AI Voice:
AI is everywhere. It's writing social media posts and helping us pick the next binge-worthy Netflix show. You're even listening to an AI-generated voice right now. But for personal injury, and mass tort lawyers, AI has the potential to go way deeper.
This is not about the next shiny tool. It's about scaling your firm, sharpening your edge, and delivering better results for your clients, all without sacrificing your time or your team's sanity.
Chris Dreyer:
As you just heard, AI isn't coming. It's already here. The smartest law firms aren't fighting it. They're learning how to make it work for them. The top litigators like Lisa Ann Gorshe and Cara Wall are proving it can do more than summarize documents. It can reshape your firm.
Lisa brings 25 years of mass tort firepower. Cara is the tech-savvy partner, rethinking how PI cases are built. Together, they're leading the AI conversation at mass torts made perfect in Las Vegas. I caught up with them while at MTMP. They're showing us how to make it practical, ethical, and profitable. This is Personal Injury Mastermind. I'm Chris Dreyer, founder and CEO of Rankings.io. This one's essential. Let's go.
Carasusana 'Cara' Wall:
I love it for depositions, your cross-exams. If I know there's a question or a point I want to reach, but I'm struggling with how to phrase my leading question just right, I'll create a prompt that's like, "Here's what I want to get at, ways to ask this question."
Chris Dreyer:
Cara Wall is partner at Zoll & Kranz. She's part of a new generation of attorneys who grew up in tech. She's using it to level up her litigation game. She's smart, scrappy, and exactly the kind of lawyer you want in your corner.
Lisa Gorshe:
We use some AI to automate some of our processes within our office, to make our employees more efficient. I use some pieces within Westlaw, which will allow me, if I have a brief, and I need a table of contents or a table of authorities, the AI that's in the system will automate those pieces. That's low-hanging fruit.
Chris Dreyer:
Lisa is the managing attorney at Johnson // Becker. She remembers practicing law when floppy disks were cutting edge. Today, she's leading a top firm into the AI era with sharp instincts, and even sharper strategy.
Lisa Gorshe:
I will upload into publicly available sources, ChatGPT, Adobe, articles that I want summarized that, "I don't know. Do I want to spend an hour reading the 20 pages or get the summary to get the high points, and then have it linked to those main points? Be like, 'Oh yeah, this is worth my time to take an hour to read, and digest this,' or, is that overview good enough?" We have used AI with services that are able to promise us siloing of our data to give us some chronologies in specific instances. That's one of the pieces that our firm is starting to investigate more. How can we harness AI in mass torts to put chronologies together?
Chris Dreyer:
I got to admit, I am using chatGPT and Claude almost every day, where me and my wife can't figure out what we want to watch for a movie on Netflix or whatever. We're like, "Okay, let's put in some prompts, and let's see what's suggesting." We're like, "Oh, that's what we want." Then we'll talk about it.
Then also, I'll see a long YouTube video that's like an hour long. I'm like, "Okay, can you just summarize this for me, and give me the top 10 points?" I'm doing some of those things myself. It enhances our time, our abilities. It just gives us more time back, and resources. Cara, what about you?
Carasusana 'Cara' Wall:
I actually use, in the firm side of things, a lot of the same uses that Lisa was talking about. The first four-way we really made into it was on the medical chronology side. Probably about two and a half, three years now, we've been using various services to help us summarize, organize, and get to the heart of our client's medical records. It's absolutely necessary on the mass tort because of the volume of clients.
But on my single event side, for the nursing homes, those are patients who had an extensive medical history, extensive treatment history. It's not unusual to have 17,000, 20,000 pages, especially now that we have EFR records. We've had to use programs to help us get to the heart of what we're looking for. We're starting to use a new system that actually integrates both the chronology aspect. That's the big one internally in our office, is using AI for the chronologies.
Same thing, we're using the ChatGPT, the Claude, all those systems. There's a program that is an aggregator of different LLMs, all complexity. We use it a lot for just quick questions on the medical side. I work with medical consultants for big picture questions. But if I just need to know what a process is, or what is this procedure, I love being able to plug it in there, and just getting a quick overview of it.
One of my associates and I, we started referring to the program as Lexi. "We'll just go ask Lexi, like she's our co-worker when we need questions for things. Then on the mass tort side, I'm involved in the litigation that's got about 20,000 or so filed cases. It does not have a mandatory census, but we want to know what's the census, what does the lay land look like for all filed cases?
We've actually employed an AI service there that can pull the information out of medical records for the data points to be requested, and put it into a format that matches the manual database we had created years before. It's helping us get our arms around a large set of data that would be just impossible to do manually. Quite frankly, firms who are responsible for it don't have the manpower often to do that work.
Lisa's very familiar. The other part is when you're doing doc review and discovery, AI really helps to get to concepts, surface those hot documents, and get to the heart of the story. That's something, she is, especially, we have less conversations actually about what tool is best for this litigation, and what do we like here and there. Kind of a wide range of uses.
Chris Dreyer:
Yeah. You gave me a nice pass here. Lisa, let's continue on that side of the coin.
