Mintz On Air: Practical Policies — Litigating in the AI Age

AI is already shaping the litigation process. And while it can benefit that process, it also introduces new risks.

In the latest Mintz On Air: Practical Policies episode, host Jen Rubin and Member Mathilda McGee-Tubb explain where AI helps and where it creates exposure, and how litigators should approach both.

They cover:

• How AI is reshaping and opening up court access, and the impact on the litigation process
• Where AI can benefit lawyers and litigants, and some of its complications
• How AI is influencing strategy, evidence, and case narratives and helping the storytelling process
• Why AI is a starting point — not an ending point — in the litigation process, and why human judgment remains critical

For in-house counsel, business leaders, and litigation professionals, this episode explains how AI helps streamline the litigation process, how it helps derisk litigation, and how to do so without sacrificing credibility or good human judgment.

Mintz On Air: Practical Policies — Litigating in the AI Age

Jen Rubin (JR): Welcome to the Mintz On Air: Practical Policies podcast. Today’s topic: Litigating in the AI Age. I’m Jen Rubin, a Member of the Mintz Employment Group with the San Diego–based Bicoastal Employment Practice, representing management executives and corporate boards. Thank you for joining Mintz On Air. If you have not tuned in to our previous episodes and would like to access our content, please visit us at the Insights Center at Mintz.com, or you can find us on Spotify.

Today I’m joined by my Boston-based litigation partner, Mathilda McGee-Tubb, who focuses her practice in complex commercial litigation, class actions, white collar criminal and administrative investigations, probate and fiduciary litigation, and insurance coverage disputes. I invited Mathilda to join me today to discuss some of the many ways that AI impacts how we, as litigators, approach preparing our cases and managing our cases, and how AI might change the manner in which we might actually try cases.

At the outset, I want to make clear that when I talk about cases, I’m talking about commercial disputes that are in, or will become, a part of the litigation process. That might mean a court case or even a matter subject to arbitration. Like many of you, while I find AI to present a grazing table of possibilities that go well beyond our law practice, what is truly fascinating is that it is already impacting the approach to the dispute resolution process. I thought a conversation between two, shall we say, sensible and practical trial lawyers about how AI impacts what we do would be an interesting topic. So thank you, Mathilda, for joining me today on the pod. 

Mathilda McGee-Tubb (MM): Thank you, Jen. I’m happy to be here.

JR: I’d like to set up our conversation around four topics that I believe play a huge part in the what, when, and how of dispute resolution for litigants and how AI might already be impacting those areas. Those topics include (1) courtroom access, (2) litigation timing, (3) litigation content, and (4) how parties might approach litigation risk. 

How AI Is Changing Access to Courts and Legal Services

JR: Let’s start with access. By access I mean how people are able to have their disputes heard in the court system or in arbitration. Mathilda, what are you seeing out there in terms of AI’s impact on court access? 

MM: We’re seeing significant increase in access to the court because of all the resources available through AI. I won’t go so far as to say that AI has leveled the playing field, but it has certainly opened the doors to the stadium. AI can really help individuals who aren’t sure where to start. Using AI, they can research potential claims. They can find an attorney. There isn’t that impediment we have seen historically of tracking down a lawyer, talking to them; maybe they’re not the right lawyer, then you’ve got to go somewhere else. You can do a lot of the research yourself and start to figure out if you potentially have a claim.

The other important thing AI can do is translate, into regular vernacular, some otherwise challenging legal concepts. For example, what is slander and what do you need to establish it or prove it? What are fiduciary duties? What is wrongful termination? These are concepts that everybody has a general sense of, but you can use AI tools to evaluate whether you would have a claim and then actually draft a complaint and file it all without a lawyer.

JR: I love this analogy to the gatekeeping and opening the doors to the stadium. It’s providing people with tickets in and translating the regular vernacular. It’s kind of a peek into the coach’s playbook, right? These secret things that you and I learn as lawyers that AI is opening up and providing to the public. But there’s a flip-side to easy access. How does this all-access, all-of-the-time impact the court system? What does it mean for the quality of legal services?

I’m joking about the secrets we learn, but you and I have trained for a long time, first in college and then in law school, and then with other, more experienced lawyers who gave us their playbooks and taught us a lot of those things. In light of that, do you see some downsides that come with this all-access pass?

