Generative AI: Understanding it from a legal perspective

Mallory Hendry [00:00:07] Hello everyone and thanks for joining us today. I'm Mallory Hendry, Content Specialist Manager with Canadian Lawyer Magazine and I'm pleased to introduce today's webinar, Generative AI understanding it from a legal perspective. The speakers we have with us today are Zach Warren, Manager of Enterprise Content, Technology and Innovation at the Thomson Reuters Institute. And Zena Applebaum, Vice President of Product Marketing at Thomson Reuters. Over the next hour, these two industry experts will provide comprehensive coverage of what law experts need to know about this technology and how to best leverage it in your day to day work. At the end of the presentation, our experts will participate in a question and answer period. So be sure to type any questions that you may have into the q&a box within the webinar software. I'm going to turn things over now to our panelists to begin the presentation. Take it away, Zach.

Zach Warren [00:00:58] Thank you very much Mallory. And hey there everybody. As she mentioned, my name is Zach Warren. I'm the Technology and Innovation insights lead with the Thomson Reuters Institute. And I'd like to welcome you to today's webinar, where we talk about a very exciting subject, one that I think a lot of people are chatting about, but may not know the nitty gritty of legal AI. There's a lot that's happening in the space, think back just even a year and a half ago. That was when chatGPT became public for the first time and a lot of people heard about what this generative AI thing could be. Over the past year and a half, there have been a lot of improvements, not only on the technology, but how people in law are applying it to their daily lives. And what we really want to do here today is go through a little bit of our research into how generative AI is being used in the legal space, but also talk about the practical. What does it mean for you either as a practitioner or a legal professional in your own career and in your own work? So as Mallory mentioned, I'm Zach Warren, I'm with the Thomson Reuters Institute. So what the Thomson Reuters Institute is, it is the research and thought leadership arm of the Thomson Reuters company, the company behind Westlaw, practical law, etc, which Zena will go into a little bit. In particular, what I'm going to be talking about today is a report that actually isn't released yet. It is a report that will be coming out next week about generative AI in professional services, including legal but also you'll see mentions of tax and government risk and fraud, compare and contrast what is happening in the legal industry, to other professional services out there. I've been tracking this stuff about 10 plus years myself, I used to work on the journalism side of things with ALM legal tech news. But since bout a year and a half or so I've really been neck deep in this research, Zena, I'll have you introduce yourself and what you're going to be talking about a little bit as well, if you could.

Zena Applebaum [00:03:02] Absolutely. So thanks, everyone for joining us and watching everybody come in. It's great to see you all, or see you all, of course. So Zena Applebaum. I'm the Vice President of Product Marketing here at Thomson Reuters. And I focus primarily on our research product. But as you'll see, as we go through this presentation, the the world between legal content and legal technology is merging at a rapid clip. And so you're gonna start to see lots more integrations, not just with Thomson Reuters products, but also with other products throughout the legal ecosystem. Today, I'm going to talk a little bit following Zach, I'm going to talk a little bit about how we actually see this work coming to fruition. And what tactical applications of generative AI exist in the market today? And what are some of the more visionary, aspirational things that are coming in the short term and in the longer term as well?

Zach Warren [00:03:52] Thank you. And as Mallory mentioned up top as well, we're going to be keeping an eye on the q&a. So if you have any questions, go ahead and pop those in. We'll try and answer throughout the presentation as well. But if there's anything we don't get to, we will leave some time for q&a at the end of the presentation as well. So with that in mind, let's talk a little bit about some of the research. So this report that I'm talking about, that's going to be coming out next week. Really, there's a few high level things and we wanted to touch on. We wanted to touch of course on the usage, who's using it, how where, but we also wanted to talk a little bit about the perceptions. How are people thinking about this? How are they approaching it? Do they think this can be used? And do they think this should be used in law? We're also going to touch on not only the law firm aspect of this, but the corporate aspect of this, and in particular, how those two interplay, what do clients want to see from their law firms? What does it mean for the future of business? And then we're going to ask, so what does this mean just in the big For the future of work as well, it's important to note for this research up top, how exactly we getting the answers that we did. So primarily, we surveyed English speaking countries. So the US is the primary one, but you'll see 16% of our answers do come from Canada, primarily our customer segments, law firms are who we got the most answers from, but a decent amount of corporate legal, and you'll see courts in there as well. And then, as a result, legal is the primary segment that we got our answers from. So just something to keep in the back of your mind. A lot of these data points that I'll be showing, I'll highlight where law firms are different from corporate legal or different from courts, etc. I don't have too much broken out Canada specific in this. But if you have any questions for me, particularly after the public release of the report next week, just shoot me an email or let me know. And I'm always happy to answer any questions. So with that in mind, let's start with a poll question a little bit because I like to level set. So I just want to know from people who are on this call, have you used generative AI before? And yes, for work? Yes, I've played around with it on my own time. No, I haven't. And the final one, what's generative AI? And I'm curious, while we're getting answers in Zena, are you still hearing a lot of what's generative AI? Or do you think people kind of know what this is now?

Zena Applebaum [00:06:32] So you do sometimes get what's generative AI, but then they've heard of ChatGPT. So sometimes the connection between generative AI and ChatGPT or large language models, and I think one of the things that people struggle with is that there's so much different, so many different ways of approaching the topic and so many different nomenclatures, you can apply whether it's large language model, generative AI, some people are just saying AI generally. Which is a bit of a misnomer, because what we've been talking about since March of 23 is actually specifically generative AI. So, yes, we do still sometimes get what is general AI, as per our survey? 

