Expert Speak Digital Frontiers
Published on Nov 16, 2017
Transforming the Legal Profession: the Impact and Challenges of Artificial Intelligence


In this era of globalisation, technology is playing a key role in the delivery of legal services, the nature of competition, and the demands and expectations of clients. Artificial intelligence (AI) has been driving some of the biggest changes by providing new ways for law firms to conduct their business, for lawyers to advise and communicate with clients, and for legal research to be undertaken.<1> These developments lead to a series of questions. What further impact will AI have on the profession and its constituents? How will the changes affect the process of delivery of legal services and the work of judges and lawyers? Will the disruption result in the creation of new kinds of jobs and, as a corollary, make some existing jobs redundant? How must legal educational institutions adapt their courses to deal with changing market needs? Will AI enhance the goal of access to justice? This essay will have three sections. The first will detail the constituent elements of AI technologies and why they have generated interest, as well as scepticism, among many in the legal profession. The second section will paint a non-exhaustive landscape of solutions by mapping recent AI-driven legal products. The last section will look at AI’s impact on the profession and the political, economic and ethical challenges that are emerging as a result. It will examine the compulsions for advancing these technologies. 

Unpacking Artificial Intelligence

Unpacking the term ‘artificial intelligence’ is relevant, because it has several competing and overlapping definitions. Perhaps, in its most basic sense, it means the capacity of a computer to perform tasks commonly associated with human beings. This includes the ‘ability to reason, discover meaning, generalise, or learn from past experience’<2> and thereby find patterns and relations to respond dynamically to changing situations. Three key aspects can be adduced in AI systems (or cognitive computing systems): the capacity to find and gather information; the ability to analyse and make sense of the information gathered; and thereafter, to generate and make decisions.<3> Within the field of AI systems, there exist several technologies. These include natural language processing, where computer systems are able to analyse text to generate content, classify information and even answer questions;<4> machine learning, where computer systems through working with a large volume of data are able to discover patterns without precise and clear programmed instructions;<5> speech recognition, which involves the conversion of speech to text functions and vice versa; expert systems which have the ability to perform tasks which need the kind of expertise only humans have possessed until now; and vision, which includes the ability to recognise and analyse different images.<6> In the past two years, the conversation on AI has been magnified within legal circles. It began with the development of an IBM Watson powered robot called ROSS, the world’s first AI lawyer which, through its cognitive computing capabilities, could provide answers to research questions by mining data and deciphering trends and patterns.<7> Earlier this year, Cyril Amarchand Mangaldas became the first Indian law firm to employ software that uses AI to identify, extract, analyse and evaluate clauses and other information from contracts and other legal documents.<8> The recognition that AI will have an impact on the legal industry has been growing. In a widely cited study by Altman Weil of 320 law firms in the US, it was found that in 2015, 47 percent of the respondents felt that paralegals would be replaced by AI powered products, compared to 35 percent feeling the same way in 2011. The survey also showed that 35 percent of respondents in 2015 felt that AI could replace the work of first-year associates, compared to 23 percent four years previously.<9> These advancements in technology prompted Klaus Schwab, founder and Executive Chairman of the World Economic Forum, to suggest that the world is on the cusp of a “fourth industrial revolution” fuelled by technology, which is combining physical, digital and biological worlds, and which is likely to create an adverse impact on jobs and security, and increase inequality, unless organisations learn to adapt.<10> An Oxford University report suggested that over 47 percent of US jobs were at risk due to automation, and in China the figure could be as high as 77 percent.<11> While discussions on the use of AI in law have been taking place in academia for more than 30 years,<12> their influence into actual legal practice has happened only recently.  Some of the key factors driving the present resurgence include the increase of computing power (exemplified by Moore’s law, which states that the processing power of computers doubles every two years);<13> the availability of a large volume of data upon which technologies can be tested and developed;<14> the evolution of new and more effective algorithms which have synergistic connections with access to better hardware and also larger data sets;<15> and finally, AI’s access to capital, which has exploded in the last couple of years with 200 AI start-ups raising US$1.5 billion in equity funding.<16>

