From Hyperbole to Real Talk, Law is Finally Ready for AI

Thomas Hamilton

Coming out of the crazy energy of ILTACON 2017, I was inspired to put together a quick summary of the three panels on artificial intelligence that our CEO Andrew Arruda was invited to moderate (see below for a rundown), and also use the opportunity to include some of my own thoughts on how the industry’s knowledge of AI is maturing.

(For those not in the know, ILTACON is the premier legal technology event, now celebrating its 40th consecutive year. This year it took place in fabulous Las Vegas, from August 13–17).

The biggest takeaway is that in the last two years, AI in the legal space has gone from science fiction, to niche subject, to mainstream interest and now all the way to over-hyped. It’s great seeing such incredible acceptance, but it also worries me that not everyone approaching the space is going to be as informed as they should be. The ROSS team put together an AI buyer’s guide here and an AI implementation guide here in advance of ILTA for folks who are both evaluating AI solutions and also trying to drive practical adoption of them in their organization, which I think are even more important coming out of the three panels I observed.

The two main themes we stressed:

  1. Find a vendor who is truly using cutting-edge AI technology
  2. Make sure they’ll work with you to engage your users while also continuing to build a solution that can truly help.
“AI technologies will be the most disruptive class of technologies over the next 10 years due to radical computational power, near-endless amounts of data and unprecedented advances in deep neural networks. These will enable organizations with AI technologies to harness data in order to adapt to new situations and solve problems that no one has ever encountered previously.” — Mike J. Walker, research director, Gartner.

What’s neat about the above quote is here you have the head researcher at Gartner, Mike Walker, clearly giving people a heads up that AI is here and will become ubiquitous soon. Strap in legal industry because things will not be the same for long!

As Andrew Arruda said at ILTACON, “folks need to stop over-selling AI and focusing on the powerful applications that firms are implementing right now. People in legal technology need to stop over-estimating where AI is today but most importantly, stop underestimating where it will be tomorrow. AI breakthroughs are happening on a daily basis, we are living in historic times.”

Big Law is already using AI

While the acceptance of AI by the legal community definitely seems to be growing, I think it’s also important to note that the awareness is growing disproportionately in large firms. While it makes sense that large firms (and their large legal department clients) would have the necessary resources to be first movers in some instances, I think it’s important we don’t lose sight of the incredible things AI can do for smaller firms, sole practitioners and legal aid organizations.

The Access-to-AI Gap

Having built out our access to justice program, where we provide ROSS access entirely for free to pro bono organizations, I know first hand how much these organizations benefit from recent developments in the legaltech space. What’s more, they also often have the most pressing use case, as they don’t face any of the innovation obstacles we’re all so familiar with in the Big Law space (minimum billable hour targets, declining realization rates, etc.). The same can be said for smaller boutique firms, where the economic buyer of an AI solution is often the end user. Historically these firms are not only some of our most active users, they are also frequently our most engaged and vocal partners who take an active role in working with our engineering and design teams to direct the evolution of ROSS.

Is there a reason for this under-representation? I’m not sure. Obviously ILTA attendance is more easily affordable to larger firms, but I think there a few easy steps to widening the playing field. If the goal of ILTA is to be the marqee event for legal technology, and therefore putting vendors in front of major decision makers (which, for the record, I have no issue with), then why not start inviting tech solution procurement teams from different state bar associations to the table? Many smaller firms are not in touch with the day to day development of legal tech and AI, but they do rely on their bar associations as guides in the field. Just a thought, and I think there are lots of different strategies we could pursue as an industry here. Looking forward to hearing anyone else’s.

Below you’ll see that I’ve included my reviews of the panels as well as audio recordings of each, so now that you’ve heard my opinion, why not see what you think of the panels yourself?

Will Computers Replace Lawyers? The Myths, Realities and Future of Artificial Intelligence and Automation in the Law (Part 1 of 3)

The first of the AI panels was comprised of Martin Tully, Co-Chair of Akerman LLP’s Data Law Practice, and Samuel Whitman, Mayer Brown’s Knowledge Management Leader. Andrew opened the intimate, fireside style discussion with an introduction of the generally accepted definition of AI (machine learning, natural language processing (NLP), vision, and speech), with the obvious qualification that on its whole AI is difficult to define.

Martin, Sam, and Andrew then proceeded to have a whole lot of fun talking about AI. The first point of agreement was that AI is really just the most recent version of previous hot topics such as cloud-based services and big data, which have garnered a great deal of attention in years past. Sam used a favorite expression of mine, which is that “something is considered AI right up until we use it, at which time we start calling it software.”

Both Martin and Sam discussed the power of AI to aid in helping automate the repetitive, low value-added tasks that lawyers often dread doing. I believe these tasks were described as, ahem, “soul-destroying” which certainly echoes how my former colleagues and law school classmates in Big Law still view them! Martin put it very succinctly when he said that AI can now function like the prep cook, allowing the lawyer to focus all of their time and attention on being a chef, and that we should keep in mind in that in the future, lawyers who eschew using AI tools may begin to face moral/ethical/professional obligations to do so.

Both Sam and Martin seemed to feel that it was time for internal champions to begin pushing for AI adoption and innovation at law firms, and that the consequences for those firms who did not take this step could be potentially dire in the years to come. Martin felt that three categories of law firms will emerge: First adopters who will rapidly advance and enhance their practices with AI, a second group that will take a wait-and-see approach but who will be forced to adopt AI over time, and a third group who will end up in the dustbin of history.