Lisa Gorshe:
With eDiscovery, I love the new platforms, what is out there that brings us beyond technology assisted research, and gets us to being able to find groups of documents within litigation that are thematic, and that are culled together. That might sprout off another sub theme that the technology is finding.
Once we have whatever that main topic is, it will find the thread that is the next most common, that relates those documents that we may not have thought about to start to go down that avenue, as well as some of the technology in sources allows you to pinpoint, and find out. This is a little bit of automation with the AI. But if you put in a particular character or employee, and find out who are they communicating to versus who are they getting communications from, that can be very telling. Where you think the hierarchy is one way, but then you don't see that in reality because their emails aren't going to ABC. They're always going to XYZ is their answer.
Chris Dreyer:
Yeah. When I'm using these different LLMs, for example, ChatGPT, I call it my lazy approach. It's the shortest prompt that it somehow uses my brain, and knows what I want, or I'll just upload a picture, and it will summarize it really nicely. But if I try to do that on Claude, it's just a big fail.
But I feel like sometimes Claude, if I spend a little bit more time with the prompt, the abilities that it has is far superior than ChatGPT. I remember recently I was asking a question about addressable market, and Claude created an algorithm to analyze these different things, and ChatGPT was very top level. How are you guys really utilizing, and entering the prompts? Are you just feeding the data in Excel sheets? Are you taking pictures? Tell me about the usage. Let's start with Cara.
Carasusana 'Cara' Wall:
My approach to the programs is actually kind of thinking of each LLM as an employee or coworker. Each one has its own skillset that it's good at. You have to explain a task to people in different ways depending on their abilities. The prompts that I might make for ChatGPT might be something I might give to an entry level or an assistant in my office who just needs kind of very basic, very high level work. Versus something like Claude or DeepSeek, it might be something like my associate who I need a deep dive in, and I give very different instructions to.
When I'm going for the more detailed, if I'm trying to create something with a very directed result from the prompts, I think about the kind of instructions I would give an employee. What's the scope? What output am I looking for? What approach do I want, and have it taken? That's how I build out those prompts to try to get as close to the end result that I'm looking for.
Chris Dreyer:
I am probably going to steal that, and probably name my different LLMs because that's how I think about it. The thing that's frustrating for me is when I don't delete or archive the history, it's like biases everything. I was joking with my team. I was like, "Look, not everything that I enter is, 'oh, that's a great idea, ChatGPT.'"
But please challenge me. Let's give it a persona to challenge, and scrutinize my information. Lisa, what do you see? Do you have a similar methodology of naming and the different skill sets? How do you go about in approaching which LLM you're going to use?
Lisa Gorshe:
I've not thought of naming, but now that Cara said that, I want to steal that, and use a different name for the different sources that are out there. But I will look at what is [inaudible 00:10:58] I am very intentional, and I don't use as many sources as Cara does. Because I'm still old, I'm still afraid, and very worried about my client's privacy. I will look at what is it that I absolutely need, and what can I put into there with the barest of minimum client-sensitive data to get me on my jumping off point? I will weigh, what is it that I need to put into the system to get my answer?
Our office, I was going to say next week, is going to be in a sandbox for one of the sources that's out there. I've gone into the sandbox myself, but I'm opening up to the rest of our firm to go into that sandbox to learn the tools to determine on a wide level, is this going to be advance the ball for us so that we can all work a little smarter, and not as hard to shave off some time so we're being more effective.
Chris Dreyer:
Perplexity is excellent, because it cites everything. It doesn't just garble everything together. Kind of, "Here's the sources, but good luck on trying to determine which is which." I want to stick in with you, Lisa. You got a ton of trial experience. What about on the trial side? Have you seen the trial attorneys at the firm go as far as, "Here's everything about the case, ChatGPT," as much as you can anonymize, and keep it confidential, what's the defense tactics? Is it that advanced? Can it craft a narrative? Look, I'm not a trial attorney. I think there's different tactics, and approaches to cases. Does it go that far? Does it have that ability?
Lisa Gorshe:
I'll be honest. I'm the skeptic when it comes to that. I don't know if it's there yet. I think you can use it to target a question or frame an argument to help you with that piece, and maybe determine during a trial between the two different docs you may want to use, and which one is going to present better based on the answers that have been given to date during a person's testimony. But I don't know.
AI cannot replace a human's gut and intuition. So much at trial is your gut instincts, where you're going to go. I don't think AI is there yet to be able to help when you're in that split decision making process. Can it help you prepare your opening argument or closing statements? Yes, but during the trial, there's too much thinking on it, just on your toes. That's really what your client is paying for, is that knowledge, experience, and being able to use your gut and your emotions.
Chris Dreyer:
Hey, folks. Chris Dreyer here. Last year at PIMCON, we set the tone. This year, we're setting the standard. We're heading back to the five-star Phoenician Resort in Scottsdale, Arizona, October 5th through 8th to take your personal injury firm from good to goat again.