Zach Warren [00:07:09] Yeah. And we're getting the results up here. And these results are actually pretty reflective of what I'm seeing in the market and some of the data that I'm going to be going over, which is very interesting. About a quarter 26% saying, yes, I've used it for work a little bit higher. 29% saying yes, I've used it on my own time. 40% said no, have not used this yet. And we did get about 4% saying yes. What is generative AI just looking for more information, Zena any responses to these particular poll results? Do they ring true to what you've seen? 

Zena Applebaum [00:07:46] They definitely ring true. And I think depending on where you work in the industry, what part of the ecosystem, you're a part of these answers will change what you answered today, and certainly over time, you're gonna see more and more people eking out of that no category into the Yes, category.

Zach Warren [00:08:07] So let's dig into some of the data that we found in our report. And one of the first things that we asked is two related but different questions. The first one we asked, Can generative AI be applied to your work? Basically, do you see a use case in your particular industry or daily work life? Then we took a little bit of a spin on that and said, Should generative AI be applied to your work? Not only do you see a use case, but do you think there is an actual application that should be in your daily work? And in most cases people said yes. You see over there on the left hand side, legal actually had the highest proportion of people seeing use cases for generative AI in their work. 85% said, Yes, we think this can be applied to our work. And additional 11%, as you can see in legal said don't know, only 4% said no. But then when we ask the should question, the percentages still pretty high? No, I did not mean to go advance slide yet. So you see 55% said yes, this not only can be applied to my work in legal, but it should be applied to my work. But as you can see, if you didn't answer yes, it wasn't necessarily a know either. 29% So they did not know which leaves 16% in legal, only 16% saying this should not be applied to my work. Now, I didn't put any of the free response answers here as to why or why not. We're gonna get into some of the barriers more a little bit later. But this is just kind of a snapshot of showing people think that there's a there there, even though we're only a year and a half into it. LM generative AI really being in the public eye, already people are starting to see the potential use within legal. And are people using it? Yes and no. So we asked this in two different ways we asked, okay, so the public tools like your ChatGPT, maybe your Microsoft co pilot, other sorts of public facing tools. Have you played around with those? And about a quarter, which is why exactly 26%, which is what we saw in the Canadian lawyer poll here on the webcast said, Yeah, we're already using this. And additional 21% said, either within the next year or the next three years, we think that this is going to be something that we're regularly using for our own personal use law firms and corporate legal, both right in line with those particular things. So we said, Okay, so that's like Zena was saying a lot of people think about generative AI, they're thinking about Chad GPT. What about legal specific tools? When there we asked right now, a little bit lower only 11% said, Yes, we use for our own personal use either a legal software or some sort of legal specific generative AI tool. But as you can see the potential growth there within the next year to three years, a lot of people anticipate using these sorts of tools. As you can see, 34% in total said, Yes, we think that legal, specific generative AI solutions are the future. So adding up everybody who's using now to within the next three years, that's about even and in both cases, whether it's your ChatGPT’s or public tools, or your legal specific tools here, about half expect to be using this within the next three years, which tells me that people are viewing this not necessarily as some futuristic technology, but something that is already here, and something that people are already planning for. So we adjusted the question slightly. Okay, we asked about personal use, are you using it? How about on an organizational level? And that is where we start to see people lag behind a little bit. So we asked, Is your organization integrating generative AI into what they do? And by and large, people said, Not yet, but they're considering. So as you can see, for law firms for the legal industry, in particular, about 14% said they're already using on an organizational level, about 20 12%, excuse me, said they're planning to use, but then the rest are split, considering whether or not to use or having no plans. So what this tells me comparing to the previous slides is a lot of people are playing around with this on a personal level, they're thinking how is it going to affect my work, but organizations as a whole still have a little bit of work to do to think about how to integrate these tools, either firm wide or company wide department wide into what they do on a daily basis. So how exactly are they using it? Well, it's a very wide variety of things, which I think speaks to earlier, a lot of people are seeing the use cases here. Legal Research at this is the representative population is people who are either using right now or actively planning to use so in the integration process as it is right now. And of those that have actively played around with the tools 80% said, Yeah, we're using for legal research 75% document review. You have document summarization, contract drafting briefs, and memo drafting is the over on the right hand side that table, particularly in law firms, more than half using for marketing, some back office tools. There is a wide variety of use cases that you can use generative AI, which I think is part of the reason that a lot of people are adopting this technology quicker than really you've seen a lot of legal oriented technologies be adopted in the past. You think about things like the cloud and how not slowly, that law picked up the cloud and moved a lot of applications to the cloud. But there is some people took some convincing to say, oh, I can see where this is going to fit into my daily work. Generative AI, I don't think there's as much convincing I think people are seeing pretty clearly what exactly the use cases for this technology should be. So with that in mind, I just kind of want to get another poll out there. Get a sense check of how people feel up. Out of this, because the very next question that we asked in our survey is essentially the sentiments that you now see on the screen, which best categorizes how you're feeling about the future of generative AI and law. Is it excited, hopeful, resilient, concern fearful? And Zena, when you're talking with a lot of clients, which one of these do you think really stands out?

Zena Applebaum [00:15:26] I think cautiously optimistic is the way most customers and most most people in the industry are feeling today. So a little bit of concern, but mostly excited.

