Law Firm CIO on AI Adoption: The Real Blockers in 2026

Jim Merrifield (00:01.586)
Well, welcome back to the InfoGov Hot Seat, the podcast where we sit down with the people shaping how data, AI, and technology actually get used in the real world. Today, I'm excited to welcome Judy Flournoy, former law firm, CIO, and someone who's been in the room making the hard tech decisions long before AI was a buzzword. is today. Judy, welcome to the Hot Seat.

Judi Flournoy (00:27.286)
It was very nice to be here, Jim. Thanks for the invitation. Looking forward to the conversation.

Jim Merrifield (00:31.922)
Yeah, absolutely. It's great to have you here. Now, Judy, before we get into the details on AI and tech, why don't you just introduce yourself to the audience?

Judi Flournoy (00:42.506)
Again, my name is Judy Flournoy. I'm a prior CIO for mid-sized law firms here in the US for the better part of 20 something years. I decided to retire a year ago and now subsequently have decided maybe I don't want to be retired. That's another funny story for another day. But I've really enjoyed the intersection of legal and technology throughout my career and most notably in the last few years, certainly since the end of 2020, deeply involved in projects that either utilize generative AI

or robotic process automation. And I think both of those technologies have tremendous impacts on law firms as they work as an operation, as well as how lawyers practice. And so I think it's just a fascinating time for our industry. And I'm really looking forward to see what happens now in the next couple of years as some of the newer companies come up and some of those that are more established try to sort out how they're really going to remain both competitive and perhaps.

I don't know, the word I'm trying to find here, relevant perhaps, that's the word.

Jim Merrifield (01:46.002)
Excellent, relevant is the word. So excellent, it's great to have you here. So let's start with a fun question. So as a CIO or former CIO, what task did you secretly hate that you thrilled AI might steal in 2026, maybe?

Judi Flournoy (02:02.574)
Fundamentally, because at the end of the day, part of the job of a CIO is to ensure that operationally things are running well, people are doing a good job, the attorney staff and other professionals are happy with the technology and the services being provided to them. It's help desk ticket analysis. Right? particularly for larger firms,

Help desk tickets are something that oftentimes get reviewed to see if, you know, what's the canary in the coal mine? Is there something happening here that I should be paying attention to? It's also a barometer for satisfaction. It's also a barometer for did this investment pay out? In other words, we made this big investment in technology. We took the entire community through a training program. Is it actually producing the result we were looking for when we made that investment? So I would want to be able to use generative AI to dig into that.

and take me beyond just simply there were five tickets related to this particular issue.

Jim Merrifield (02:59.484)
think that's excellent. That's using data right to your advantage and making decisions. So excellent. So be honest with us. What really slows AI adoption the most in professional services? Is it messy data, nervous partners, lack of talent, funding? What do you think?

Judi Flournoy (03:02.413)
Yep.

Judi Flournoy (03:15.694)
you

Honestly, I don't think it's any of those. I think it's not understanding the problem you're trying to solve when you buy a product. And so I'll use this sort of as an example. When you buy a document management system, you understand when you buy that, what its purpose is, what your community is going to actually utilize it for.

It's going to store documents, going to secure documents, et cetera. Hypothetically, find abilities to get solved if you have a document management system. With generative AI, the question is, well, what are you trying to do with it and then find the right product or service and apply that technology to the problem you're trying to solve? I think that for our industry, the last few years has been somewhat interesting and obviously challenging, but also exciting around this idea that

firms have gone out and made these massive investments in some of these products and services. And if you talk to the people inside, you might find that the adoption rate is lower than anyone anticipated, that the results are not as robust as they had anticipated, that the promises made by the vendor around what their product can do perhaps didn't get realized. But on the other side of that, there's some great success stories where the

problem was identified, the right application of the right technology, and voila, you get a really good result out of that and it helps the lawyers serve their clients effectively, efficiently with better results. And so that's, you know, I think that's really the issue. If you can provide the lawyer, this has been my experience, with the basis for which you are asking them to do something different.

Judi Flournoy (05:07.05)
and show them why it matters and show them how that it can improve their client service, their result at the end of the day, perhaps even drive more business in their direction. They're not as resistive as people think they are, but you have to be able to make that case.