MM: Definitely. But before I get into the downsides, you mentioned how we’ve trained for a long time, and that really does play into the value we add to our clients. I do want to note that I’ve seen how AI can support more under-resourced lawyers. I do a lot of class action defense against some very good plaintiff-side class action lawyers, and they tend to be small shops, lean and mean, and they haven’t historically had the human capital to engage in extensive motion practice or to be on our level in terms of what we can turn quickly. But their briefing with the use of various tools they have available now is significantly improved. In some ways, when you have a lawyer who’s trained the same way you and I are, Jen, those parties are getting closer to a fair fight because of the resources that AI provides. 

But in terms of downsides of general access: When we look at how this increase in complaints or increase in ability to access the court system without a lawyer, how that’s playing out, the courts are swamped. Anybody can draft a complaint that on its face looks like it states a claim. Any lawyer might tell you it’s not worth the paper it’s written on, but it gets pretty far through the process because it’s so easy for somebody who is unrepresented or underrepresented to move forward with it. 

Now, why is that a problem? Fundamentally, we believe everybody should be able to use the court system to right wrongs they’ve suffered. But not having some sort of threshold to get into the court means that everybody is experiencing delays in access to justice, that judges are frustrated, that court staff are overworked, and everybody is more likely to make mistakes.

One thing we’re seeing in Massachusetts right now in the Superior Court, the courtroom clerks are so far behind in docketing the papers being filed that you could file something on Monday and it won’t show up on the public docket until three weeks later. That is seriously compromising the public’s constitutional right to access to the courts and to know what’s happening in our courtrooms.

Another downside: When you have AI-generated filings with the court, that can be misleading to the court and also to the public and can be problematic for a defendant who is responding to potentially false or inflammatory claims, and even to citations to cases that don’t exist.

JR: I think this is absolutely true. And we’re seeing this all across the country. If you’re opening the doors to the stadium and the crowd rushes in, you know what? We don’t have the system and the infrastructure to support this all-access, all-of-the-time. As you point out, and I completely agree, we are an open public court system and everybody should have the fair opportunity to have access to that court system. And AI’s impact on all of this is going to be interesting in terms of dealing with these things as we move forward. It’s creating challenges as well as opportunities, including, as you point out, enhancing some additional tools for lawyers to even better represent their clients.

How AI Is Transforming Litigation Speed and Discovery Efficiency

JR: How else is AI impacting the pace and speed of litigation? Can you give us a few examples and speak to that? 

MM: I think the biggest one is in discovery. We exchange a significant amount of documents in a lot of our cases. Often we have to produce thousands of documents and review hundreds of thousands to get there. For most of the time I have practiced, that has been a pretty time-consuming process. When I first started practicing — and Jen I’m sure when you first started practicing too — we were literally flipping through hard copy documents. 

JR: Oh, yes. The documents. 

MM: The transformation has been profound. We can now use continuous active learning and these large language model–based approaches to get through a significant volume of data in a very short period of time. We’re talking thousands of emails, Excel files, Word documents that can be processed quickly and cost effectively. I am seeing the courts impose much tighter time frames for discovery because of all of the resources available to streamline that. Frankly, that’s a benefit to our profession, because we don’t spend four years in college and three years in law school to sit there flipping through documents. We want to get to the meat of the issues. 

JR: Now, talking about the meat of the issues, there’s a flip-side to this, right? AI can help enhance our ability to digest large volumes of data and to quickly generate summaries and things that can enhance our ability to try the case. But faster doesn’t always mean better, right? 

MM: Absolutely. Sometimes faster means more mistakes or it means skipping over the cerebral, fun part where you brainstorm. What are our best claims here? What are our best defenses here? How can we use the human element to our advantage? How can we tell a story? I do think we need to continue to use AI strategically, carefully, thoughtfully, so we aren’t relying exclusively on machine-generated content to develop our case themes, for example. 

One of the things I think a lot about, and I know you do too, is training the next generation of lawyers. We want them to use these resources, but we also want them to have the benefit and the privilege of engaging in the thought process of developing the client’s story and the client’s rationale for how they acted the way they did and how that’s going to persuade whomever the decision maker is. Because at the end of the day, you have a human — whether it is a judge or jury or an arbitrator, or even the parties in a negotiated resolution like a settlement — who’s deciding the outcome of your case.

JR: You know this from being in the courtroom: Part of the trial experience is experiencing the reaction to the evidence that you’re putting on the witness who’s sitting in the box. How that witness responds to questions. I think it’s important not only for other lawyers, but for litigants to recognize that when you interact with those folks who are trying your case and making decisions about the validity of your case, you need to have an understanding about how your story is unfolding. It’s not all about the data, and it’s not all about summaries and compilations, and it’s clearly not all about boxes of documents. Having that human in the loop is important in terms of that presentation. At the end of the day, a trial is telling a story, right? Part of what we do as litigators is presenting that story.