Zach Warren [00:15:38] A little bit of concern, but mostly excited. I think that sounds about right to me. Let's go ahead and see the poll results as they're popping up here. And we're getting a lot of pot positive sentiment. So 40%, saying hopeful 20% saying excited, seeing some hesitant 22% 15% concern 3% fearful, one said none of these. So a lot of positive sentiment here, which I find particularly interesting. I guess, if you're listening to this webinar, probably you're here because you kind of want to know what the future looks like. But I will say that these positive sentiments are more than we saw when we surveyed the industry at large. So when we asked that question in general, hesitant was number one, about a third of people said they were hesitant, even higher in a law firm setting, as you can see highlighted there 37%. But hesitant was kind of the neutral option. If there were comparing the positive to the negative, hopeful and excited, definitely higher than concerned or fearful. And I think that that's something that we just saw in our poll here as well. So Zena, I'm kind of curious, your thoughts, both the poll that we just saw, as well as the numbers on the screen? Do those ring true to life for you.

Zena Applebaum [00:17:00] Think so I think that's fairly accurate to say that people are more positive than negative. And I think hesitant is probably a good word to describe. It's not that you're not going to use it. You're just going to use it with caution.

Zach Warren [00:17:17] Sir. So I think this is before I move too far away from some of the numbers that I just put up there. There's a few questions in the q&a that I want to address real quick in terms of the research. So one of the questions is do you test your responses from legal specific tools against open source or subject agnostic tools? So when we ask the question, the only distinction we made is public versus paid for proprietary. So we put a lot of the open source in that public option. We put ChatGPT, probably is the one that most people talked about. But when we asked the question we not did not distinguish beyond that. And I think it's Zena. Did you have anything further zoom just gave me okay. Google zoom was just giving the little note there, but I think it might have just been zoom. And then the other question is about how we source the government court. So one thing I will note is just in terms of sample size, law firms, corporate legal departments are much higher than government courts. So the government court numbers a little bit wonky er, than maybe some of the other numbers here specifically from law firms and corporate legal departments. That's why I'm trying to highlight the law firms as much as possible. But we did get court respondents from really across the board, Canada, US UK, a lot of them did tend to be judges, barristers, etc. We did get a few from court personnel, clerks, etc, as well. So yeah, when you see courts, that's where that is coming from usually lower sample size, but primarily judges or somebody on that level. So, with that in mind, let's talk a little bit about some of the barriers. So we talked about this earlier and where is some of that hesitance coming from and a lot of the hesitance are things that especially if you hear about ChatGPT in the news, it's usually something related to this, the hallucinations, the potential for inaccurate responses, some concerns about data security, data privacy, number five there, you got a little bit cut off, but that one is ensuring that generative AI tools are ethically used. So a lot of these at least me personally, I think they do stem from people playing around with ChatGPT trying to figure out these tools for the first time. but not necessarily some of the legal specific tools that I'm not going to say they've solved these problems, because there's a lot in the technology that can't be solved. But a lot of the legal specific tools have been purpose built to try and address some of these concerns, specifically around privacy or around data security, potential for hallucinations, and Zena is going to talk about that in a bit as well. So I talked up top about Okay, so we asked law firms questions, we asked corporate legal departments questions, then we asked clients questions about their firms. So the next couple of slides are going to be about that. So one thing we asked is, okay, corporate legal departments. Do you think your vendor firms should use generative AI? And by and large, they said, Yes, we do. As you can see, 58% of corporate legal departments actively think their vendor firm should use generative AI? And similar to what we saw earlier? If it wasn't a yes, it wasn't necessarily a no, either. A decent amount said don't know, only 15% of corporate legal clients said their vendor firms should not use generative AI. Alright, so that begs the question, what are they going to do about it? Well, first of all, they don't know whether their firms are actually using generative AI. We asked. So when we asked that question, only 16% of corporate legal only 12% of all total legal clients said, Yeah, we know that the firm's we're working with are using generative ai 67% said don't know, or there was particularly on the court or government side, a substantial amount that don't actually work with law firms. By and large, a lot of clients and a lot of firms have not had this conversation quite yet. And part of the reason is, nothing's really forcing that conversation. We ask clients as well, are you actually including this in RFPs? And no, by and large, especially on the corporate side, they're not providing any direction really, around the use of generative AI. So if you're somebody in a law firm listening to this, you're probably thinking, Oh, well, how exactly? Are we going to get answers? Should we be the one to bring up these questions with our clients? And personally, my answer is yes. Because as we saw earlier, a lot of people think within the next year, three years, half of people think that this is going to be used on a daily basis. So it's probably good to begin that conversation now with clients to figure out, okay, we're using this personally. What does that mean for our engagements? How do you want us to be approaching this? And that's particularly key because a lot of people aren't even providing training internally for how to approach this. We ask the question. So you whether you're in a law firm, corporate legal, we also asked the same of tax risk and fraud and government. Are you providing training or education about this across the board? And by and large people are not? You see, they're the law firms highlighted in that table in the right, only 18% of law firm respondents said their firms are providing education or training. 77% said no, and 5%. And no. So what that tells me is there's a hole here. And there's especially if you're somebody who is hopeful, excited, like our poll revealed, probably you are in a good position to try and influence some of this training some of this education to make sure that you're taking advantage of what generative AI has to give. And what exactly is that that they have to give? Well, I'm going to turn that over to Zena for the next part to talk about some of the nuts and bolts of generative AI.

