11/06/2026: ClioCon recap, AI focused law firm internships, and more

0:00No, I don't think so.

0:05hi everyone sorry for the delay we had some uh technical difficulties um yeah it wouldn't be it wouldn't be oh uh oh all right uh we still have some technical difficulties uh welcome to uh legal tech week i'm gonna try to figure out what's going on with my camera so let's just go around the horn uh let's start with you joe

0:32No audio, Joe.

0:37I think this episode is cursed. I think this episode is cursed. All right. Want me to go while he figures it out? Yes, go ahead. My name is Nikki Black. I'm the principal legal insight strategist. At 8 a.m., I write legal tech columns for ABA Journal, Above the Law, and The Daily Record. And I am just impressed that we all are here. So I think this is a good thing.

1:04Hey, I'm back. I had the wrong microphone turned on. Anyway, Joe Patrice from Above the Law. And yeah, well, in the interest of time, I'll just say that and move on. All right, let's go to you, Steve. Hey, Steve Embry. I write the blog Tech Law Crossroads and also write for Above the Law. And Julie. Hi, everyone. I'm Julie Shabawali. your resident Canadian on the panel, writing for Canadian Bar Association, The Walrus, and other publications. Good to see you all. And we can see you, Victor. Yeah, I don't know. This thing's on the fritz. I just bought it. I don't know why it's all messed up all of a sudden.

1:44I will note that somebody in the chat said that Julie made it on time. I think technically speaking, we started late, so we don't really know if she made it on time. I agree. Just to correct the record. It's true. I think I made everyone late, and somehow I'm making Victor's camera. Yeah, I don't know what's going on with this thing. I apologize. I might just have to start it off. And then we have a guest with us. Jake, can you introduce yourself? Sure. Hi, everyone. I'm Jake Marr, a legal tech reporter for Law360. It's my first time being out here. So thanks for inviting me. Thanks very much for having me.

2:25Looking forward to it. Yeah, usually we're in a little bit better shape than this. You know, Bob leaves, everything goes to hell. So, but anyway, yeah, so let's just start off with, you know, well, here, we'll let you go first because you're the guest here this week. So you wrote an interesting article this week about Wile. Why don't you go through that? Sure. Sounds good. And yeah, my editor and I saw some headlines kind of from around the legal tech journalism world about a few different AI native firms that were making bids or publicly announcing that they were making bids for certain data assets from Weil as the firm gets ready to dissolve.

3:10and you know that news was was out there and there was some discussion about it so my my editor and I wanted to dig a little more deeply into kind of beyond the headlines to look into those bids and ask like is this even possible what would something like that look like what kind of data actually could be for sale and like what would the value of something like that be and so we touched base with a lot of legal tech experts and got their read on it. I think as a lot of people on social media and just like the world pointed out, there are tons of client privilege, confidentiality issues that make it.

3:49So it's really, there's a lot of hard limits on what a firm could, what kind of data could be available to buy if WILE was even interested, which is also not in any way established, whether this is actually for sale. These are just unsolicited bids.

4:10But I'm sorry, I was reading the chat echo. No, okay. But so we got into the question of what would all of this, if there is something to be sold, what would it be worth? And what might be able to be sold is info on firm operations, things that are scrubbed of any client identifying information to clear privilege and confidentiality issues. Things that, you know, I think one quote that one of my sources told me but didn't make it into the article was, everything is worth something. And I think, you know, that's, somebody might be interested. I think he pointed out that it speaks to the moment we're in with terms of Gen AI, where large quantities of internal data from a company could indeed be worth something if you were going to train something on it, especially for an AI native firm that doesn't really have I'm sorry involved in making offers no again while they not respond to request for comment and there's not really any information out there about whether they welcomed this kind of thing these were unsolicited bids mostly so yeah it's not clear like how if this is actually going to go through but I think when we started reporting it out the bigger question we got into was how law firms think about their data, how they treat it, how they're starting to think about ways that AI can turn that into more of an asset that they can draw on.

5:35And that's really where the reporting ended up going mostly. So it's an interesting article. I prefer you to check it out.