AI in Legal Strategy: Shaping Narratives and Trial Themes

JR: That’s a great segue because I want to talk about the substance. We know that every case is taken on its own merits and its own facts, and on the witnesses and how the witnesses are presented and how the documents are presented. As lawyers, we try and work with our clients to present those cases to the fact finder as best we can. And of course, we always hope the fact finder will rule in our favor, whether it’s a verdict from a jury or a judge’s decision. Mathilda, how has AI influenced how we present case strategy and how we tell those stories? 

MM: There are so many ways we’re able to use AI to influence how we tell these stories and how we identify factual gaps, how we think about different tweaks in our story. The same facts can be framed different ways depending on what story you think will be most compelling. We’ve got some great resources available that can help us think about and evaluate questions like, “Is this witness going to be credible? Is this argument going to pass the red-face test, as we say?” I think it’s helpful to get both human input and AI input, because you’re going to have your case evaluated by people who have most likely gone in and utilized an LLM to think about how to help guide how they’re thinking about the case.

One of the things we grapple with a lot now is the credibility of evidence being presented in a given case. We’ve all heard and seen stories about AI-generated photos, AI-generated videos — and now we are even confronting AI-generated emails that look real but are not. And it can be incredibly problematic to prove that these are not real. It goes to the theme of, what are the benefits? What are the challenges? I think we’re still learning how to properly deal with inaccurate presentations of evidence in a trial and helping the ultimate fact finder, whomever that may be, understand what’s real and what’s not. It’s the question that’s faced trial lawyers since the beginning of time. What’s the truth? What’s not the truth? We’re just confronting it in a different way now. 

JR: It’s so interesting to me because it’s credibility, right? That’s what we’re really talking about. Who’s credible? Do you believe this witness? Do you believe this document? AI is both helping with credibility, because it’s helping us as lawyers to quickly assess and reduce a lot of data to help make those decisions. But by the same token, it also causes mistrust in the process. If you don’t believe the email, you don’t believe the image, you really lose faith. It’s so interesting to me how something could be so beneficial and at the same time so challenging.

Trust but Verify: Best Practices for Using AI in Legal Research and Writing

JR: What about the skills that lawyers need to develop, Mathilda? We’ve talked a little bit about this, but what are your thoughts in terms of both lawyers who are just starting out and lawyers who have been practicing for a long time? I’ve been practicing for 37 years and I’m finding it’s still teaching me how to be a better lawyer. Can you talk about how we can use AI to help hone our skills? 

MM: I agree with you, all of us can use AI to help hone our skills. And you’re right, the new generation of lawyers need to begin with understanding how they’re going to use AI. Then the rest of us can continue to use it to improve how we practice.

One of the most important parts here is going back to the old Reaganism, “trust but verify.” I think that new lawyers, the younger generation of lawyers, and those of us who have been practicing for longer, need to understand that what you get out of AI is only as good as what you put in, and you still need to verify it.

So validating sources, using your own independent legal judgment to say, “Does this look right? Let me check what this LLM has told me by going to a treatise or by looking at the statute myself.” Then you look at them together and you say, “This makes sense, I’m glad I sought the assistance of AI because it helped me think about this in a different way. But I know that the information this AI tool is relying on is the right information because I vetted it myself. So I can use what I’ve gotten from this AI tool to inform my own recommendation to either the more senior lawyer or to the client about how we should proceed.” AI is a starting point, not an ending point, I would say.

JR: I think that’s great. It’s a thought process at the beginning. And I would add to the list of things I would do, something I always enjoy, which is talking to my colleagues. Not just colleagues who are more senior to me, but also colleagues who are junior to me. When you work collaboratively and across generations of lawyers, it’s interesting how people have different perspectives and it can really enhance your thought process as a human. So I’m just going to put in a pitch for humans here if I can. 

Using AI in Litigation Risk Assessment

JR: I want to pivot to the last topic. Litigation has always been about managing uncertainty. That’s one of the big topics when clients come to me with a particular dispute. AI may give lawyers and clients more data, but it also can create new kinds of uncertainty. Managing that uncertainty leads into a discussion about risk and de-risking the litigation process. One way to de-risk is to avoid a claim completely. I wish we could always figure that out; we can’t always, as you know. Another way is to resolve a claim. And a third way is to avoid a jury completely by avoiding the court process. What are some of the ways you see AI impacting how we assess risk and mitigate it in the courtroom or in the litigation process? 

