The Robots Are Coming… But Not Yet
Par Kerrin-lee Whyte, Student
Kerrin-lee Whyte, Student, McGill University
Katarina Daniels, Lawyer, Liaison Librarian at Nahum Gelber Law Library, McGill University
In our modern world, information is available 24 hours a day, 7 days a week. With everything at our fingertips, consumers have started to demand more from traditional services, such as those offered in the legal field. When a person can easily obtain the laws governing their situation through a quick search on a free, user-friendly database, the purpose of paying lawyers fees’ for legal research, for instance, becomes unclear.
Advances not only in mobile technology, but in technology more generally, combined with increased dissatisfaction with costs and delay of legal services, have led to the rapid growth of legal technology, or “legal tech”. Legal tech may be defined as technology affecting the provision of legal services, generally rendering work performed by lawyers more efficient. Buzz in the media has surfaced surrounding legal tech, much of it negatively labeling it a job destroyer and signaling the end of the legal profession. This perception, however reasonably founded in the remarkability and mystery of new technology, lacks information about just what legal tech can and cannot do.
Artificial Intelligence in Legal Tech
Much of the reason legal tech has been labelled a threat to lawyers has to do with its increased association with artificial intelligence, or AI. In particular, two important attributes of AI are prominently featured in the legal tech market: machine learning and natural language processing. Machine learning is the attribute which allows AI to learn from experience, identify patterns and continuously improve its performance beyond what it was programmed to do, while natural language processing is the capacity of a machine to understand the meaning of written and spoken human speech, and to apply that understanding to perform human-like analysis. Both are undeniably impressive feats for a machine – especially where they surpass human capabilities – however it would be incorrect to assume that we are close to the image of AI as depicted in dystopian and sci-fi movies, where human-like robots are described as more than capable of out-performing biological humans. Specifically, AI cannot yet replicate important human attributes including human judgement, inference, common sense and interpersonal skills. Rather than eliminating lawyers altogether, instead AI is changing the way lawyers do business and interact with clients.
Technology Induced Change
One way AI is creating this change is by taking over some tasks in law firms, including work normally accomplished by lawyers. These include document drafting, including contract drafting, document management, including in eDiscovery, and document review, including in a due diligence context – in other words, tasks that are easy to automate as they often involve repetition or are standardized.
For example, Kira Systems offers an automated data analysis product for due diligence work that swiftly reads though numerous documents and identifies liabilities, ultimately allowing lawyers to make recommendations. It uses machine learning to continuously improve its ability to extract and analyze documents, and can easily be taught to identify new content; some users have reported an accuracy of over 90% in finding relevant provisions, as compared to 60-70% accuracy in human review. Krause and Hecker attribute lower human accuracy to “lack of experience or excessive workload.” Indeed, due diligence is usually carried out by associates or paralegals. Not only is Kira more accurate, however, it is also significantly more efficient. Kira boasts a time saving of anywhere between 20% and 90%, depending on the complexity of the task, translating into a financial savings for the client.
AI is also changing how lawyers and firms approach risk aversion. eDiscovery tools are already well-established in the legal tech market due to the exponential increase in electronically stored information (ESI), including emails, texts, voicemails, word processing files, etc. over the past decade. While they are used to acquire evidence for litigation, similar tools are harnessing the same technology behind eDiscovery tools, namely, machine learning and predictive analysis, to look for potentially harmful evidence before litigation even happens, in an attempt to predict and possibly prevent potential litigation. Using these tools, ESIs are searched for content that has a “pattern-match” indicating, for example, potential discrimination or sexual harassment.
Intraspextion is an example of one such tool. Intraspexion’s website describes it as a warning system that monitors a company’s emails, and reports emails with particularly risky content, allowing the company to can stop litigation before it happens or at least get a head start on preparation. With AI taking over this burdensome task of filtering through lots of material, in-house lawyers can spend time tackling more important tasks, while saving their employers time and money.
While automation of these tasks necessarily reduces billable hours, we are nowhere near complete eradication of lawyers. For now, lawyers are necessarily required to train the AI, filter through legal tech output and validate results, and make final recommendations, among other tasks. In addition, the reduction of time spent on these tedious tasks allows lawyers time to devote to more interesting work, including client development. Legal tech should be embraced as a means of changing how lawyers currently work.