5:43Yeah, I saw some people talking about it online and obviously people were a little freaked out by this. I mean, did you get the sense that this is something that other firms or other companies who have gone under that something that that that's happened quite a bit or that they've done no that was what i i asked all my sources that they knew of any comparable situations and no one um could really point to anything and again i think that kind of speaks to the fact that up until this time um in terms of like the technology there really wouldn't have been any market for something like like no one would have really had any use for something like this so it's kind of a novel situation i think that again there's a lot of fuzziness about what what could happen here what's even possible that the article gets into.

6:27But yeah, it's kind of a new situation. Well, it's interesting because the other day I got an email, clearly I don't have my own law firm at this point, but offering me$500 ,000 for my law firm's data. I don't know the legitimacy of the sender. I don't know the legitimacy of the offer, but it was interesting. It was a tech company, a purported tech company offering to pay not a small amount of money for law firm data, which then triggers so many ethics issues. I mean, what type of data are you talking about? Like they called it operations data. That includes all sorts of confidential information.

7:00So I just don't even, I can't imagine any law firm being able to splice and dice their data in a way where they could sell some of it off. But I thought, I think that's definitely going to be a increasing trend of value of this type of data, becoming more valuable and companies trying to get access to it. For sure. One of the sources I talked to said that they've gotten many unsolicited, kind of like you described, random bids for their data. And again, yeah, it would be a lot of work to make that clean enough to use for anything in terms of confidentiality issues. So that gets to the point of like, would it actually be worth the trouble to go through that?

7:39And apparently some companies think it would be. I'm looking forward to AI companies realizing how unorganized law firms are on the back end um i look forward to to that um because and nikki that's such a nice offer 500 000 for data that doesn't exist um i thought but of course i don't have a law firm so they don't know that you can just make up some data right just give it to them here here's my data that's some good synthetic data but i just wanted to mention do you guys remember um companies who were trying to do, what is it, what were they trying to name it? Like legal AI analytics for inside law firms, right?

8:21Where they were like, leverage your billing data, leverage your, you know, this so-called non-client confidential data. But the problem is not just the ethical concerns of turning over the data and all of that, but a lot of these AI tools are linked to other AI tools. So when you give something to one company, you're kind of giving it to a lot of different companies all at the same time. So I can't even imagine how you would deal with that as well, because we know all these vendors also have other vendors that work with them.

8:58Hey, Nikki, when you got that letter, did they ask for your bank account number so they could just send you your$500 ,000? Yeah, my social security number and my database, my kids' names. I sent it all. I mean, I figured, you know, I'll try. We'll see what happens. You should represent a bunch of Nigerian princes and they're your clients. That's what I was going to say. This Nigerian prince really wants my law firm data. He says it'll help him get his money out of the country. Yeah. Julie, you may be wild. Sorry, go ahead. I was just going to say, Julie, you make an interesting point about tech companies finding out what the data looks like on the back end.

9:41because I spoke to one sort of data career guy who described the way firms sometimes treat their data storage as a dumping ground, basically, with not a lot of organization, not a lot of ways to track or really make use of it. And only recently firms are putting in the effort into performing that a little bit. But yeah, it's an early stage process, it sounds like. Yeah. Yeah. And let's remember law firms are the second most targeted for cybersecurity after retail because of the credit card information is really easy to grab. So it's just like, just like, why would card information have to do with billing?

10:24Again, like you said, it's a dumping ground. So I, but we're also in a funny time right now with AI. So I just wonder if actually who owns this data? I'm sure you talked about this in your article, but because the law firm is dissolving, right? Like because of the way we deal with data in firms as a dumping ground, like some of it technically like, you know, people who are going to go to work at other firms, they'll take their clients with them. But will they take the data with them or does that technically live with the firm? Well, and some of the data could belong to the client. I mean, when you're talking about attorney-client privilege, it's the client's, I mean, client's privilege, it's the client's data.