Zena Applebaum [00:24:15] Exact and hopefully you all found that a little bit interesting in terms of level setting what the industry is actually thinking. And I'd be curious to hear from some of you, obviously, I can't hear but I'd be curious to know what your reaction is to this idea that corpora corporate counsel still don't entirely know how they want their firms to be using it. But they do know that they want their firms to be using it. And I did see a question in the q&a around whether or not we queried law firms as to whether or not they would pay sorry, corporate counsel us whether or not they would pay more, if they're law firms were using generative AI to enhance their offering. It's a great question and one that we can definitely think about talking to putting into our surveys for Later, I think it's clear that the world is changing and that the legal market is changing alongside the world, I think, ChatGPT became a technology and large language models became a technology that hit the market really hard as some of you are aware. The ChatGPT hit a million users in five days. And it took something like Netflix three and a half years to get there. So the groundswell of use of this technology in everyday personal lives took over such that you end up in a situation in the legal market, where for the first time, you have partners asking about a technology at the same time that you have associates or summer students coming into the environment, asking you about a technology. And it's rare that we see that kind of adoption, and that kind of interest across the entire ecosystem of the firm. We're seeing, you know, people in their, in their marketing and business development departments in their research departments, everybody's really asking how would they can be using these tools, we find in the corporate counsel side, that across the C suite of organizations, they're being asked, and they're being challenged with, how are you going to use this technology to help mitigate costs as well. And we hear, we'll hear GCS coming to us and saying that they've been challenged, their CFOs have been challenged. Their chief human resource departments, or their chief people, officers have all been put to task to say, what are you going to do? And how are you going to use generative AI. So let's talk a little bit about what's happening in the legal market, we can go to the next slide. So what we see happening in the legal market, and I'd say we're squarely in the middle of wave one right now is that we will figure out what the technology does, how the technology can be used, what those use cases are, for all the various types of law that gets practiced and the various workflows that happen within a law firm or a legal department context, or a government department context. And then we will start to optimize those workflows. Think back 20 years ago, 25 years ago to the way ediscovery took place in the way ediscovery takes place. Now, it took us about 2025 years to get there, I don't think it's going to be another 20 years, I think this is going to be a much, much more truncated spectrum of how quickly workflows change. But workflows will change. Once workflows have changed, we're going to move into work wave to where you're actually going to start to see an industry that has fundamentally changed in the way they do things has fundamentally changed. And it may mean that there's reengineering of the firms, it may mean that there are new commercial offerings or new commercial arrangements between buyers and sellers of legal services. And then we're slowly going to move into wave three, where we're going to start to see AI winners emerge. What I mean by that is, you're going to start to see organizations, whether in house counsel, government departments, which is a great place to see these things, or nonprofits like legal aid clinics, or even law firms where you're going to start to see the service delivery model and the pricing for those services fundamentally change. And when that happens, you're going to start to see agentic illegal AI models deployed into the market, you're going to start to see a very dynamic and competitive and fast moving market, such that the legal market hasn't been really in the same way. Until now you're gonna see something more similar to maybe some of the way some of the other professional services have acted over the last several years. When I show this slide to people, and I've been doing a lot of these road shows across the US across Canada, shout out to anybody in Calgary who's going to be there in a couple of weeks, please join us. What's gonna happen is, some people say, yeah, it's gonna take 10 years other people say this is gonna happen in five. And so I think there's a broad spectrum of understanding about how long this is going to take to happen. But what people are not disputing is that we are at the beginning of a change. And I would actually argue that we're past. We're past the beginning. We're sort of in the beginning of the beginning, but we're past the beginning stages, we can go to the next slide. We do believe and polls that we've seen, this is not one of ours. This is one from Wolters Kluwer. But we do believe that the effective use of generative AI will separate successful and unsuccessful law firms over the course of the next five years. And you saw that in the previous slide where you're gonna start to see these winners emerge. What I what I would caution you to say here is it's not that if you don't use generative AI, you won't be successful. It's that you will be more successful if you do. We see that General Counsel are saying we want our firms using this. We know that government agencies want their firms using it because it will help make things faster and more efficient. And we also know that the technology is moving so quickly, that it will be harder for firms who are not engaging at the beginning of the beginning where we are Today, if you're not engaging, if you don't have some roadmap, if you don't have some AI committee setup, if you're kind of turning a blind eye and being, I want to be a fast follower, and I want to wait and see what everybody else does, before I jump in, which is, tends to be typical in the legal industry, the farther you fall behind, the harder it will be to catch up because the technology and the products available in the market. And the urgency in the market is changing so rapidly, that the farther you fall behind, it'll be harder for you to catch up, because the adoption will take longer, for example. So all this is to say, the next five years, this is going to have significant impact. So what are the things that can actually be impacted or replaced by generative AI, all the skills that you see on the left hand side or on my left hand side of the screen, around drafting document assembly, analyzing things at scale tasks at scale, these are all things that computers do really, really well. And we saw this with, again with respect to ediscovery. But you can say the same thing with respect to even just document production today versus 30 years ago, associates come in today, they type their own documents, they're not handing it off to a legal assistant to type their documents for them, they're not dictating anymore or not, for the most part. So many of the things on that side of the screen tasks that sale, drafting communication, things that need to be done to deadlines and pull out those deadlines, reviewing documents and reviewing large, large datasets, that is something that AI can do really well can do it really efficiently and can do it with a smaller margin of error than when you leave it to the human. On the other side of the screen. You have things like attention to detail, commercial acumen, experience judgment, these are the things that corporate counsel are coming to you to actually pay you to do. Because these are things they want you to be able to do. This is the things technically these are the things that I hear people say I went to law school to do. These are the things that we cannot yet replicate Critical Thinking abstract thinking, analogous thinking in a litigation context. So what happened in a similar case with similar fact patterns? How can I apply that to my case, these are things that for the moment, the generative AI is not as good at doing because they do need to generate something, it doesn't need to produce something. And so the thing is, on the left are the things where we're actually going to produce something and what's not there is also research, although arguably reviewing can be part of research. So if we go to the next slide, we can talk about how this is actually manifesting. We know from Zach slides, that there's a lot of concern in the market. We know that there's a lot of issues around trust, we've seen both in the US and in Canada, most recently in British Columbia, where people are lawyers are putting memos or other documents in front of the court that are using faulty citations. I would say that this is not a problem of the technology. This is just bad lawyering. But what if I could take that all away? What if I could take all your concerns away, and I could give you a legal assistant that was like by your side riding shotgun with you that was reliable, trusted, spoke plainly. So you don't need to think about legalese. They were by your side, wherever and whenever you needed to work, whether you're doing research or drafting, whether you were just thinking out loud and wanted somebody to bounce an idea off of an assistant that you could delegate. And I would add here into this that you edit an assistant that you could delegate substantive legal tasks to. And that's where it's a little bit different than some of the other platforms. We're not talking ChatGPT, we're talking an assistant that you can delegate substantive legal tasks to, and help deliver results faster and more efficiently. If I could give you this in a box in a way that you could trust it would some of the hesitancy that we saw earlier disappear.