Despite the many benefits legal tech has to offer, critics have raised a number of legitimate concerns. In particular, Marchant draws attention to issues regarding training and hiring, as much of the work that is being automated is work normally performed by new law firm associates. Further, AI may also negatively impact the public’s perception of the justice system. Incomprehensible algorithms behind AI create a black box scenario, where facts go in and decisions comes out. This may be met with skepticism about its seemingly arbitrary nature. Braegelmann describes the potential negative impacts of legal tech in the courts. For instance, Online Dispute Resolution, although improving access for some, comes at the cost of interactions in courtrooms. Will judges be robbed of some aspects of their reasoning if they do not see or hear a witness in person? Will victims feel cheated if their issues are not resolved in person? Will parties be less empathetic in negotiations when they can hide behind a screen?
What the future holds
Currently, these questions remain unanswered. Concerns that can be put to rest in the present, however, are those about the elimination of lawyers. Legal tech is only starting to replace lawyers in activities that are easy to automate, such as drafting documents or managing or reviewing contracts, where arguably lawyers were not needed in the first place. Legal professionals should not be particularly worried about work loss though, as replacing tasks will also create tasks. For instance, AI will require lawyers to train and continuously work with the systems to adequately deliver services. Some experts predict that legal AI will actually augment lawyer work rather than replace it. Micha-Manual predicts that in the future legal advice will be a hybrid of “human knowledge and experience […] enriched and intertwined with digital and statistical experience and predictions.”
The media’s portrayal of legal tech entering the playing field represents extremes; the legal profession is not ending. Although some uncertainty remains surrounding legal tech, this is our reality. Lawyers need to start assessing how legal tech can be incorporated in their work to provide more efficient and accessible legal services. Our society is moving forward technologically, and so must the services provided to it.
Interested in how legal technologies will impact the practice of law closer to home? Be sure to attend this year’s Legal.IT Conference, hosted by the Young Bar of Montreal. Our opening panel will feature a discussion between our guest of honour, Me Sonia LeBel, Minister of Justice and Attorney General of Quebec, and Me Karim Benyekhlef, Ad.E., professor of law at Université de Montréal, moderated by Me Dominic Jaar, Ad. E., Partner and National Leader, Forensic Technology Services at KPMG. Just one of a multitude of exciting plenaries this year! Details and registration available here.
Les chroniques du CTI sont rédigées par un ou plusieurs membres du Comité Technologies de l’information (CTI)
dans le but de susciter les discussions et de soulever les réflexions
au sein de la communauté juridique à propos des nouvelles technologies
et le droit. Les auteurs sont donc seuls responsables du contenu des
articles et l’opinion qui y est véhiculée n’est pas celle du JBM, mais
bien celle des auteurs. Si vous désirez rédiger une chronique, envoyez
un courriel au firstname.lastname@example.org.
To register at the Legal.IT Conference, click here.
 Hartung, Markus, “Thoughts on Legal Tech and Digitilization” in Markus Hartung, Micha-Manuel Bues & Gernot Halbleib, eds, Legal Tech (Munich: Verlag C H Beck oHG, 2018) 3 at 5.
 Bues, Micha-Manuel, “Artificial Intelligence in Law” in Markus Hartung, Micha-Manuel Bues & Gernot Halbleib, eds, Legal Tech (Munich: Verlag C H Beck oHG, 2018) 265 at 265.
 Susskind, Richard, The End of Lawyers (Oxford: Oxford University Press, 2008), ch 1.
 Krause, Nils & Ronja Hecker, “Commercial law firms under the influence of Artificial Intelligence – Status report and outlook using the analysis software Kira as an example” in Markus Hartung, Micha-Manuel Bues & Gernot Halbleib, eds, Legal Tech (Munich: Verlag C H Beck oHG, 2018) 49 at 54.
 Braegelmann, Tom H, “Online dispute resolution – ODR” in Markus Hartung, Micha-Manuel Bues & Gernot Halbleib, eds, Legal Tech (Munich: Verlag C H Beck oHG, 2018) 233 at 240.
 Ibid at 244.
 Hartung, Markus, “Four theses for the future” in Markus Hartung, Micha-Manuel Bues & Gernot Halbleib, eds, Legal Tech (Munich: Verlag C H Beck oHG, 2018) 397 at 398.
 Krause, Nils & Ronja Hecker, supra note 4 at 55.
 Bues, Micha-Manuel, supra note 2 at 273.