16 Sep 2016

The next big thing in legal: carthorse to racehorse artificial intelligence

Chrissie Lightfoot takes a detailed look at how AI is used by law firms and what is on the horizon.

By Chrissie Lightfoot

As a futurist, an entrepreneur, and a lawyer, I always get asked, ‘What do you think is the next big thing in the legal world?’

I always begin my response with a catch-all reply: ‘The next big thing is anything that helps you attract and keep a client. No client equals no business. A bit of a cliché, I know. But you must constantly rethink how to do things better and be more efficient by using the latest research, thinking, and innovations.’ In 2008 I commented that the next ‘big thing’ would be the adoption of social media and social networking, providing an innovative, cost-effective, smart way to market, brand, and sell using the latest in digital technology.

Facebook’s pioneering founder Mark Zuckerberg in 2010 remarked that ‘There’s going to be an opportunity over the next five years or so to pick any industry and rethink it.’ True to form, the legal world caught on to this latest wave of ‘rethinking’– but it took far longer than other professions and industries. It has taken more than eight years for social media to be widely adopted and truly embraced in the mainstream legal ecosystem. Fast-forward three years. Back in 2011 I would answer with my catch-all reply and add smart technologies, such as relationship intelligent products; cognitive computing (CC); artificial intelligence (AI); and robotic automation technologies. Leap forward another three years to 2014 and, in a nutshell, I believe the next big thing for legal is adopting a variety of fee-earning and business processes that encompass AI technologies, and which blend human talent and machine intelligence, delivering all kinds, and enabling new kinds, of legal services. Today, in 2016, smart CC and AI technologies exist that are available to pervade every aspect of fee-earning and business processes within a law firm; including marketing, business development and customer relationship management. 

Suspicions and reinventions

Not surprisingly, it has been a slow burn for these pioneering smart technologies to ‘break into’ the legal world. It still remains a huge challenge for the vast majority due to the same suspicions and cautious approach that lawyers and law firms hold towards anything that may promulgate change or cause major disruption – or they simply do not understand. Not unlike how social media and social networking was viewed with scepticism back in 2008, so too emerging intelligence – augmented and artificial – at present.

However, due to the evolution of AI and its potential impact on the legal ecosystem, Professor Richard Susskind recently warned that the legal profession has five years to reinvent itself, and Riverview Law CEO Karl Chapman has predicted that the use of AI will be mainstream in the legal profession by 2020. I too have long held the belief that the profession needs to reinvent itself, and in 2011 declared that the impact of AI and robots would be felt by 2016 or thereabouts, but certainly before 2020, and definitely not decades away, as Tony Williams predicted in the Jomati Report ‘Civilisation 2030’.

I held and hold steadfast in this view, despite ridicule and scoffing, because we, the human race, constantly surprise ourselves with our technological breakthroughs. For example, when scientists were given the green light to crack the DNA code and sequence the entire human genome, everyone (scientists, futurists, thought-leaders, etc.) predicted it would take 100 years: it only took eight (1995 to 2003). Fast-forward 13 years to 2016. On 10 May this year Harvard University hosted a closed-door conference (in absolute secrecy – no journalists allowed) with more than 100 scientists, lawyers, and business leaders to discuss the creation of a synthetic human genome and its feasibility.

The Human Genome Project (2)’s (HGP2) aims to synthesise the full human genome from synthetic materials. I have no doubt that this project will go ahead (once all the regulatory and patent squabbles have been agreed, not unlike the 1990s) and in all likelihood it will take far less than ten years if the previous track record and history in genome achievement is to be taken into account; imagine the technology in the 1990s and compare it to what we have now, factor in the exponential development of technology, and what do we have?

From AlphaGo to AlphaLaw?

Drawing a parallel with exponential advancement in AI evolution and computing intelligence breakthroughs, and factoring in what could be deployed in the legal world, in March this year we witnessed Google’s DeepMind AlphaGo’s stunning victory over Go legend and world champion Lee Se-dol.

DeepMind (an AI system of neural networks) was acquired by Google in 2014 and uses games as a testing ground for AI algorithms that could have real-world applications. AlphaGo was imbued with the ability to learn through the human-like process of practice and study. It is not a hand-crafted program in which software engineers distil information from, for example, a human player’s head into specific rules and heuristics. AlphaGo’s founder Demis Hassabis comments that an aspect of ‘intuition’ has been introduced into the neural networks, a characteristic which distinguishes top Go players.

The same could be said about top lawyers, lawyering, and legal service provision. In a handful of years – from the introduction of chess-playing computer program Deep Blue in 1996 to AlphaGo’s self-learning in 2016 – we have reached a point where a machine proved its superiority over a world-class human being’s creativity and intuitive insights.