Zena Applebaum [00:34:00] We go to the next slide. So we believe that the answer is in delivering that platform, using a combination of the authoritative content that you're used to trusting Thomson Reuters for that is grounded in facts. So our AI is not going out to all of the information ever produced on the internet. And all of the queries that everybody in the world is putting into ChatGPT that's being collected as part of the corpus of data. We want to ground it in authoritative content. We want that content to be produced by domain experts, whether in case law or transactional work, we further want to refine the ability to search that case law and to search that authoritative content by leveraging data scientists and AI software engineers, many of whom are at the highest levels of education to back that up, so it's not just domain expertise and subject matter experts on the legal side. It's also that on the technology side, and then we want to connect these things and provide you with the broadest spectrum of legal workflow connecting subject matter expertise, as well as World Class AI talent in a way that is secure, maintains your ethics and maintains governance across the entire workflow. I will call out to Chris Sharma, one of our heads of product here at Thomson Reuters, who is all about ethical AI, and she's a huge champion for it. I encourage any of you that are unfamiliar with her and maybe Zach, you can type into the chat or you can somehow get people out to know about Kurt about Kriti. And follow her because she will talk about the things that we are doing to keep our AI ethical, we can move on to the next page. We're going to bring you that assistant. And our vision is to really bring you that assistant using all of those levers that I just showed you doubling down in trust, to give you the most robust set of trusted generative AI skills that are available in the market today. And we continue to grow. And so that is CO counsel co counsel is your little shotgun, legal assistant that rides shotgun with you, and is with you every step of the way. Go to the next slide. So what is CO counsel counsel is the most comprehensive, comprehensive set of skills available today. And today, you can use CO counsel to do things like review documents, as we talked about, summarize, documents search a database, you can create a timeline or chronology in the litigation context. So this is a thing. This is a skill that we know, many young litigators or other people in the in the workflow will be tasked with taking case files, and figuring out all of the different events that happened in a chronology. To do that, you have to read through the documents, pull out the salient dates, and then create this technology. And it can often take several days to read through everything and create the technology. With co counsel, you can do it really quickly because it can read documents faster. And it can look for the markers of what makes it a chronological event and added, you can prepare for discovery, you can draft correspondence. And then other things that are coming are around, you know, analyzing red lines, verifying a brief things like this. And so these are things that you can do today with CO counsel. And CO counsel just kind of lives on your desktop, and it's there and it's available, and you can feed it information, and it will take whatever task you're looking for. And it does it in plain language. That's one of the most beautiful things about it. There's not a lot of on ramp time. Because the beautiful thing about chat functionality or large language model functionality is that the idea is that it's about a chat interface. You can talk and ask questions as though you had a colleague in the room with you. And by chat, I mean type, of course, but you could actually use voice commands as well, and it would work. The idea here is that you ask it a question the way you would ask a question of a friend and it will give you that answer. I actually don't know for sure voice commands would work. So don't quote me on that one. But maybe it's a future enhancement