11:12And I know when you change firms, if you're going to take business with you, you have to get permission. So that would be hard to sell data that would belong to a client. Exactly, Steve. not hard but that are if it's yeah exactly that was the main issue raised by people i spoke to you and people just in the general commenting world that yeah but you know if it's client data belongs to the client and it can't be it's not really for so but there's like again like more like operational things like nikki was talking about those are um you know not really attached to a particular client basically and also begs the question like sort of like how would you how would you privatize it so that um yeah i mean i mean i mean they must have like i don't know terabytes petabytes of like this stuff how so how would you how would you how would you go through all of it or would you just use ai to like kind of go through all of it and and and um you know uh was it redact all the i'll redact all the info when i left that raises a whole bunch of other issues like how good is your technology at the end of the day yeah but companies were doing this before i remember writing about this as early as 2016 um and and they were really good even then of being able to you know like you're saying anonymize the data but you can track easily billing there's so many people who were trying to sell you know let's optimize your billing give us all your billing and we'll run it through machine learning and predictive coding and all this other stuff um we've kind of reached a sophisticated point with AI where I don't know if anonymized data is even really a thing anymore.

12:53Good point. Yeah. Well, hey, I mean, Wild's got to raise money somehow, though, right? So this might be the way to go. Well, it's also, I mean, I don't know if your article covers, but wasn't there also something about how they're also looking to sell the brand as well? there was some talk about that that didn't my the article didn't go as deeply into that um but i i did see some chatter about like the uh the brand itself maybe being included in some of the bids that were being mounted um that kind of got into a whole other direction of branding and and legacy and whatnot that we didn't we didn't it was that was outside the scope of what we're working on but yes there was i think i do remember seeing something about that i don't even know what you would do with that i mean i mean unless you're gonna start your own successor firm called wild gotchal but i mean i think it's very yeah i feel like the name's how i damaged a little bit at this point yeah yeah i'm not sure what that would be put to use for exactly so that's another unresolved question yeah well on that note let's uh so i'm gonna work on my segways here they aren't They might not be as good as Bob's, but let's go to a firm that is not closed yet.

14:12So, Steve, you had a story about Allen and Overy Sherman. I hate these firms that put the and before the last name. Allen and Overy Sherman and their multi-year, multi-billion dollar commitment to AI. Yeah, so the article that I came up with was actually in a funny place because it was one of the few that was behind – it was not behind a paywall, which is why that link is there. But apparently the firm has announced that they're going to invest$1.1 billion over the next three years into – it looks to be like building AI tools for the firm, which that in and of itself wasn't particularly interesting because other firms have done that, although a billion dollars is a billion dollars.

15:06But the thing that interests me is they were going to fund that investment not with hard dollars, but in part by requiring each practice group to contribute, I guess that's the right word, 3 % of their billings in a form of non-billable time as an investment in trying to build the tools. In other words, if they had 10 ,000 hours practice group billable time for the year, then 3 % of that should go to developing AI tools. And I thought that was just kind of interesting. I've not seen a firm do that, and I'm not sure how they're going to calculate that. I mean, are they going to, like at the beginning of the year, say your practice group quota is 10 ,000 hours, so do the math.

15:56we expect you at the end of the year to have non-billable time of whatever 3 % times 10 ,000 hours would be. And that only gives you a time number, but then you've got to calculate the dollar number. And interestingly, the article says that each practice group head could sort of figure out who was going to do it, either by hiring a full-time staff person or spreading it among the associates and partners, which probably means spreading it among the associates. But in any event, and it also struck me that if you were going to hire a full-time staff person that kind of defeats the purpose, I mean, it would seem to me the purpose would be to have lawyers engaging with AI tools and learn how to develop and build tools and platforms specific to their practice group, which a full-time person could do, but you're going to like.

16:51So anyway, it just struck me there was a lot of questions to it, but it was sort of an interesting approach that 3 % of your billable hours, I mean, that is a dollar number. It is a – you can calculate what it is. It's not like cash, hard cash, but if you assume that there's plenty of work to do, then that is a solid number, whatever that number is. So I would be interested to know how they calculated it and how it's going to turn out in the end. But I thought it was just a sort of a unique approach to coming up with a way to encourage lawyers in the firm to work on developing AI skills and tools.