Despite the fact that Lee Se-dol has since beaten AlphaGo, this does not detract from the significant step forward made in machine-learning capabilities. After all, even Hassabis was surprised at the rate of progress he and his team achieved in such a short timeframe – 18 months, from inception to triumph. I cannot imagine an 18-month-old baby being able to study and learn at the same rate as AlphaGo, can you?  There is, therefore, a possibility, in theory at least, that a machine could learn through practice and study to become a world champion lawyer. And this is deeply significant. 

But until such time as an AlphaLawchampion lawyer exists (which may be sooner than we all anticipate), let us deal with the reality in a legal world where:

  • Smart CC, AI, and robot technologies in relation to the four key elements of legal service provision – commoditisation, research, reasoning and judgement – exist and can support or replace many aspects of a lawyer’s job;
  • Both fee-earning processes and business processes can be supported and/or replaced by expert systems, CC, robotic-automated systems, AI, and machine learning;
  • Clients demand and expect more speedy, accurate, expert, creative, intuitive, and accessible legal advice;
  • Pressure is growing both from within law firms (or in-house teams) and from clients to respond to the demand for customer-designed services;
  • Technology-related projects that are both user and client-centric need to be implemented successfully; and
  • In the long term, everyone will be using these kinds of technologies, despite the fact that AI deployment in the legal world has a last mile problem, comments Rick Seabrook, managing director of Neota Logic, Europe.

Four aspects of legal services

As we witness the evolution of AI and its take-up in the world at large, I predict we will experience a more measured approach and gradual warming to the evolution of AI in the legal world throughout 2016, which will then lead to a tsunami deployment within a couple of years; similar to the social media warming but on synthetic, exponential steroids. There will be AlphaLaw pioneers, just like there were social media pioneers.

Due to the deployment of smart tech, CC, AI and robots in the legal ecosystem where lawyers, firms, general counsel and clients are beginning to embrace such technologies, we’re seeing that lawyering is increasingly becoming more productive, efficient, accurate, better quality, less labour intensive and time intensive and the role of the lawyer is gradually shifting and changing.

There are, at present, several key providers of smart CC, AI, and robot technologies that relate to all aspects of legal services provision, and they deal with different aspects of the lawyer’s role. If we break down a lawyer’s tasks in a legal project from beginning to end, we find there is a technology that can handle the majority of these four tasks far more quickly and accurately than a human lawyer.

CC, AI, and robots are currently (and mainly) used and being considered for use in the legal ecosystem as a carthorse; for example, e-discovery, research, business processes, basic analytics or analysis, hypothesising, etc.; where they speed up much of the mundane elements of legal work. For example:  

  • Big Law American firm BakerHostetler is the first licensee of Ross, hailed as the world’s first AI lawyer and a ‘very smart artificial co-worker’, which interacts using natural language and is used for bankruptcy-related legal research matters. Ross (owned by Ross Intelligence Inc.), an AI product and service resulting from the use of Apple’s Siri slick voice capabilities and IBM Watson’scognitive computing prowess, has been in the development and pilot stage throughout 2015 but will now be a huge boon for lawyers, law firms, and clients.
  • BLP has been using RAVN ACE(applied cognitive engine) AI technology in the UK since 2015 in its real estate department and commercial practices. Used for ‘deep research’ and processing, such as extracting specific pieces of information from large documents, the technology has been nicknamed Lonald by BLP associates, who welcome the machine as being more efficient, productive, and accurate than they could possibly be. Lonald is due to be ‘rolled-out’ in all BLP departments over the next three years in an attempt to boost efficiency and improve the morale of its lawyers.
  • The first Magic Circle firm to go public with the use of AI, Linklaters has also signed a deal with technology provider RAVN systems, the details of which remain confidential. The Financial Times also reveals the company has developed Verifi, a computer program that can sift through 14 European and UK regulatory registers to check client names for banks. The company said it could ‘process thousands of names overnight.’
  • Hot on Linklaters heels is Clifford Chance, the latest UK Magic Circle law firm to adopt the use of AI after announcing that it has entered into a partnership with software provider Kira Systems.
  • DLA Piper in mid-June announced that it has also partnered with Kira Systems to implement a global artificial intelligence tool for document review in the M&A due diligence process.
  • Riverview Law launched KIM– an AI virtual assistant in January this year –followed up in April with an AI triage of managed services technology and an AI platform; a huge advantage for general counsel and law departments in corporations, and for clients of all company sizes.
  • Pinsent Masons has developed a programme that reads and analyses clauses in loan agreements. Its TermFrame systemhelps guide its lawyers through transactions and points them towards the correct precedents at each stage of a process.
  • Hodge, Jones & Allen has been working with academics from University College London to create software that assesses the merits of personal injury cases since 2014/15. 