AI-driven Models for the Legal profession

In order to ascertain the potential impact that such advancements in technology can have on the industry, a scoping of existing technologies is useful.  The major areas in which AI is being utilised include in the conduct of predictive analysis, legal research, e-discovery document review, and self-help, as well as for administrative assistance and to enhance cyber security. Uncertainty is commonly associated with litigation because of the number of variables such as the forum, the judge and the evidence that can have an impact on a case. In order to address the constraints that emerge with risk in the profession, a number of solutions have emerged. Lex Machina is an example of a legal analytics platform that uses large volumes of litigation information to provide insights into the workings of judges, lawyers, parties and the cases before them.<17> Lex Machina emerged as a platform initially focused on providing data-driven strategies for intellectual property cases, so that risks could be assessed and decisions taken on the scope and strategy for the litigation. Premonition AI, another legal analytics firm, provides information on the effectiveness of litigators before particular judges by mining what it claims is the largest litigation database in the world. It aims to determine the track record of judges and lawyers to ensure that parties make choices based on empirical insights.<18> These platforms are designed to reduce the inefficiencies associated with litigation by enabling lawyers to develop strategies, predict outcomes and offer data driven actionable solutions to their clients. Another area where there appears to be great scope for innovation is in terms of the manner in which traditional legal research is undertaken. As research involves multiple processes of identification, categorisation and review of information, the need for tools to enhance the process is clear. The purpose of technologies, such as the one developed by ROSS Intelligence,<19> is to help lawyers find cases and secondary material using natural language processing. The tool allows researchers to ask questions in plain English and thereafter scans and examines its database to provide answers and readings from leading cases, and articles, relevant to the question. As the company’s CEO explains, “ROSS pretty much mimics the human process of reading, identifies patterns in text, and provides contextualised answers with snippets from the document in question”.<20>  Its focus is primarily on bankruptcy and insolvency law. Older legal research platforms like Westlaw and Lexis Nexis are also incorporating natural language processing into their searches. AI is also making inroads in terms of innovations regarding discovery of facts. OpenText is a platform that uses analytics and machine learning to identify facts that are central and important for litigation, as well as for compliance and governance. It allows the user to filter and focus the research by identifying facts and relationships that are important from the context of the case, through analysing communications as well as other information such as terms, sources and types of files. <21> While legal research requires critical engagement and analysis, document review is often seen as a task that is both time-consuming and mundane. Moreover, it is prone to human error on account of the large volume of information that often has to be processed. Kira is an AI powered platform designed to identify and analyse data by extracting information such as clauses and concepts from contracts and thereby allow the user to analyse trends and patterns between documents. It is being used in due diligence, contract analysis and lease abstraction, to ascertain the risks and challenges that can emerge, by comparing the documents concerned with vast volumes of previously assembled data.<22> Ravn, another AI platform, organises, analyses and summarises documents, with the aim of allowing organisations to increase efficiency and productivity while reducing risk.<23> In addition to tackling these different aspects central to the litigation process, whether predictive analysis, legal research or document review, AI has also been used for the development of chatbots that can respond to queries from users in the form of self-help tools. Lawbot is one such example of a solution designed to help victims of sexual assault in the UK.<24> Another application DoNotPay helps people with filing for compensation for delayed flights, as well as in deciding whether or not they should pay parking tickets. <25> Further, AI powered platforms have also been used by the industry to power their cyber security and build more robust analysis of the threats companies face.<26> It can also do the job of administrative assistants, who help organise tasks such as scheduling meetings, making travel arrangements or managing expenses. <27> Given these developments, the next section will attempt to situate the impact of AI on the legal profession by looking at the responses to technology, the impact on ways of working, and the regulatory challenges.

Transforming the Profession?

A mixed response to using technology

 The increasing pervasiveness of technology in the legal industry is also driven by the fact that clients expect value for money, lower risks and greater efficiencies in terms of time and outcomes from legal providers.  According to a Thomson Reuters study, there has been a 484 percent increase in the number of patents filed globally with respect to new legal services technology in the last five years, with 579 patents filed in 2016, up from just 99 in 2012. <28> This signifies that there is great demand from law firms for new types of legal services such as those that provide litigation support, risk analysis, due diligence and review services, to keep themselves competitive in terms of cost, precision and efficiency of services. It also indicates that clients expect new legal services to keep up with the digital age, reduce risk and costs while improving communication between consumers and providers of legal services.<29>  The economic case for the use of technology – in addition to the increasing capacities for the firm – is that it enables monotonous and tedious tasks to be completed by AI solutions, allowing lawyers to focus their energies on strategic and high value functions.<30> The introduction of these new technologies, however, will take time and planning. It will require a buy-in by all stakeholders, as well as clearly thought out implementation and integration strategies.<31>  In terms of law firms, a new survey in 2017 by Altman Weil<32> found that 25.9 percent of US law firms surveyed were wholly ignorant of the role of AI in the legal industry, 7.5 percent were already using some of the tools, while 28.8 percent were exploring options. However, the remaining 37.8 percent of firms, despite knowledge of the developments in AI, were not using it in their operations.  The large percentage of disaffection indicates a continued scepticism about AI in the law firm community and the reluctance to fully embrace the technology.  Law schools, a key constituency in the need to facilitate an embrace of the changes brought by AI, are having to adapt their courses and introduce elements such as ‘tech audits’ to prepare their students to become more tech literate.<33>