Jim Merrifield (05:25.106)
Yeah, at the end of the day, it all goes back to the revenue, right? How can we make more money? You know, what's in it for me? So let's talk a little bit about that. By 2026, do you think that AI actually disrupts the billable hour model or does it somehow survive everything? Because we've been talking about this flat fee model for like ever, right? Probably the better part of your career. I know probably my whole career. So what do you think about that question?

Judi Flournoy (05:28.212)
It sure does.

Judi Flournoy (05:43.31)
Judi Flournoy (05:54.153)
I think it cuts both ways. think that for bespoke work, there will always be a billable hour. I think that there are some firms of a certain cadre that will always garner that type of relationship with clients. you have a, you know, bet your business litigation and you're going to bet your business litigation law firm, you're going to pay whatever it's going to take to keep you, your company from having to pay out, you know,

billions of dollars, millions of dollars, depending upon what it is. So I think that there will always be the billable hour But I do think, to your point about flat fee, I think we've seen it already. I think we've already seen different pricing models, particularly as firms have begun to understand through data analysis what it really costs for a matter. And so they can actually now budget. But they're using technology to help them budget. And then they can sit down with a client when the client says, here's the problem I want you to solve for me.

They can sit down with client and based upon prior work, they can actually tell the client, well, this is what it's going to cost. And it can be a fixed fee arrangement, and it actually can be a truly fixed fee arrangement where a client has cost certainty. So I think both exist in a world with generative AI. I don't think that changes.

Jim Merrifield (07:12.56)
Yeah, I agree with you. There's plenty of firms out there that are creating a pricing arm within their finance groups that are actually connected with the AI team, right? And as you alluded to, being able to look back on previous history of data and actually, you know, mine the data and understand what a matter actually cost these days.

Judi Flournoy (07:18.466)
Yep.

Judi Flournoy (07:22.158)
Mm-hmm.

Judi Flournoy (07:28.301)
No.

Mm-hmm.

Exactly. And why some firms, and I think you alluded to this, Jim, they hire data scientists now, right? People who can actually dig into the data and help surface the real trends. And then firms really understand where profitability lies within their matters. And that helps them do a number of things as a business, not just simply engage with the client.

Jim Merrifield (07:54.044)
Yeah, absolutely. So here's an interesting question. When AI gives bad advice in client work, right? Because it does, it does hallucinate. Who holds the bag in 2026? Like who's, I don't know, like who's on the hook there?

Judi Flournoy (08:00.463)
You

Judi Flournoy (08:10.393)
Well, I think that's a great question. It's an interesting question because I'm actually currently taking the Stanford course called AI Strategy for Legal Leaders. And the instructor for this course is Wei Chen, and she just presented this in the module I just went through. She just presented this very question about who's responsible. And she utilized the EU's AI Act as a framework where within the EU's AI Act, and you may be familiar with this, there's the provider.

there's the legal AI tool and then there is the law firm, right? So you have the provider that might be Gemini, Claude, Chat GPT You then have the legal AI tool that might be, you know, we talk a lot about Harvey these days, Harvey, Lexis, name it, right? Or it ends then the actual provider is the law firm. So from my perspective or my opinion, I think at the end of the day,

the lawyer is always the last mile. At the end of the day, regardless of what technology they're using, regardless of how they got to that answer, regardless of what resource tool they use, at the end of the day, it's up to them to verify the information they're using. When they cite a case in a brief, they need to go back and they need to go to trusted resources that are kind of old school, right? Your Westlaw, your Lexis, right? They have these massive data sets and verify that the case that got cited

actually exists and that way they can stay out of trouble. So I think at the end of the day they are responsible. But with that said, I think it's extraordinarily important for the law firm and the lawyers to understand whatever product they're using, how it was trained and what it was trained on, right? Because generative AI is probabilistic.

Whereas machine learning is deterministic. So you really have to understand that as a user, is that it's going to give you what it thinks it should. It wasn't trained to be right. It was trained to give you an answer. In LLM, chat GPT doesn't know the difference between right and wrong. It just knows you've asked it a question and it needs to give you an answer. And that answer can, to your point, Jim, it can hallucinate. It can be a hallucination.