PIMCON isn't just another legal conference. It's a high impact experience built specifically for personal injury attorneys who want to scale smarter, connect deeper, and get real results. No fluff, no filler, just proven and actionable strategies from brilliant attorneys that actually move the needle.
General admission gets you the goods. Full access to all speaker sessions, entertainment, and networking with some of the most respected names in plaintiff law. VIP goes even further with luxury accommodations included, exclusive offerings, a white glove concierge, and more. A fair warning, VIP is very limited, and they'll go fast.
Ticket prices go up as tiers sell out. Now's the time to lock in your spot. Get your tickets at PIMCON, that's P-I-M-C-O-N dot O-R-G, or visit the links in the show notes. The strongest firms show up at PIMCON. The rest, they wonder what happened. Let's go from good to goat together. See you this fall in Scottsdale.
Carasusana 'Cara' Wall:
I echo a lot of that. I've actually got four trials between May 12th, and the end of July that I'm simultaneously doing now. I'm with Lisa. The on your feet, actual lawyering abilities part of it, it will never replace. Where it really is a fundamentally helpful tool is breaking through your writer's block, and breaking out of the own box that you're thinking can keep you fit.
I love it for depositions or cross exams. If I know there's a question or a point I want to reach, but I'm struggling with how to phrase my leading question just right, I'll create a prompt that's like, "Here's what I want to get at, ways to ask this question." Then I can Frankenstein from the options given to me what I ultimately want to ask. It's good for crafting questions.
Then it's also great for helping with metaphors and analogies. We do that constantly with our openings, and our closings when we're dealing with experts. Often, the metaphors and analogies we can come with are always limited by our own experience with the world. That's how we see things as we frame it. The AI is really good at breaking past that, and getting input that you would never occur to you.
Because I don't watch NASCAR. I wouldn't have thought about a racing analogy for something. But it could be exactly what I need in a moment. I really love that for challenging my thinking, and getting me out of where I might be stuck in something.
Chris Dreyer:
It's funny. I give my prompts, personas, and how this individual would analyze a particular thing. I'll use say Carmack, which was the CTO at Meta. He's an engineer. He uses first principles, and everything is super logical. Then I'll use somebody else when I want a creative. Then if I'm having fun, I'll maybe have Samuel Jackson give me some feedback, just depending on my mood. But yeah, super interesting.
Also, I want to send a follow-up question, Cara, you first, and then Lisa. Talk to me about your entries. Okay, I am getting there to the point where it's almost scary where I'm talking into the phone. It's like a super weird feeling when you get the back and forth, especially if you get the audio, read it back to you. But I'm wondering just how you're entering into the LLMs. Is it mostly text-based prompts? Are you speaking into it? Talk to me about the entries.
Carasusana 'Cara' Wall:
Yeah, it is, I would say, overwhelmingly text-based prompts for me. Unless I'm uploading something I want analyzed, then I'll upload documents. Occasionally a picture, oftentimes it is an article, or a piece of writing on the internet, but usually it's me. It might be my control freak nature that I want to be able to say exactly what I want, and so I'm entering via text.
Chris Dreyer:
Yeah, that's great. Then sometimes if you don't delete the history, you use those prompts in those other individuals, it's like, "Oh, wait, I need to shut that off, and change things."
Lisa Gorshe:
I am almost exclusive with texting, and typing it all in, what is my query, to control what it is. Unless it's something personal for myself. If I'm doing something that's personal, that I just want something off the cuff, or I'm just using the Google or Microsoft Co-pilot, the AI in that, because it's just general, then I might do an audio question. But it is always, I am very intentional with how I'm crafting the prompt.
Chris Dreyer:
This has been a really fun conversation. I can see the uses in the legal space, how it enhances your capabilities, and enhances capacity. I guess one final question for each of you is, where do you see this going for the future in adoption for the legal space? I'll start with you, Lisa.
Lisa Gorshe:
For me, I don't think AI is going to go away. It's going to continue to grow, but it's a tool within the toolbox. It's never going to replace a lawyer. It might replace pieces of what a staff member can do for you in order to support you, and make you a better lawyer, but it's not going to replace lawyers out there.
Carasusana 'Cara' Wall:
I agree. I think that's one of the philosophies, at least, I really share at the heart of it is, there will always be a need for lawyers. You just can't tell a robot to take over the case.
I really think of it as akin to the change from books to electronic case law, in terms of it being a tool that if you don't adapt and develop, and at least be able to understand it, you're going to be doing a disservice to your clients. Because you won't be lawyering at your best ability. But making the switch from a book to electronic case law doesn't take out thinking, it didn't do the work for you, it made it easier to access things, but ultimately, the human oversight and intelligence is still there.
Chris Dreyer:
Big thanks to Lisa Gorshe and Cara Wall for pulling back the curtain on how AI is really being used in mass torts, and what it means for the future of the legal practice. Whether you're managing 20,000 files or writing your next opening statement, the message is clear. AI isn't replacing lawyers, but it's giving you the edge. If you got value from today's episode, do me a favor, share with a colleague, leave a review, and make sure you're subscribed. This is Personal Injury Mastermind. I'm Chris Dreyer. Catch you next time.