Artificial Intelligence in Law: Top Products in Action (Part 2 of 3)

The second panel followed quickly on the heels of the recap of the first (now in a larger room), which was a really fun way to start things off and allowed us to continue with the great momentum and buzz.

The panel consisted of Katie DeBord of Bryan Cave, Ron Friedmann of Fireman and Co., Neil Cameron of Neil Cameron Consulting Group, Stephen Allen of Hogan Lovells, and Steven Harmon of CISCO. Unlike the first, more intimate panel, today’s followed a more formal, one-at-a-time format. While there was still a great deal of discussing legal technology at law firms, there was also (owing at least partially to Steven’s presence) more discussion of client needs than I typically hear at this sort of panel, which was refreshing.

A few major takeaways. Firstly, AI is here to stay, and is addressing major industry needs. For now, the focus of dialogue needs to be not on replacing lawyers (if that’s ever even going to happen) but rather on focusing on what lawyers can’t do, won’t do well, or even better that lawyers don’t want to do and clients don’t want to pay for.

At the same time, not all lawyers are approaching AI in the same way. Consistent with adoption by large corporate legal departments, very large firms are disproportionately the ones to adopt AI. Katie Debord’s discussion of Bryan Cave’s approach to AI adoption, and the lessons they had learned in doing so, was illustrative as to why this trend exists.

Larger law firms not only have access to a C Suite (notably CIOs, CKOs) with official mandates to focus on technology and AI adoption. They are also then provided with the resources they need, both in terms of team members and budgets, to invest in both internal AI initiatives as well as evaluating and then deploying “off the shelf” AI solutions to rapidly automate high volume, low margin work.

Across the board it was agreed that while AI will be a gamechanger, it has also been accompanied with too much hype. In both in house departments, and with law firms, buyers are interested in clearly demonstrable value and ROIs, which dovetailed perfectly into the final AI panel.

Lastly, the idea that companies are now slapping AI into their product lines was discussed — this is one of my biggest pet peeves and something I’m sure we will continue to see more of as companies pivot from what they were doing before into becoming an “AI legal tech” company.

Artificial Intelligence in Law: From Theory to Practice (Part 3 of 3)

Audio recording of the third AI panel in the series can be found here.

This panel was comprised of Anna Moca of White and Case, Amy Monaghan of Perkins Coie, Jonathan Talbot of DLA Piper, and Julian Tsisin of Google. The panel was coordinated by Ginevra Saylor, of Dentons.

The panel was broken into a series of discussions, which allowed it to serve as a great summary and dialogue on much that had been discussed in the two previous days.

There was a large consensus on the importance of both consultative interaction with potential legal technology partners prior to making a decision, as well as the need to identify great partners who will give you a seat at the table in the continued rollout of their AI solution.

It was agreed that once that partner is identified, internal education is crucial with respect to addressing misconceptions about new technology initiatives with lawyers. Something I found very encouraging was the agreement on the need for someone to oversee change management with any AI implementation.

Once a solution had been prepped for rollout, it’s important to identify key stakeholders, as well as the ideal pilot group. Firm wide / legal department wide rollouts up front did not appear to be the goal, and capturing ROI and quantitative gain data was key.

Everyone seemed to feel that there were pros and cons to building v. buying AI solutions, but that generally you wanted to leverage existing experience, while continuing to evaluate other outside solutions that can either aid in your existing processes, or provide entirely new capabilities that the firm doesn’t currently possess.

Lots of different great strategies were discussed in terms of building excitement and buy-in among users, including starting with smaller deployments to then land and expand, as well as creating “just in time” training programs for lawyers who are about to be staffed on a file where they will be expected to leverage a new AI solution. Above all, it was clear in the adoption phase that avoiding over-hyping a new solution as AI was crucial — internal champions and change managers have to focus on communicating the current and future value to their end users.

Overall, the third panel was a great wrap up, as well as a summary of a number of the subjects and themes being explored in the previous 2 panels, and at ILTACON as a whole. I was impressed with how much awareness there was of the need for change management at large firms, although as mentioned I would have appreciated more discussion or at least awareness of how smaller, more easily deployable solutions could be used, especially outside of Big Law.

By 2022 A.I. in Legal Will Be…

What’s clear is that AI in legaltech is here to stay. What’s less clear is what that will look like 6 months from now, 18 months from now and onwards. I hope we’ll see a continued expansion both across all aspects of usage in Big Law, but also more importantly into all the other parts of legal services, so the incredible benefits of this software revolution will be shared evenly by every segment of the population. As with every new wave of technology, the rewards will fall disproportionately to the first movers, as well as those that take the time to invest properly in the training, change management and collaboration required.

At ILTACON there was a board that said, “By 2022 A.I. in law will be…” and attendees added their predictions. What stood out to me was, “…a regular part of the practice of law” and “ubiquitous and status quo.”

With AI, we have a toolset capable of transforming the legal industry for the better. We owe it to everyone to use it properly.

Thomas Hamilton

VP Strategy and Operations at ROSS Intelligence, and as employee #1 he co-ordinates efforts across the company to ensure that sole practitioners, legal aid groups, law firms, government agencies, corporate law departments, state bar associations and law faculties are able to benefit from cutting-edge developments in artificial intelligence research.