Last month, Thomson Reuters announced a plugin with Microsoft 365 Copilot, meant to assist with legal research, drafting, and client collaboration, and outlined new generative AI capabilities in its online legal research service Westlaw, including its Practical Law reference too and an AI-powered contract and document-analysis tool, Document Intelligence.
No Jitter had a chance to speak with David Wong, chief product officer at Thomson Reuters, about how generative artificial intelligence (AI) will shape the experience of practicing law, what we still need humans for, and where the company will go next. An edited version of the conversation appears below.
Q: Can you tell me how this suite of tools could help like fresh out of law school associates acquire the kind of experience and perspective they'll need to get more out of Copilot?
Wong: What I would say is that it's still very early. So you know our clients, law schools, Microsoft, everyone, we're all still working out how this affects the industry sort of end to end.
These types of tools and these types of integration will offer will help people save time and effort, improve the quality of the work they do and ultimately allow professionals to deliver value in new ways.
As a junior lawyer, you'll probably work with a senior and say, hey, how do I do this? What are the questions I should ask, and we envision that features of tools like Ask, Westlaw or ask Practical Lawwill help those more junior lawyers to be able to answer those questions and to be able to get up to speed a whole lot faster.
There is a hypothesis right now amongst many law firm leaders that the experience of new law school graduates in their firms will pretty dramatically change that there will be less rote work that they'll have to do as sort of, you know, earning your stripes. They can get on to learning what it means to be a good lawyer. You're getting to judgment and reasoning faster.
Q. Getting to judgment and reasoning faster -- is that the natural cognitive or skill progression as you as you practice law?
Wong: What our customer research and what the feedback from our customers indicates is that yes, there is this progression from rote work that then helps junior lawyers to then gain legal judgment and then take on progressively more and more complex tasks. There's also sort of a scope of judgment, which also grows so you can first learn how to apply judgment to simple tasks like contract review, which would then allow you to then get judgment for full contract drafting and then to full transactional matter work. So that's sort of the evolution from, you know, just being that junior associate to a senior associate to a, you know, eventually a partner where you're going to be the one leading a piece of work.
Q. Can you describe for me how Practical Law and copilot will be working together? So if you do have Bob, who's drafting a contract, are they also able to then go to Practical Law and be like, "I have no idea what I'm doing here. Please help me?"
Wong: One of our kind of fundamental beliefs on the application of AI is it's here to augment professionals. What we hear from customers, and what we believe is necessary in the market, is to be able to create the combination of technology with human judgment and experience to be able to deliver again better results, faster, fewer errors, etc.
Here's a bit of background on Practical Law as a product. Sort of the tagline of Practical Law is sort of "the law partner down the hall," so the classic use case for Practical Law is for the general counsel or an associate general counsel at a corporation that has to deal with a matter which they are not familiar with.
The kind of unfair expectation of a general counsel at a corporation is, you know everything about the law, right, like you're our general counsel. But the law is just too complex, and usually general counsel will be specialized in particular area -- maybe they're good commercial lawyer or a transactional lawyer or privacy lawyer or what have you. So when something new comes in that they're not familiar with, there's two generally two things a general counsel can do. They can call up a colleague that knows something about it. If they don't have that, then they call up an external counsel, they'll say, well, I hire this expensive law firm often to provide me with guidance on these topics. So let me call the partner.
What Practical Law does is, it helps to provide that first line of knowledge. So that's why it's the partner down the hall, which is the step before you call external counsel. You look at Practical Law and ask for some practical guidance on how to deal with a particular matter and it'll often give you enough so that you can get started, that you can again start to work in practice and maybe the next step is calling up external accounts. You realize, oh, no, this is a big deal. I need to call up external counsel.
Q. So this tool basically teaches you how to learn about a subject that you're not necessarily an expert at, then tells you what you do and don't know already?
Wong: That's right, but it also is a compendium of practical guidance, so a lot of what's in there are checklists and summaries of the implications of changes in law within your particular business.
The example in our copilot demo was asking Practical Law about a particular contract clause, because reference materials are one of the most frequently used capabilities of Practical Law. So there's what we call sort of a gold standard contract a template that you could start with that is safe as a place to start. You edit it. You can have add and remove from it and that's probably a good place to start from if you're going to be doing a sales agreement or, you know, submitting an IPO or things like that.