17:36Yeah, I guess I'd be interested in sort of wondering, like, because, I mean, one thing, the article does bring up that they were Harvey's first customer. Oh, yeah, that's right. I forgot that. yeah and so i'd be interested in so i mean not to not to start any you know stuff with harvey but like you know like you know i mean obviously you know like like harvey's not going to cover everything that they need but why not but so so what so yeah like so what tools are they going to be focused on that that to address areas that maybe harvey doesn't touch for them so that that that'd be pretty interesting sort of like what what their approach is going to be i guess it mentioned um well i mean specifically caught like they built contract matrix as well with harvey and Microsoft.

18:12So I assume those areas would be covered and there'd be other things that they'd be looking at. But yeah, it doesn't go into details. But yeah, I mean, look, 1.1 billion is not nothing. Well, maybe it is for them, but it's not nothing. And didn't Kirkland also weren't they also investing like a billion plus as well into AI? Yeah, that's right. I don't remember with the Kirkland deal whether they were kind of doing this song and dance with contributing non-billable 3 % of your billable should be non-billable time spent on this but yeah and Morgan and Morgan of course they made a big splash about I can't remember how many billions that they threw in the pot but I'm sure it was a lot it was one yeah so this is literally just trying to one-up Morgan Morgan's number by the smallest possible margin.

19:12Yeah. That's interesting how, I mean, the cynic in me would kind of read that and conclude just what you did, Joe, that we're trying to one-up and we're doing it with this sort of little bit of funny money with this billable hour idea contribution. I mean, yes, it may all be legitimate, but on the other hand, there's a lot of ways to massage that where it really isn't anything but a funny dollar figure at the end of the day. So I guess we'll wait and see. I don't mean to slam Allen and Overy at all. I know how those kind of numbers and schemes can work. So there is room for some maneuvering within that.

20:03Well, it seems to me that telling people to, it feels to me a lot like what you just said, sort of this smoke and mirrors way of one-upping their competitor, because realistically, the likelihood that these associates are not going to feel the need to bill 100 % of their billable hours is kind of low. I mean, they're going to feel obligated to, you know, bill 103 % of their billable hours then, you know, with that 3 % tacked on. I mean, and I speak from experience because in 2002, I told my company, the firm, that we needed a website. And they're like, well, what do we need a website for? I'm like, because you need a website because everyone has websites.

20:43And so I bought Dreamweaver. They bought it for me. And I was allowed to track the time and, you know, have it be billable. But at the end of the day, I built this website and one of the partners said to the other part in front of me, did you see the website Nikki built? And he's like, no one's going to hire a lawyer over the internet. So clearly most of the partners weren't even bought into this. And this idea that I'm like spending this non-billable time on a website that most of them don't even think is valuable. So I feel like that's kind of, I think that the firm thinks it's valuable, but I still felt like I had to bill all of my regular time.

21:16And I'm sure that on top of that, and I'm sure that these young associates are too. So it is a really interesting way, Steve, I think, to make it look like you're spending money on something where you're probably really not. Yeah. And I really got to sort of wonder, too, you know, let's suppose that one of these three percenters comes up with this great tool to do, you know, client work at a fraction of the cost for what it was. So you've now you've spent three percent non billable time of what you would have billed in order to create a tool that's going to reduce our billable time by another 10 percent.

21:55That sounds like a sure-fired way to do a promotion or partnership. I don't know. I don't know. I hate to be a cynic about all this stuff, but, you know, there's reasons to be cynical. Well, it feels like that's just what's happening, not just in law firms, but everywhere, is everybody's spent so much. And we talk about this all the time. It's not a novel concept I'm bringing up, but everybody's invested so much in it, and they're trying to show the benefit of this investment, that productivity, efficiency, and they're not seeing it. So they're trying to force them now to create it somehow so they can at least show some ROI from the money they've put into this.

22:33The significant amounts they've spent both in terms of employees, sending them out to these conferences to try and choose the tools and then the actual money spent on the tools. So I think we're hitting the point where the pedal hits the metal and what's actually coming out of all this. I think that's right because Go ahead Joe, I'm sorry I mean Victor, I'm sorry I was just going to say the irony of it all is when you look at the AI tools that do some of the back office kind of stuff that's easy to say, okay, I can see the return on investment and we saved $5 million because we could reduce staff and reduce overhead but then when you start looking at the tools that actually reduce the time spent on substantive tasks, the ROI really enures to the client.

