In this Q&A our China Consultant Levana Huang interviews our Senior Advisor Kenneth Tung, to discuss the future of AI within the legal industry.
Kenny Tung is an in-house community thought leader and APAC Senior Advisor for SSQ. Tung’s previous positions include Chief Legal Counsel of Geely Holding Group Co. Ltd and Greater China General Counsel at Kodak. He also worked as Legal Director/General Counsel in Greater China or Asia at PepsiCo, Goodyear and Honeywell.
Timeline of AI Adoption in the Legal Industry
Q&A with Kenneth Tung
L: Can you share with us how your initial interest in artificial intelligence (‘AI’) and machine learning (‘ML’) burgeoned?
K: It may sound apocryphal, but in college I thought seriously about studying AI, and took courses in computer programing, logic, physiological psychology and statistics, before eventually moving to a degree in history and economics to pursue a career in law and business. The notion of AI 25 years ago was different from the mindset that there is today.
L: We have all heard of the concept of AI, but not many of us can dissect what it means or entails. Can you help us understand how AI can be used to augment our work efficiency?
K: First, there are many branches in AI, such as natural language processing and machine learning. Unfortunately they are often mixed up and lumped under the AI umbrella. Programmed and hardwired expert systems have for a long time been considered the main body of AI, but the salient features of many AI algorithms today perform with agility and flexibility above expertise systems can deliver when faced with evolving and complex circumstances. The notion of AI is dynamic and shifting. The common calculator was a modern and innovative piece of AI in the 60s when many students may still remember how to use the slide rule (or the abacus!). Automated spell check did not remain innovative for long, before it became a nuisance. Sometimes AI ceases to be AI when it has been deployed and another box of previously exclusively human domain has been checked. So today, with the common expert systems in the market, machine learning is leading the campaign to identify and deliver valuable outcomes in many industries, and the legal service sector is no exception.
L: Can you give us some examples of how AI/ML are transforming the legal sector? (i.e. predicts legal outcomes, DD, contract review, etc.)
K: Perhaps the most commonly machine learning technology is in the e-discovery space where algorithms take the front line to scan and categorise verbiage in terms of relevance before human lawyers follow up with validation and further proof. The optical character recognition of images of words into digital words on our screen is also a close cousin whether they might have been born from expert systems or machine learning. LexPredict and LexMachina are also leading examples of applications in the legal space that are based mostly or partly on machine learning. However, what LegalTech needs to focus more on are applications designed and born of problems to be solved for customers rather than limited to efficiency and cost cutting for legal tasks.
L: According to Dan Jansen, the first CEO of NextLaw Labs, the legal industry (in the U.S.) spends less than one percent on R&D. Do you think the legal industry is falling behind in terms of adopting AI usage and the reason(s) for this?
K: Yes, the legal industry is behind, and no, the industry’s adoption of technology, let alone AI, lacks successful designs and adaptions, therefore has not gained enough ground to be considered as having fallen behind. However, we are not unique in this quagmire, because many industries and organisations often have the cart (technology) before the horse (people and process) and bandy technology notions about without being integrated to an underlying business strategy. People shop for technology expecting to solve some inefficiencies, to head off competition or simply to brand themselves as being digital, and many technology providers who have something off the shelf with minimal tweaking are happy to oblige. The best known secret is the dismal rate of technical adoption as the user’s real problem to be solved usually remain as it was, such as the legal service provider and their client’s failure to scope, identify and prioritise value which usually goes beyond cost reduction. Another challenge is the lack of a culture to invest lawyers’ time alongside coders and algorithm to capture lawyers’ workflow and decisions.
L: As programmers train AIs to think like lawyers, do you foresee the work of paralegals and legal assistants being diminished to tasks such as inputting data? How do you think the associate role is changing as law firms integrate technology on a more sophisticated level?
K: Digitization of information has transformed legal practices. To the extent that legal assistants and some associates are given more manual tasks which aren’t automated. While increasing efficiency is one vector of change, efficacy is arguably the other more important vector. What is happening here is clients are questioning why some tasks of legal services even need to take place and challenge the legal profession to re design their way of work from objectives to outcomes. Machine learning simply is a proof of concept that it is best to take over generating predictions in decision making while people should focus on weighing and working out outcomes within complex organizations.
In this new world, the legal part of the decision process is no longer a separate sub routine of the overall business decision making, but an integrated one. The more junior associates and legal assistants will work more in 1) the front end of decision process, data, its origination, effectiveness, insulation from biases and 2) validating the outcome of the process. This way will be a more satisfactory experience than the previously insular and repetitive tasks that many have been experiencing.
L: How will the prevalence of AI technology bring about staffing reforms of law firms? (i.e. less need for paralegals and interns, establishing R&D departments, higher demand lawyers with tech backgrounds, etc.)
K: In addition to legal advice, law firms also offer the fire power to plough through words, whether in the case law to prep lawyers in navigating applicable rules or the warehouse full of contracts and materials to find the “smoking gun,” and even to mark up documents to manage document versions. Digitalization of information has changed most of that, and this is before AI applications have even entered the picture. Automation of certain legal tasks aside, emerging AI tools will make the greatest gains for the legal sector in helping lawyers to help clients navigate the if-then scenarios. To be useful in these tasks, lawyers and legal assistants need to be the T-shaped professionals that have been emerging in other fields and act as part of the solution to clients’ problems rather than practicing law for the sake of practicing law.
L: As large law firms launch internal R&D units, new legal tech platforms, establish subsidiaries, (i.e. Ashurst/Advance, Clifford Chance/Applied Solutions, Simmons & Simmons/Time off for innovation, Reed Smith/GravityStack and many others – https://prismlegal.com/rd-big-law/) all within three years, is there a specific model of adopting AI that you foresee to be the most beneficial for law firms?
K: These are early days for R&D in the legal sector. In most fields, R&D develops products and services that represent new and better value propositions for customers; however, most clients of legal services today are looking for value. Indeed, LegalTech and new products in the field have been focusing on making lawyers more efficient, but rarely help clients to solve problems, let alone deploying AI to harness value from the clients’ data. Lawyers need to be truly and pervasively commercial while clients need to be strategic and make this a part of their operations and strategy. Only when the two have started to integrate legal work into business can they explore people-process-technology and look for outcomes that profit from AI and other technology.
 T-shaped professionals are characterized by their deep disciplinary knowledge in at least one area, an understanding of systems, and their ability to function as “adaptive innovators” and cross the boundaries between disciplines.