All of these AI systems and computer programs handle large quantities of structured and unstructured data, and assist with the process, management, research, and reasoning elements related to legal issues. Two years ago they were not even in play in the legal market, though they were on the horizon.  

Time equals money

For entrepreneurial lawyers and law firms, CC, AI, and robots also provide the opportunity to rethink legal service offerings. For example, with RAVN’s AI technology, a law firm that has a client with tens of thousands of employment contracts across the world could offer the existing client an ‘employment contract analysis’ in relation to risk and compliance issues (for example, when there is new regulation or when the law bites). RAVN can configure a robot that will tell the law firm or client what their exposure is.

Providing this kind of legal service would have been impossible previously due to the disproportionate spend in time and cost; however, RAVN takes a routine, cognitive process (traditionally undertaken by junior lawyers labouring over a long period of time, usually months) and transforms it into an automated process, which can be completed in a matter of minutes or hours.  Consequently, something that is incredibly difficult for a human to do that a robot finds very easy could be offered as a valuable new legal service. RAVN is currently in the process of developing a whitepaper on the subject of the potential of deploying AI technologies in the legal world, which should prove highly informative and helpful to curious minds.

This has been, and is, our current and main level of understanding, thinking, application, and deployment of smart tech, CC, and AI in the legal industry. And yet, a search for ‘AI in law’produces 105,000 results in just the news section alone of Google’s vast index. From the web as a whole, 25,900,000 results can be found, meaning there is appetite, curiosity, and interest in searching for how AI and the robots may be of use in the legal ecosystem more widely.

These technologies could, to some extent, be used in every aspect of legal service provision, not just the mundane ‘everyday’ basic legal matters of ‘doing’ but also the more lucrative area of ‘thinking’ and ‘advisory’. 

Racehorse thinking

Law firms have, since the start of this year, only just begun to think of using artificial intelligence (AI) technologies as a racehorse, and not a carthorse, and have started to implement in their racehorse thinking and advisory elements unique to cognitive thinking and a human mind. Global players Taylor Wessing and Norton Rose Fulbright have each independently collaborated with software and AI developer Neota Logic in creating apps, named TW: navigate PSC App and ContractorCheckrespectively.

Taylor Wessing’s app, introduced in April of this year, acts as a guide and adviser with respect to the recent changes to the Companies Act, where there is now a requirement for companies to identify and name persons of significant control in their returns to Companies House. Norton Rose Fulbright describes its app, introduced in late 2015/early 2016, as ‘an innovative tool designed to help organisations accurately characterise their people to manage risk, so that contracts truly reflect an individual’s contractor or employee status. This smart identification tool has been tailored to apply the key characterisation indicators across multiple jurisdictions.’

Both firms have rethought how they deliver specific aspects of lawyering and are beginning to focus on racehorse elements of lawyering; which are the high-end, intellectual capital, reasoning, and judgement elements usually locked away in a lawyer’s mind. Neota Logic works in collaboration with law firms, and elicits knowledge and experience from a lawyer to then ‘brain dump’ that knowledge and experience into software. Algorithms are then created which respond to the client or prospect without the need for a person. In theory it means a lawyer or law firm could serve hundreds or even thousands of clients, prospects, or queries seamlessly while they sleep – and get paid for it. 

Each bespoke AI app mimics and replaces what a human lawyer does in relation to the process, reasoning and judgement element (the knowledge and experience aspect) of a lawyer’s specific area of law. For example, a lawyer will ask a series of questions to a client who will reply with their answers: throughout, the lawyer pulls on their knowledge and experience of the law and provides the client with the answers they seek. This kind of AI technology is beginning to prove popular for: 

  • Online compliance;
  • Online legal assessment; and
  • Providing online advice solutions.

Using this kind of technology is also proving to be a smart way of marketing, creating, and building relationships with firms’ new and existing clients.

Inviolable lawyers

Currently, there are automated, cognitive computing (CC), and AI systems that support lawyers and which replace lawyers. Nevertheless, as machines become more intelligent and lawyers become more comfortable in deploying and working with these systems, it is inevitable that the machine will make a switch from carthorse to racehorse functionality, and will replace more and more of a human lawyer’s key roles in relation to the four elements of legal services provision (commoditisation, research, reasoning, and judgment), edging into the realm lawyers think they are immune and untouchable. 