Zach and Zena [00:38:04] I think that would work. I think so I

Zena Applebaum [00:38:06] Would think so right? Let's move to the next slide. So what our customers telling us. And I will point out here that we have small law firm associates, we have global large law partners. We have solo practitioners, they love this capability. They love the ability to have this little co counsel writing with them that can answer their questions and do substantive legal work for them. It doesn't replace the work that they do. But it augments it in such a way that it makes it faster to move to the next step. One of the questions we often get asked is, well, this is how we train associates. If we take the associates out of the if we take the grunt work out, how do the associates learn. So a couple of different things. One, you can reverse engineer the work product and help the associates learn that way. The other thing is, the tolerance for doing this kind of work is shrinking over time as technology becomes better. And the other thing I would talk about is that you know, you get to a point where an associate can have a more substantive legal conversation earlier in the process, because they don't have to worry so much about getting the work done quickly. And effectively. You can get a first draft or a first version of an answer. And I call it that specifically, it is a first version, you would never say that this is a polished version that you would send off to court. But it's the first version of things that you can get to really quickly to start to have the next questions. And to start to dig into more of the substantive discussion around the legal issue that you're exploring. We can go to the next slide. So what are we doing in Westlaw? So we've seen that you can do a whole bunch of technical things with respect to co counsel, but what are we doing with Westlaw edge with Westlaw edge what we're doing we can go to the next slide is we're grounding the discovery process against the Westlaw Edge database So within the Westlaw Edge database, you have things like Keysight, you have things like our key numbering system, things that are unique to Westlaw, you can use that Westlaw unique proprietary taxonomy that is underwriting Westlaw, what will happen is you'll ask a question, it'll first go through its regular discovery process. And instead of bringing you back a whole list of citations, it sends those citations over to the large language model, the large language model will then use those citations to create an AI assisted answer. By doing this, you've grounded your search criteria into only that trusted content that we talked about a few slides ago. And it allows for significantly higher results that you know, you can trust. Because you know that it's only being based on Westlaw content that you know, and trust today. So you're not, for example, going to get a citation coming back from 2026. Because 2026 Hasn't happened yet. We do see examples out in the market where things like that have happened in other platforms, it won't happen in Westlaw. That particular example, we can move on from the retrieval augmented generation slide. But I will say that that's not a proprietary technology. What is proprietary is the grounding and Westlaw and all of the editorial enhancements that go along with Westlaw, that are unique to Westlaw, that's something that you're not going to find anywhere else. And when you start with better content, you're going to get better results. Similar with practical law, we're moving into more transactional content, although there is litigation content and practical as well. And it too, is going to be grounded in a similar retrieval augmented generation technology, whereby we're only searching practical content to bring you back your practical law answers and we can move to the next slide, you can see an example of what this looks like. So you asked practical law question. The question there is about drafting an executive employment agreement for a CEO in a particular jurisdiction, it gives you an answer for how to do that. And it cites exactly where it got you that answer from as well. And in the Westlaw example, and I'm sorry, I didn't put up a screenshot for Westlaw. And again, these are the US contexts these products have not been released in Canada yet, they will be soon. But it gives you your citation. So you're not guessing you're not asking where it came from. And you can see the blue hyperlink in the middle of the screen there, you could click on that, and it will take you directly to where something was found in practical law. And the answer will be similar, it'll look very similar when you are in Westlaw. When you click on it, it will take you directly to the citation. And you'll have footnotes, you click on the footnote, it will open up a summary for you right in the interface. We can go to the next slide. Again, practical law, people are very excited about this, both in terms of in house, particularly when you're a solo in house, the ability to search faster and get to the content faster is always easier. And then again, on the law firm side. You know, as an associate, you're thrown into a lot of different things, you're expected to know a lot of different things and get up to speed quickly practical off course helps you get there really quickly and practical law, ask practical law or generative AI within practical law helps you get there even faster. So we can go to the next slide. So one of the places that people find that generative AI lends itself really easily to is in the drafting context. So you have here on screen is the sale of goods agreement, you have another screen that's popped up beside it. This is a vision for what we see of our drafting solution that's going to come out later this year as well. And in that drafting solution, it says copilot on the top of that page there for a Microsoft plugin, you could not be on copilot and you just have co counsel there, which is our legal assistant is that you could use a practical law precedent document for sale of goods agreement. Or you could use your own template your own precedent from within your own DMS or net docs or whatever DMS you're using. You pull it up on screen in Word. And then you have this legal assistant beside you. You have co counsel beside you. And CO counsel can help you query against good law on Westlaw. It can help you query against practical law looking for clauses or looking for other elements in your sale of goods agreement that you want to swap out that you want to change that you want to verify. And other things that you want to do with in your drafting experience. And the idea here is that it's right, embedded in the tools that you use every day, like Microsoft Word. So this is a vision of what's coming later in the year. We know that there are this is an animated slide if you want to move a little faster, or just There we go. We know that there are challenges to adoption, we know that bringing all of this out to you all at once is maybe where there's that hesitancy that we talked about earlier today, that information can be siloed, that you have a scarcity of time to get these things trained in your organization, that there are other issues that your firm's or your organizations are dealing with in terms of a hybrid work challenge or in terms of intergenerational challenges, as well as an unpredicted. Well, economy, although the rate changes stayed today at 5%, thank you, Bank of Canada and a scarcity of time to learn all these new tools. The reality is these chant challenges to tech adoption. Don't go away. But what is going to happen if we go to the next slides, because the market is not changing the market is not telling us. We don't care about all that adoption. We don't, we don't want to do this, the market is telling us the opposite. client expectations are that they want transparency. There's a changing workforce, we do think that, like I said, and you know, maybe phase two or towards phase three, we're going to start to see changing pricing models. And we're going to see competition, the new market entrants coming into the industry as well. And so I do think there is an imperative notwithstanding some of the barriers to adoption we've seen before with this technology, I do think this time, the technology is much easier to understand or to interact with, even if you don't understand necessarily what's happening behind the hood. And so I think there is an imperative in the market to get on the generative AI train, I would encourage you to think about what that means for your organization. Think about use cases in which you can do things easily start with something you do every day like legal research, and apply a layer of generative AI to your legal research and something like a Westlaw or a practical law, look at co counsel to ride shotgun with you and help you do some of those tasks that take a lot of time and take some of the time suck out of it. So that you can realize more potential and free up billable hours to be doing other things or more billable things, but generating more free time for you. And if we go to the last slide, I think I just say thank you here, but Oh, no. It's more of that visionary that I talked about earlier. But having that CO counsel, generative legal AI assistant with you, and weaving it through everything that you do, this legal assistant is going to travel with you throughout your day, whatever it is, whatever part of the workflow you're in, whether you're moving from research into planning your litigation, whether you're starting from a share purchase agreement, and moving into sharing that out through high Q through a customer portal, wherever you're going or integrating with other third party apps like Microsoft Word, or iManage. These are all places where having that assistant with you will help make the work