New ways of working

Among the key reasons holding back the use of AI, is the fear that technologies driven by it will result in a loss of jobs for lawyers. Jordan Furlong, a leading strategic legal consultant, has argued that junior lawyers should be open to technologies because their purpose is to add value, not to take it away from the lawyer. Furlong advances that if he was a young lawyer, he would like to know the following: “First, what is happening in the legal marketplace? Secondly, tell me what kind of skills, attributes and knowledge I need and who out there is going to engage me for that kind of work. And third, I need help being assimilated.<34>  He states that much of the debate so far has been on automation and not augmentation, the former being where the machine does everything relating to tasks like due diligence, or document or case review, as opposed to the latter where the machine facilitates the lawyer’s work.<35> While the former may result in job losses, a lawyer’s job also includes several other elements such as ‘strategy, creativity, judgement and empathy’ that are needed for tasks such as advising, negotiating and appearing in court, none of which can yet be automated.<36>  Thus, although the advent of technology is introducing a new vocabulary to the legal profession, with terms such as data analytics and algorithms becoming commonplace,<37> and this could be unsettling, lawyers, law schools and regulatory bodies should see these changes as enablers rather than threats and should embrace such change. According to the Canadian Bar Association Legal Futures Initiative, the key to adapting the profession to the future is innovation, which goes beyond mere development and adoption of new technologies. Innovation would require rethinking the ways in which lawyers are trained, how the profession is regulated and how the public is protected. <38>  In a dynamic new environment, where several legal tasks are rendered redundant, the report suggests that new disciplines will emerge, including those of legal knowledge engineers who can build online legal systems; legal process analysts who can develop means of distributing complex legal work in organisations; legal support system managers who develop and manage processes such as workflow systems; and legal project; and risk managers who introduce project management and monitoring techniques.<39>  Richard Susskind, author of Tomorrow’s Lawyers: An Introduction to Your Future, argues that legal work is going to evolve and transform from being a ‘bespoke’ service to one which, through a series of standardisations, systematising and packaging, will become a ‘commoditised’ service.<40> This will result in a decomposition of legal work both in terms of litigation, to include disaggregated tasks such as ‘document review, legal research, project management, litigation support, (electronic) disclosure, strategy, negotiation, tactics, advocacy’ as well transactional work, which would include ‘due diligence, legal research, transaction management, negotiation, bespoke drafting, document management, legal advice, and risk management’<41>. These changes will result in new ways of doing work in terms of how practices are structured, fees are calculated, projects are managed and clients engaged, as well as flexible models for working, including part law-part technology practices, legal hubs and marketplaces, virtual firms, etc.<42>

Regulatory challenges

With such rapid developments, several issues are likely to arise on the regulatory front. The first is the question of bias. Several courts in the US have started to use AI products for sentencing to determine the recidivism of persons convicted. While algorithms are usually coded to be neutral, the programmers who code them may not be so, or are likely to operate with some assumptions that may lead to bias.<43>  As Kranzberg first law of technology states, “Technology is neither good nor bad; nor is it neutral”.<44> A study by US news organisation Pro-Publica, of a risk assessment tool called COMPAS (Correctional Offender Management Profiling for Alternative Sanctions) bears testimony to this fact. It found that the tool was biased against black defendants who were incorrectly judged to be at high risk of recidivism compared to white defendants who were marked as low risk.<45>  This raises the question of more transparency in the methodology being used to develop such technologies, and also a degree of accountability and responsibility for the findings when they do not adhere to principles of fairness and equity. The second related challenge is that of unveiling the ‘black box’ of who is responsible for how an AI driven product operates. In many instances, the product is a result of multiple codes that analyse data and make decisions – leaving the developers unsure of how each component works, or how the tool makes its decisions.<46> This results in many questions regarding the elements of decision-making processes, the rationale behind particular connections and the reasoning for certain programming outcomes. There have now been calls for giving robots an ‘ethical black box’ which will ensure that these products keep track of their decisions and can explain how they were arrived at.<47> The third challenge is related to data confidentiality. It is essential that, as consumers of large volumes of data, AI service providers are also held to standards regarding how they access and use data. <48> Finally, though the field of AI is rapidly developing, there is need for the legal regulatory environment to develop a new vocabulary of guidelines and frameworks, not in order to stymie development in this field, but to evolve new ways to think about criminality, ethics and responsibility from a bottom-up understanding of how the field is developing,<49> as well as one that is self improving. This too could perhaps be driven by AI.<50>


The influence of AI on the legal profession presents a very exciting means of improving efficiencies through legal research tools and administrative assistants; of doing good through the use of chatbots as self-help tools; of reducing risk and uncertainty through predictive technologies and e-discovery tools, and as a tool for compliance and governance. Rather than mere automation, it has the potential to augment the profession, as it will open up new kinds of jobs, specialisations and influences for lawyers, law schools and regulatory bodies through their association with disciplines such as engineering, finance, management, political science and risk assessment. It however, requires members of the legal profession to be willing to embrace change, seeing AI not as a threat but as a tool through which their productivity, opportunity and potential can increase. Finally, given that this is a space seeing rapid development, it is essential that regulatory frameworks are developed to reduce the opaqueness and secrecy that surrounds the development of these technologies, the use of algorithms and data. It is vital that these products meet standards of transparency and accountability for their performances.