But they are neither immune nor untouchable. All areas of law, including corporate, commercial, intellectual property, employment, property, tax, bankruptcy, family, litigation, etc., will not escape the march of the machine. Premonition’s pioneering technology is testament to this in the litigious domain. Within the legal ecosystem, we will see AI evolve from CC (which demands a significant amount of data) and mimicking the human brain to iterative intelligence and design algorithms (rules-based programing and self-learning where the AI or robot learns for itself); then, eventually, artificial general intelligence will come about via the Google Brain, White House BRAIN Initiative (Brain Research through Advancing Innovative Neurotechnologies), Obama brain, or DARPA AI projects – possibly in around 15 to 20 years’ time.

It is not insurmountable that we will witness a world champion AI lawyer, not unlike a world champion AI player in Go, the Chinese strategy board game, within the next few years. The rate of technological development and AI innovation is exponential. Accordingly, the likelihood is that many aspects (but not all) of lawyering as we know it now can and will be done with little human intervention within five to ten years. This prediction is in line with the global research that, within ten years, 57 per cent of blue-collar and white-collar work, globally, will be replaced by automation, AI, or robots.

So, in 2016, you have two choices: the first is to embrace the rise of smart CC, AI, and robot technologies; management may be responsible for providing the environment to facilitate change and innovation, but the responsibility to embrace change rests with each and every lawyer in our firms. The second is to ignore emerging technologies, do nothing, and get left behind. 

Paper-pusher pushout

Another prediction is that massively overpaid paper-pushers will be sidelined within one to three years. If you understand that smart AI technology is not here to replace you (well, not immediately) but to make you a better lawyer that forces you to use your instinct, emotional intelligence, and capitalise on your mind, then you have made the first step in future-proofing your career and business. The brave new legal eco-system will require ‘algorithmic angels’, in which the primary lawyerly role will be: 

  • To interpret that the AI or robot is correct about the law; and
  • To provide a supportive relationship to clients and general counsels.

There will only be three kinds of human worker or lawyer within the next ten years when the machines rule:

  • Those with high-end emotional intelligence and relationship skills; or
  • Those who provide support to the machine systems; or
  • Unemployed. 

With exception to those firms that have already gone public in relation to the deployment of AI systems (some of which are named earlier), companies and firms already using and/or considering using these technologies in the UK and beyond include a selection of those from: 

  • The Big Four accountancy firms; 
  • Magic Circle firms;
  • The top 50; and
  • Those outside the top 50, i.e. those with distinct specialities that aim to use AI to win new clients and compete with the top 50. These types of private practice include real estate, immigration, pensions, and finance.  

Renovating private practice

Rethinking legal services may also entail rethinking what type of business of law will be most suitable going forward. Law firm consultant George Beatonhas recently released a fascinating piece of research on why ‘big law’ should start to reinvent itself due to the mix of legal service provider predicted to be in the market in ten years’ time. These types of private practice include big law; traditional or remade traditional law; new law; standalone automated legal service provider (for example, Riverview Law); and in-house general counsels. The findings predict that a staggering 50 per cent of big law and traditional firms will be ‘remade’ and will lose significant market share to the other types of businesses within the next decade.

Accordingly, my advice, as ever, is to future-proof you and your business now. It is inevitable that smart tech, AI, and robots do – and will continue – to support us positively in our current roles, but eventually they may replace lawyers entirely in carrying out particular tasks once AI exceeds human lawyerly intelligence; we’re talking a handful of months or years here, not decades. Ask yourself the following:  

  • Can you afford not to be prepared for the time when ‘the machine’ evolves and inevitably does your fee earning and performs business processes quicker and more accurately, more creatively and smarter than you?
  • Can you afford not to embrace the machine, when your existing and new competitors are already using it, or about to?
  • Can you afford not to deploy CC and AI technologies when your existing and new clients are beginning to expect some form of these systems as part of a legal service offering? 

If you ask me today what I believe the next ‘big thing’ in legal is going to be, my reply is as follows. Do not assume that smart technology, AI, and the robots will not be able to do significant elements of your job, even the complex stuff. Prepare for the possibility that they could and most likely will. 

This content is available to subscribers only. To continue reading...

Sign in to your account

Take a one-month free trial

If you aren't a subscriber, please sign up for a one-month free trial to access all Robotics Law Journal content, including:

  • All premium online content
  • Daily newsletters
  • Breaking news alerts

If you require further information, please email subscriptions@roboticslawjournal.com or contact call us on +44 (0) 20 7193 5801.