Zena Applebaum [00:47:18] easier, more efficient, and really hit the needs of people around you that are asking you how you're using this technology if you're using this technology, and it is technology that? Well we are at the beginning of it. And we can ask questions like will you use it in five years, the reality is in five years, it's just going to be embedded into many of the applications that we use today. Whether we want it to be there or not think of it a little bit like spellcheck. It's around everywhere or AutoCorrect. It's around everywhere. Those are other forms of natural language processing and generative AI. And I think I'm sorry, not generative AI just other forms of AI, like natural language processing. And I do believe that there will come a point where we won't be having this conversation anymore, because generative ai ai will just be something that is built into much of what we are already doing. So with that, Zach, I think we can turn it over to some q&a, I see that it's popping up here with a lot. 

Zach Warren [00:48:14] Yeah, definitely. We're gonna get into some q&a. As we do that. We're gonna have one more poll that is coming up on the screen, which is just basically, okay. You've listened to us speak for the past 50 minutes? Are you feeling more confident about generative AI? Now? Yes, or No? or need more information? And as people answer this one, I'm gonna get started on some of the q&a. And there was one question that came in, which is pretty simple, but also one that I think a lot of people are really struggling with, which is essentially, we want to get started, where's the starting point? What is an easy jump in for us to beginning to, for us to begin using some of these generative AI tools, and Z and I know, you just talked about how it's going to be baked into a lot of these tools. It's going to be just a part of how people work. But for people who want to get started today, where's an easy entrance point?

Zena Applebaum [00:49:07] I think the easiest entrance point is in summarizing your documents in helping you get work product done faster, the work product that is not necessarily billable, and so co counsel is a great place to start. 

Zach Warren [00:49:19] 
Sure. So we're seeing the results of the poll that makes me happy about 48% said yes, feeling more confident, a decent amount need more information. And I think that makes sense. There is a learning curve to this. There is a lot of trial and error, a lot that people need to play around with to figure out how this is going to work. So I'm gonna dig into the q&a, and I'll throw some up there. There's a few questions around the business applications of all this that I'll pose to you Zena. There was one basically any insight about how firms are dealing with finding efficiencies and the impact on the time for firms Now, Bill, if firms are being more efficient than they are billing less time, but they're also paying more for AI tools. So, especially in your conversations with clients, how exactly are firms reconciling this these days? 

Zena Applebaum [00:50:12] Yeah, it's a, it's a really great question. And it's one that's on everybody's mind. It's not just the law firms It's on. It's on in House Counsel's minds as well, and government lawyers as well. So I think a couple of things are going to happen. You will bill less for individual matters, but you'll be able to take on more matters, because you'll be free up capacity, and you'll have more capacity to do more work. So that's one thing. The other thing is we believe, and maybe this is maybe this is too optimistic, but that maybe lawyers will get some work life balance back in their lives. Not that billable targets are going to necessarily come down inside of firms. But the way the peaks and valleys will sort of level out a little bit, and you'll be able to have a more steady workflow instead of hitting those peaks and valleys as you're working. The other thing is, we're gonna I think, probably start to see more value based billing coming about over the course of the next several years, where certain tasks that would have been billed on an hourly basis will start to move towards value based billing. Now I know, we've talked about this forever. Those of you that have been around, you know, Susskind was talking about this 30 years ago. But I do think there is a moment in time and a technology that is the catalyst for actually making that happen today. So clients won't necessarily want to pay more for these efficiencies. But you will get more work from them if you're being more efficient. And therefore, you'll be able to take that same time to be able to reallocate it to doing more work.

Zach Warren [00:51:43] Certainly, there are a few related questions in here as well around essentially, does that mean that big firms will require fewer associates in the future? Or does this mean that there will be consolidation? I'm seeing you shake your head already. 

Zena Applebaum [00:51:57] Look, the consolidation question is, is not just about workflow, right, the consolidation question will bring a whole bunch of other economic factors into that conversation. So I don't want to I don't want to put I don't want to predict what I think is going to happen there. In terms of the other part of the question, which was, are they going to need fewer associates? Time will tell I actually think that because there will be more work. Because there will be opportunity to do more work, they may not need fewer associates, I do think the mix of who gets hired in firms to do what will change the same way. We no longer in most organizations, you no longer have a one to one relationship with a legal assistant and a lawyer, you'll start to have different ratios of people and different hiring patterns within firms and different skill sets that are hired to do different tasks within firms. So I think the workforce won't necessarily shrink, but may change over time as well. 

Zach Warren [00:52:57] Yeah, one maxim that I've heard a few times that I do like is, AI isn't going to replace lawyers. But lawyers who know how to use AI may replace lawyers who don't, in terms of the headcount, it's people need to adapt with what is best to serve their clients. And if what is best to serve clients in the future? And what clients are actively asking for our AI solutions? Well, that's an adaptation like lawyers have adapted to different new ways of doing law in the past, it's just a little bit more of a technologically advanced way that people are going to have to adapt.