Jim Merrifield (10:23.344)
Yeah, absolutely. There's a lot of lawyers that listen to this podcast. So we'll see if they agree with you. yeah, yeah. Hey, who doesn't like a great debate? So let's look ahead a little bit. What CIO skill would matter most in 2026 that barely mattered five years ago? Lots changed five years, right?

Judi Flournoy (10:28.751)
I invite the conversation and the debate.

Judi Flournoy (10:45.143)
yeah, lots changed. think that truly influential CIO needs to understand every single aspect of the business and not just the technology. I think you need to understand what's happening in the business development department, what's happening in the accounting department, what's happening in the practice development department or professional development or e-discovery, information governance. I think that

focusing purely on the technology doesn't answer the broader question, which is how do I as a CIO provide the most value for my law firm and more in particular, the lawyers that I work for at the end of the day in a private law firm, they are my employers, X hundred of them, no matter what size your firm at the end of the day, you're as a CIO, you work for them all. Now, some of them may be a little noisier than others, but at the end of the day, you work for them all, right? You can get the link so you understand what I mean by that.

The point there being is if you don't understand all of those elements within the firm, then what you don't have the opportunity to do is capture opportunities around how can I utilize the data that's actually in the accounting system? What technologies can be applicable to help the accounting department, the billers, for example, capture more time? Is there opportunities there to realize more revenue? Is there a way to shorten our collection cycle?

If it's on the biz dev side, how can you work with the business development team and the marketing team on better results coming out of a website, collaborating with them on a new website deployment? If you are working with your litigation support function, what are discovery tools are you utilizing today? Is there an opportunity for generative AI? Is there an opportunity for using robotic process automation in doing some type of analysis or getting data, pulling data out of the internet?

producing results for clients. So I think it's understanding all of those elements because you can't serve an entity which is a law firm without understanding all of the things that make up that law firm and all the people within it.

Jim Merrifield (12:56.528)
Yeah, 100 % agree. I mean, you can't have silos on the, especially on the administrative side. Cause we're all, we're all in it together. Cause I always talk about, you know, a lawyer starts, a business professional starts at a firm. Day one, they have no idea, you know, who does what. They just know that there's an administrative professional that can help them. Right? At the end of the day. And still sometimes five years later, they still don't know who does what. They just know.

Judi Flournoy (13:02.116)
especially.

Judi Flournoy (13:13.935)
No.

Judi Flournoy (13:18.603)
Exactly. Exactly.

Jim Merrifield (13:24.562)
that there's administrative people that can help them. So excellent, excellent point. So all right, let's wrap this up. We're gonna get into the lightning round. So this is the fun stuff, all right? So listen, one word answers only. No pressure, right? We're gonna ask a few of them. No pressure, no pressure. One word answers. If you have two words, it's not a big deal. So one word, what's the most underrated AI capability?

Judi Flournoy (13:25.271)
Right? Right?

Judi Flournoy (13:33.76)
Okay.

Judi Flournoy (13:37.807)
Okay. All right. No pressure.

Okay.

Judi Flournoy (13:51.372)
It writes.

Jim Merrifield (13:53.372)
Okay, it writes. What's the biggest blocker to AI adoption?

Judi Flournoy (13:58.287)
fear

Jim Merrifield (14:00.176)
Most exciting AI shift by 2026.

Judi Flournoy (14:04.879)
I don't have a one word for that. Consolidation.

Jim Merrifield (14:07.89)
It's okay, a few words.

And finally, biggest AI risk for firms.

Judi Flournoy (14:16.227)
blindness.

Jim Merrifield (14:17.906)
Awesome. You're a good sport. I love it. I felt like I was doing a sports center, you know, a sports show show on that one. So that was great. Awesome. Well, listen, Judy, this has been fantastic. Thank you so much for bringing the CIO perspective to the hot seat. We so appreciate that. And to everyone listening, if you enjoyed this episode, make sure to follow, subscribe to the info of hot seat, wherever you get your podcasts and we'll see you next time. Thanks so much.

Judi Flournoy (14:20.527)
Ha ha ha ha ha ha ha ha ha ha ha ha ha ha ha ha

Judi Flournoy (14:46.511)
Thank you so much, Jim. You have a good afternoon.

Jim Merrifield (14:48.742)
you as well.