging how quickly things are moving, in what direction and what is likely to happen in order to make judgements about whether it is safe to pull out, change lanes, speed up, slow down, etc.
In the legal sector, lawyers apply their knowledge and experience to a situation in order to make judgements about what to advise their client to do, or at least to assess the risks of different courses of action. As with the driverless car there is a need to factor in other humans’ reactions to your course of action. In the legal context there is also the need to make your client feel comfortable that they are getting the right advice, to factor in their (rational or irrational) concerns, and allow them to ask further questions. For the time being the client interface is likely to require human input much of the time. However, advisors can be better equipped to assess risks and probabilities and to predict outcomes and costs if they have access to the computing power to analyse large volumes of unstructured data in an intelligent way. Imagine a doctor having access to a computer system which has up to date information about all the symptoms experienced by millions of patients and their diagnoses and treatment. That doctor is likely to be more effective than one who relies only on their personal experience of a relatively small number of patients, plus their reading of textbooks and journals.
Why you should care about AI is because of its potential to help lawyers deliver their services using less lawyer-time and help with decisions based on the analysis of large amounts of data and the experience of often repeated work processes. And if your firm isn’t doing this but your competitors or your clients are, then you might expect to come under pressure to reduce costs and/or timescales without the tools available to those who have chosen the AI path. However, AI should not be viewed as a panacea which will guarantee increased efficiency, but rather as having the potential to reward firms which are prepared to invest in ways of improving the efficiency of their lawyers.
How is it likely to impact the legal sector?
The application of AI should be considered in the wider context of the general business community to understand how it will specifically affect our sector. As with the rise of intelligent mobile phones and tablet devices, legal has often been a follower in terms of its adoption of new technologies and we would expect this to be the case with AI. A good example here is the recent release of the Apple watch device – Apple would not have been targeting the legal sector with this, but lawyers and others will be aware of them and start buying them. One time recording vendor has already released an Apple watch app which can track time into and out of locations to provide more accurate figures for time spent in client meetings.
However, the ability to analyse large quantities of structured and unstructured data is the area that law firms have been focusing on to date, using tools such as RAVN and KM Standards (formerly Kiiac). Tools for doing this in the context of large volumes of disclosure data have been available for some time and are gradually being adopted.
Another area where this could have an impact is in insurance work where the large volume of data provides firms with the option to find patterns that can assist with decision making. Where patterns can be found for cases that the firm always pays out on or indeed loses, or conversely where the firm always defends successfully, these could be picked up programmatically to save time and most especially cost on low margin work. A small percentage of incorrect decisions would be more than outweighed by the overall savings for clients and increased profitability for law firms. Any cases falling in the middle would simply continue to be assessed by experienced lawyers. Neither scenario is AI in the true sense, but it represents a step forward towards using technology to undertake tasks which the human brain alone would struggle to do. It enables law firms to scale up without adding headcount 1.
True AI is nevertheless on its way. In August, Dentons announced that “NextLaw Labs, its collaborative innovation platform, has ... signed a deal with its first portfolio company, ROSS Intelligence Inc., a startup developing a legal advisor app powered by IBM Watson”. ROSS is a legal research assistant developed by the University of Toronto to sift through huge quantities of cases, legislation and other documentation in order to answer natural language research questions. ROSS learns from its experience (in a similar way to Amazon which learns about your preferences and those of others in order to offer you relevant products) and its answers are refined over time, much like a trainee….
What will help or hinder its adoption?
Surveys already indicate that law firms admit they do not use their existing technology to its full potential 2. This is for a number of reasons including perhaps the underlying belief that a profession which has managed very well on the basis of a long established way of working will continue to do so into the future. This confidence may be leading to a lack of investment in technology talent (or a refusal to take advice from such talent) and, if Professor Richard Susskind’s latest book 3 is to be believed, may be misplaced.
Risk taking and failure are generally thought of as anathema to lawyers, but some firms are investing in their futures by setting up innovation funds (with no guarantees of returns) which allow them to experiment with radically different approaches. This commitment to innovation via a ring-fenced budget sidesteps the perennial law firm issue of setting money aside for projects which do not have a proven return on investment. Some of these innovation funds will generate successful projects, at which point the rest of the market will need to worry.
Paul Longhurst, 3Kites Consulting Limited, November 2015
3Kites Consulting is a limited company registered in England and Wales. Registered number: 5644909. Registered office: Chancery House, 30 St Johns Road, Woking, Surrey, GU21 7SA. www.3kites.com
1 A similar approach has been adopted in analysing large volumes of US IP cases in order to forecast the outcome of cases and to anticipate a particular Judge’s likely reaction to certain approaches, see Lex Machina.com.
2 See for example a survey of CIOs in European legal firms by RSG Consulting.
3 The Future of the Professions Richard and Daniel Susskind, 2015
© Copyright 3Kites Consulting Limited 2015