Zena Applebaum [00:53:35] There's a great question here about how do associates know what questions to ask and or be able to assess whether the output is good. And this is a great question that demonstrates exactly why the associate training component isn't going away, right? Because the question may be, hey, can you go draft me a share purchase agreement? Or can you look at this particular indemnification clause within a share purchase agreement? The associate may not may have enough information to go and ask ask practical our Ask Westlaw or sorry, Westlaw edge with CO counsel be able to ask that question effectively. But whether or not it's good output, the question would be the same if they were drafting from scratch. They lean on their partners to help them understand whether or not the output is good. It's just a starting point for the conversation, but it gets you there faster. So again, it's not like we're taking away the ability for associates to learn or the ability for partners to engage with associates and teach them it's just that you're gonna get output faster.

Zach Warren [00:54:45] Certainly, there are a few questions in here as well around privacy and security that I want to bring up because as we were going through the research, that was something that is a big question and just reading one of the questions that's in here, so how can these tools complete document review? Considering the lawyers code code of conduct of confidentiality? The person who submitted the question said there is a hesitation to put client information and Docs into an AI system. So what is your thoughts around the confidentiality and privacy piece of all of this?

Zena Applebaum [00:55:17] Two things. One, those Docs are already sitting on your servers, that would be no different than your private install of CO counsel sitting on your server or sitting in your instance. So the privacy and security question it's locked in, it's only yours only you would see your own documents. And we can get a security expert on to explain more about that or to share with whoever asked the question more specifically around that, but ultimately, at least I can speak for Thomson Reuters products. When you buy a product from Thomson Reuters. It's not being shared with anybody, nobody else is being trained on that data, that data is yours, and only you see your own documentation. I don't actually see the question. But one of the things that you said Zach, was complete document review. And none of this is completing, right, it's a step in the process to get you towards completion, it still requires oversight of the professional. We're not replacing lawyers, we're just giving you really good tools to make it easier to be a lawyer. So again, this is all in the, at least at this stage. It's all about increasing efficiency and increasing ability to do things that machines do better and faster than humans.

Zach Warren [00:56:35] Certainly, there's one question that just came in that I think is a really good one. Because I know when I give these presentations, I tend to focus on the law firm side of things. But this technology, at least in my experience, so far has been utilized more in house with corporate counsel, then with law firms, a lot of in house counsel are really jumping on this. And we're seeing that in our research as well, that usage rates are a little bit higher. So I'm kind of curious people that you've talked to Zena, what are some applications that are particularly relevant to in house practice, versus maybe what we see in the law firms?

Zena Applebaum [00:57:09] Yeah, so a couple of things with the legal research side, because we did launch in the US in November, I have a little bit of a sense of how us in house counsel have been using it, they're able to ask questions really quickly. So questions that they maybe would have relied on outside counsel, or they maybe wouldn't have been asking to help them understand where their risk tolerance is before making the next call to their outside counsel. So they're not stopping the calls outside counsel, but they're just able to have a better conversation with outside counsel, before they make the phone call, we find that they're using legal research tools to do that, or things like Westlaw with CO counsel in the US. In terms of practical law, they're able to get up to speed faster that you do all the things that they would have done previously, much faster as well. And so for in house, where you are a cost center, versus a timekeeper in a law firm context, it's the better the faster you can do things, the more you can spend time on proactively servicing your business. And so that is a really great value proposition for in house counsel, that these tools allow you to be more efficient and cost ultimately cost less, or cost the organization less, even if the prices are a bit higher for the actual tools. 

Zach Warren [00:58:26] Excellent. We're coming up on the hour. There's a few more questions in the q&a. But a lot of those relate to Thomson Reuters and how we're rolling out products as well. So for more information, how exactly Zena would you like people to either contact Thomson Reuters, or what is the best way to reach out for more of that product information?

Zena Applebaum [00:58:47] Yeah, so I would encourage you to go to the Thomson Reuters website, I would encourage you to sign up for our generative AI newsletter, I would also encourage you to reach out to your sales reps. They can answer all of your pricing questions and all of your questions around product specificity as well. So thankfully, I'm not on the sales side. And I can't answer those questions.

Zach Warren [00:59:11] Same here. So that's why if both of us are pushing that question off, there's a few questions around pricing timelines for rollouts. I'm personally not on the product side of things. So I can

Zena Applebaum [00:59:23] I can say Westlawn Practical are coming to Canada later this year. Draft will be launched in the US first generally, we launched in the US first get some feedback, and then launched in the UK in Canada. So we're sticking with that timeline, and you'll start to see products hitting the market in the next several months.

Zach Warren [00:59:43] Excellent. So any final thoughts as we have a minute left here?

Zena Applebaum [00:59:48] I just think it's a really exciting time to be in the legal industry. I know we've all been saying that for years with the with the technological advances, but I think this is one technology that you can't get away from and I Um, the sooner I'm better you can get in and start playing with it, I think the the more exciting it will be for you and the more efficient the industry will become 100%.

Zach Warren  [01:00:10] 
To the people, we put up that first poll, have you played around with this? So it was about 40% said, No, I think now is a good time to start jumping in, even if you're not necessarily using it for all of your legal work right now, understanding how it functions so that when the time does come, that maybe these tools are being integrated to what you're doing on a daily basis. You know, how it works, where it's fits, where it doesn't fit, and how it may impact the future of your practice. Well, thank you very much for your time today, Zena, and it's been a pleasure presenting here with the Canadian Lawyer. 

Mallory Hendry [01:00:44] And I'll just jump in there on that note, thank you so much to Zach and Zena for, you know, sharing their insight and expertise. We appreciate it. And thank you so much to everyone in the audience for joining us today. Keep an eye out for other upcoming webinars and everybody enjoy the rest of your day. Thanks again.