Q. One of the things that popped up in the demo is mention of "a deep well of expert content," especially in your Westlaw product. Where is the data coming from for that deep well of expert content?
Wong: To give a bit of a picture of what is the content that's behind Westlaw --there's two big bodies of content. Primary law content are cases that have been argued in front of courts that decisions have been made on, as well as statutes which are the law as defined by the federal government or state governments or what have you. And then you have commentary which is authored content. So that would be things like law reviews, books. You know, uh expert opinions, things like that. Again, those are the two broad categories of content: primary law content and commentary.
Usually when we're talking about the expert content, we're referring to the primary law-- the case and the statute content, which is most frequently used for case law research, because those are the cases which are used to be able to form arguments when you're litigating.
Where do the experts come in? That's a good question because case law is in the public domain, so it's available from the courts and what have you the expert analysis that we put on top of it is what we call editorial enhancement. So you could just use the text of the case and that case will be many pages long and will be very difficult to sort of understand where it sits within the body of law, right. So what we do is our legal experts will read the case. analyze it and do a number of things.
We will summarize it. So instead of having to read sometimes a 20-page case, we'll give an abstract which is just one paragraph. So that's much easier to search, much easier to do research on.
We will classify it into a sort of a master index of the law. So we have something called key site, which is a classification of the law. So it'll be every topic of the law in a in a tree hierarchy that you can think of. We'll place the case into that that tree.
And then the last thing that we'll do is we'll annotate the case to extract relevant information and that relevant information may be mundane like what was the jurisdiction right? Like this was argued in North Carolina, so that's usually easy to determine.
We are also increasingly extracting out much more deep meaning and knowledge from the case. Things like what were the material facts in deciding this case? Or what was the judgment -- something which you would have to really not only read the case, but have to understand the case and have sufficient legal knowledge to know how did this case get ruled and also what were the causes of action for that case. We've attempted to use artificial intelligence to be able to extract out, but we found that it only kind of gets you part of the way there that you still need humans to be able to do that annotation.
Q: You seem to be highlighting one of the core misperceptions that swirls around the AI, which is that you're just throwing a big body of data at the algorithm, and the algorithm is figuring it out. But what you're telling me is that you do have a lot of hands-on human intervention to provide appropriate context and build out complex use cases. So how do you see the use of AI enhancing the way that we that that these tools will help people practice law?
Wong: You know that that if if I were to summarize how how we see success with the application of these tools, we're going apply AI to be able to further enhance and make content more accessible, more easy to search and more readily available to legal professionals of all you know, all experiences and domains.
We've been experimenting with enhancement score search algorithms, for example using the AI technology for years. You know, search in Westlaw has been employing AI for 30 years. We predate Google in terms of you know as a search algorithm and we have incorporated a sort of every new search technology advance as time has gone on in and including now with generative AI.
We're going to apply generative AI to vastly improve the collection of this content, so the breadth of information which will become available will be larger. We've been looking at applying AI to our content collection and creation for years. We, we suspect we'll be able to further improve the efficiency of that over time.
Q. Westlaw seem like such an intensely human endeavor, with the labor that you're applying to assemble your master index of law and keeping it up to date reading cases, analyzing them, summarizing them. Is it a ripe use case for AI to take out some of that work -- or rather begin some of that work?
Wong: We're already employing it. I think it's just increasing the proportion to an extent, right. So the tooling that these teams use right now. At the end of the day, I think what we'll see is that it will become increasingly rare that a that a lawyer will not have some kind of AI assistant to conduct their work.
One of the pieces of feedback we've heard from our customers is that they totally get the efficiency argument, but what they're most sort of tantalized by is the way that generative AI can help them to explore different theories and ideas that it helps to connect the particular matter that they're working on to adjacent topics to consider alternatives, and some of those might not be of any relevance, it's it could be dead ends, but simply the ability to expand the aperture at the right time when you're doing research is super valuable.
Again, we've all experienced it when we're doing problem solving, you have a partner that can help to open your mind a bit. And then you can have come back to what you think the answer should be out of it.
