Debunking the AI myths in the legal sector
Posted: 25 January 2017 | By Darcie Thompson-Fields
For those in the artificial intelligence community, AI has been around a long time. Originally dating back to the 1950s, AI research has been a fixed part of the landscape since the 1980s becoming a multibillion dollar industry long before the millennium. But for the wider public, AI has only received real attention in the last year or two. During that time, AI has made big news. Google images for AI show a range of anthropomorphized images typically illustrating scare stories of robots supplanting humans in a whole variety of jobs with a clear message echoed recently by Mark Carney: you will soon be redundant.
This dystopian vision is especially common among legal journalists. In commercial law firms of the future, so the story goes, AI systems will be putting young lawyers on the scrapheap as intelligent machines take over all their work. Bad news sells. Across a range of legal media, multiple commentators have suggested that the number of lawyers would be decimated as AI takes over in the next twenty years. It’s not a new story, of course. The distinguished crystal ball gazer, Professor Richard Susskind has been forecasting the end of lawyers since the 1980s, albeit that his original algorithmic rule-based programming approach was long since dismissed by leading players in the AI community.
New approaches to AI are now underway, and tomorrow’s winners will be the tech companies at the forefront of machine learning, or what the IBM marketing department likes to call cognitive computing. The IBM AI, Deep Blue, defeated world Chess champion Garry Kasparov back in 1997, but perhaps the most celebrated is IBM’s Watson, the best-known global supercomputer since it won the US game show Jeopardy in 2011, while just last year the British AI company DeepMind, acquired by Google in 2014, overcame Lee Sedol, one of the world’s leading GO players. The AI community is now attempting to break the poker challenge, a harder task for machine learning algorithms as the best human players win by obfuscating their betting patterns.
Generic artificial intelligence doesn’t exist
At the heart of all modern AI systems you will find neural networks and machine learning code, powered by arrays of parallel processors. All these AI systems are able to rapidly analyse vast datasets, to identify patterns and to develop conclusions, all infinitely more quickly than humans could ever achieve. But unlike the human brain, there exists no generic artificial intelligence, and instead each system has to be tailored to a specific application. Machines beating humans at anything certainly piques public interest, however narrow the application.
More significantly, Watson has begun to make a practical contribution to the real world, including diagnostic applications by doctors. In 2016, after puzzling doctors at Tokyo University, it correctly diagnosed a rare form of leukaemia when a patient’s genetic information was plugged into a medical version of the Watson program for answers. After sorting 20 million cancer research papers, Watson produced the proper diagnosis within 10 minutes, suggesting a new treatment that has since been more effective.
In the legal sector, English courts are also starting to acknowledge AI. Last year saw the first court approval of document review using predictive coding: an AI system was programmed to evaluate volumes of big data to identify and classify documents. Law firms too are now using AI applications, often based on Watson, as part of their legal research functions, enabling lawyers to provide faster and more comprehensive analysis of contracts and precedents.
So should young people today considering a law career abandon their plans? When assessing how AI already benefits the legal sector and what it has the potential to achieve, one factor is important to remember: in all its applications AI will remain a tool. Very useful in all sorts of ways, but a tool nonetheless. Lawyers will not be replaced by intelligent machines. Instead, AI will allow them to spend more time on other work and with their clients.
The evidence to back up this statement rests in what the UK’s biggest law firms are already doing. Commonly referred to as the magic circle, the top five London-based law firms are among the largest in the world, employing more than 10,000 lawyers between them across an international network of offices. The first of them to sign a major deal with an AI service provider, UK-based RAVN, was Linklaters, which happened early in 2016.
They were soon followed by Clifford Chance and Freshfields Bruckhaus Deringer, both of which formed separate partnerships with Canadian AI software provider, Kira Systems. Slaughter and May chose Cambridge-based Luminance Technologies to test its AI technology, designed to streamline the due diligence process. Meanwhile, Allen & Overy created MarginMatrix, a bespoke service to help banks deal with complex regulations. This system automates document drafting reducing the time taken from three hours to three minutes.
The magic number
So have these technological advances reduced their lawyer headcount? The simple answer is no. Overall, the number of magic circle lawyers employed has increased slightly over the last year. Perhaps the best indicator of the impact of AI in law lies in the recruitment figures of the junior tiers that are predicted to vanish: the number of trainees that each firm hires. This too has not decreased. In fact, compared with hiring numbers over the last decade, trainee numbers have increased.
Although they are fully committed to exploring the benefits of AI systems and incorporating them into their client offering, magic circle firms are not then cutting their headcount. A broader picture across the UK commercial legal sector shows that, according to the latest published figures, overall training contract numbers increased by 9.5% in 2016 from 5,000 to 5,450. So much for the hyperbole of the scaremongers.
Lawyers should therefore be inspired by what AI research can achieve in saving them time, and not be scared about losing their jobs. As they become increasingly integral to our work patterns, new AI systems will further assist them to enhance their practice in myriad ways. New machine learning techniques will offer a wealth of exciting possibilities.
Just like the doctors, lawyers will increasingly use AI as a diagnostic tool, increasing their efficiency and accuracy. But there are limits. While the current approaches will allow AI systems to display dazzling subject knowledge, they will never achieve wisdom, which is an extrapolative, non-deterministic and non-probabilistic process unique to humans.
Our AI machines will appear increasingly clever and capable, but they can never become human. We are not just a series of machine learning algorithms; the human characteristics of empathy, moral discernment and spontaneous creativity, all essential qualities required of a talented professional advisor, will not emerge from the current approaches to AI. For all the hype surrounding it, AI will never be more than a remarkable tool for the lawyer, and whatever the headline writers might say, it will never replace them.
Robert Morley is chief operating officer at Excello Law