<1> “Q&A: Richard and Daniel Susskind on the Future of Law | Canadian Lawyer Mag,” <2> “Artificial Intelligence |Encyclopedia Britannica,” <3>Cognitive Computing: Transforming Knowledge Work,” Thomson Reuters, January 24, 2017. <4> “ Artificial Intelligence in Law: The State of Play- Neota Logic” <5> “Demystifying Artificial Intelligence,” DU Press, <6> “Why Artificial Intelligence is the Future of Growth _ Accenture” <7> “World’s First Robot Lawyer ROSS Hired by US Law Firm - Livemint,” <8> Kian Ganz, “Cyril Amarchand ‘First’ to Sign up for Machine Learning Contracts Software, but Is AI the Death or Future of Lawyers?”. <9> “ Law Firms in Transition-2015” <10> “The Fourth Industrial Revolution, by Klaus Schwab,” World Economic Forum, <11> “Robots Will Steal Your Job: How AI Could Increase Unemployment and Inequality,” Business Insider, <12> “An AI Law Firm Wants to ‘Automate the Entire Legal World,’” Futurism (blog), January 30, 2017, <13> “Demystifying Artificial Intelligence.”; Mark Purdy and Paul Daugherty, “Accenture: Why AI Is the Future of Growth,” n.d., <14> Ibid <15> “What’s Driving the Machine Learning Explosion?" <16> “Artificial Intelligence Explodes: New Deal Activity Record For AI,” <17> “Lex MachinaTM | LexisNexis,” LexisNexis® IP Solutions (blog), <18> “Premonition | Legal Abalytics | Law | Court Analasys | Litigation,” <19> “ROSS Intelligence,” <20>“YC’s ROSS Intelligence Leverages IBM’s Watson To Make Sense Of Legal Knowledge | TechCrunch,” <21> “Axcelerate EDiscovery & Investigations Solutions - Recommind,”, <22> “Kira Systems | Machine Learning Contract Search, Review and Analysis,” <23>“RAVN - IManage,” <24> “Cambridge Students Build a ‘Lawbot’ to Advise Sexual Assault Victims | Education | The Guardian,” <25> “Rise of the Robolawyers - The Atlantic,” <26>“Preparing for Artificial Intelligence in the Legal Profession,” <27> Ibid <28> “Thomson Reuters Analysis Reveals 484% Increase in New Legal Services Patents Globally,” Thomson Reuters, August 16, 2017, <29> “Global Legal Tech Is Transforming Service Delivery,” <30> “Artificial Intelligence Disrupting the Business of Law,” accessed November 5, 2017, <31> “Artificial Intelligence | Canadian Lawyer Mag,” <32> “ Law firms in transition-2017” <33> Samar Warsi, “How a Law School Is Preparing Its Students to Compete Against AI,” Motherboard, April 14, 2017, <34>“How Will Artificial Intelligence Affect the Legal Profession in the next Decade?,” <35> Ibid. <36> “A.I. Is Doing Legal Work. But It Won’t Replace Lawyers, Yet. - The New York Times,” <37> “Artificial Intelligence Law | Legal AI Solutions,” TOPBOTS, June 10, 2017, <38>“CBA legal futures initiative- Futures: Transforming the delivery of legal services in Canada” <39> Ibid <40>Richard Susskind, Tomorrow’s Lawyers: An Introduction to Your Future, 1 edition (Oxford, United Kingdom: Oxford University Press, 2013). <41> Ibid <42> “ The Future of Law and Innovation in the Profession” <43> “Artificial Intelligence | Canadian Lawyer Mag.” <44> Melvin Kranzberg, “Technology and History: ‘Kranzberg’s Laws,’” Technology and Culture 27, no. 3 (1986): 544–60, <45> “How We Analyzed the COMPAS Recidivism Algorithm — ProPublica,” , <46>“The Dark Secret at the Heart of AI - MIT Technology Review,” <47>Give Robots an ‘Ethical Black Box’ to Track and Explain Decisions, Say Scientists | Science | The Guardian,<48>The Ethics of Artificial Intelligence in Law” . <49> Ibid; “Preparing for Artificial Intelligence in the Legal Profession.” <50> Mark Purdy and Paul Daugherty, “Accenture: Why AI Is the Future of Growth.”
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.