MM: As you noted, Jen, a lot of cases do settle. That’s the reality. And what’s interesting is most settlement agreements are typically private and we don’t know the terms. That’s why clients come to us. They need the advice of counsel to know what they can reasonably expect to be the outcome if they take the case to trial or if they settle. What are the different circumstances or what are the different terms they might be able to reach? We’re seeing a great increase in what AI can tell us about settlements. I don’t know fully what the sources are that some of these settlement tools can use — they’re primarily drawing from public settlements, of course, but they’re able to aggregate data in a way that would take a significant amount of human time to do. That wouldn’t be worthwhile. But when you can aggregate it in this way, you can more accurately advise a client on the likely damages they would face if they took this to trial. 

I’m typically defense side. So I’m advising the client on what they might have to pay if they lost everything — or if they lost some of the claims, here’s where they’d be. Being able to have more precise data to guide clients in their assessment of how they want to handle a given litigation is a significant step forward for businesses to be able to assess risk, maybe not mitigate it, but make an informed decision about how far they’d like to see a case go to the extent they can control it.

We also have a lot more resources available now for getting a proxy for that human element. Nothing substitutes for the human element, but if you were to try this case to a jury, for example, what would a jury say? We do mock juries, which used to be, and still in some cases are, 10, 15 people sitting in a conference room hearing a presentation of a case. Those are great experiences and are really informative, but also expensive. Online platforms driven by AI can give you an approximation of that for a fraction of the cost and can give you real-time feedback on your trial strategy. You can even utilize something like this during trial. We’ve used what’s called a shadow jury that, in the old days, would have been a number of people sitting in the courtroom watching the trial and giving you real-time input. “Oh, this witness is credible. This witness is not credible.” Now you can do that with the benefit of AI in a more cost-effective way. Of course, there are some limitations on access to the courtroom for that, but we’re seeing a lot of improvement in those types of resources to evaluate case presentation and overall strategy as you’re going through it.

AI in Trial Presentation: Visual Evidence, Simulations, and Jury Engagement

JR: Speak a little bit about the presentation side. When you’re in the courtroom, what are the types of things you might use as demonstrative evidence or otherwise?

MM: This is an area that is fun, particularly for people who like visual presentations. We’ve developed a lot of amazing tools for demonstratives at trial. These visual presentations can be made to the court, to the jury, so they can better understand the evidence, particularly for visual learners. We’ve come a long way from the days of using PowerPoint and Visio to put together blocky text and pictures. We can now do things like simulations. For example, in auto accident cases, you’re seeing video recreations of what happened at the scene, informed by the evidence that has been admitted at that trial. It’s not something that’s been created in advance of trial, hoping it will be allowed, but something that’s been generated by what we’re seeing come into evidence at trial.

I handle a lot of cases that involve people who speak different languages, and particularly emails in different languages. We’ll often have a case where a lot of the emails are in Mandarin or they’re in German. Rather than needing to put a translator on at trial and have them provide translations, we’re able to work out a lot of this in advance through user-friendly tools that help us translate anything that needs to be translated, in a cost-effective way. There have been a lot of exciting improvements in that area over the last couple of years. 

AI in Litigation Is Here Now

JR: My takeaway from all of this is that people need to stop thinking about AI as a future state technology, especially for trial lawyers, because it’s already reshaping case strategy, discovery, settlement valuation, and courtroom practice. It’s not so much that AI is going to make litigation faster and cheaper. I think the real promise is that lawyers and clients can make better decisions earlier, maybe faster, maybe more data-informed. But of course, only if AI tools are used carefully, transparently, with great judgment and with that human oversight and thought process. 

So thank you, Mathilda, for these insights. This has been a great conversation and a continuing one, of course — because as we move forward, we are finding out how AI is impacting our practice on a daily basis.

MM: Thank you, Jen. 

JR: Like you, I’m optimistic about all the resources that AI has offered to us now and in the future. As a reminder to our listeners, Mintz has a full service litigation practice on the east and west coasts, and I assure you that we at Mintz are rapidly deploying AI in our own practices. We are undergoing daily training and using these tools and finding them fascinating. For more information about our practice group or other thought leadership, I invite you to visit us at the Insights Center at Mintz.com. Or you can follow our Mintz On Air: Practical Policies podcast on Spotify. Thanks again for joining.

Subscribe To Viewpoints