23:26So you've got this sort of paradox where the law firm is expected to spend buku's amount of money on AI tools that in essence does nothing but reduces their revenue and their profits. And the clients reap the benefit of it. So it puts the law firms in an odd space, I think. And I'm not sure how that's going to ultimately play out. Yeah, I really see this all as what Nikki said and also the billable hour and the fight against the billable hour. Because if you're a firm who doesn't have billable hours, you're fine. You can spend money on this and create new tools or products or, you know, get a whole new client base, right?

24:09Like, just like you said, there's AI tools that are focused on exactly that people who are doing flux plans, alt pricing, fixed fees, etc. From from wills to corporate stuff and other stuff. But if you're, you know, a big AM 100 firm that lives and dies on the billable hour, I see these and I'm like, what is what are we going to do with one point one billion? because we don't need$1.1 billion to actually build an AI tool that's saving administrative time, chasing after the clients to get paid faster, whatever the goal is. You don't need that much money. So I'm also extremely cynical about this.

24:48And as somebody who has written articles about lawyers should code, lawyers should experiment, I'm realizing now do that if you are actually invested in getting away from the billable hour, it doesn't work if you're still going to do billable hours because you don't have the incentive to reap the benefits of what the technology is actually going to do, right? That's just where we are. Yeah, I think that's right. And it's kind of funny, I've heard this anecdotally and also saw somebody wrote an article a couple of days ago that some of the clients are now saying, you know, we sort of like that billable hour model because now it's only going to take you an hour to do what used to take you 10.

25:37So why would we want to make a change? Well, on that note, I figure I'll bring up these two little, two briefs that I saw this week. So the first one, sorry, I just put in the link. the New York State Bar issued a report for guidance for lawyers talking about what they shouldn't bill and what they shouldn't count as far as their billable hours and whatnot. And what they said was that the time that it takes lawyers to learn how to use AI should not be counted, and the time that they save as a result of using AI. So if something took you to take them four hours and now it took them one, they shouldn't bill for the original time.

26:28So I was wondering what you guys thought about it. I mean, I thought it was a fair. It seemed like a pretty fair opinion, like a fair report and whatnot. I mean, it seems pretty straightforward. It doesn't seem like lawyers should be able to pass these kind of costs on and whatnot. But yeah, I was just wondering what your take was. The thing that struck me, Victor, about it was, yeah, I can see you can't bill the three hours you would have spent but for AI. But I can see circumstances where a client and an offer would get together and the client would say, we really want you to use AI. I tell you what, we're willing to pay for the time you take to learn how to use it adequately.

27:14In return, of course, for reduced bills on the back end. And that actually might be a good incentive for some clients to take. So yeah, but I don't know that the opinion prohibits that. It was just an observation. And even on the other side of the client says, well, I know it would have taken you three hours. I'll pay you the three hours. I mean, if a client says, okay, you know, I guess that's fine. If it's informed consent, it's what happens, which is all too often when the law firm or the lawyer tries to bill for time, they didn't really work. I bookmarked the report because I'm in New York and I like to write about it, but I haven't read it yet.

27:55But that particular finding or conclusion, if I recall correctly, that was in their original set of guidelines. And it's been one that has been repeated a couple of times in other opinions, and it mirrors legal research and how lawyers were allowed to bill for online legal research and the savings that came from that versus the book research, which is that, you know, you can bill for the time formulating the query, whether that's legal research or whether that is a prompt with AI, but you can't bill for the time that you saved by not doing it the old-fashioned way. And I mean, I think that that analogy makes, you know, the way that they handled it with legal research was correct in applying that.

28:41I always like to say, you know, when I used to talk about and write about social media and cloud computing, you know, the online, look for the offline corollary of the online activity, right? And this is sort of look for the, you know, before times, before tech times, before AI came along, look for the before tech times corollary. And that's, you know, legal, online legal research is a great way to do that. And I mean, I just think it makes sense. Why would you be able to bill for the time that you saved? You know, you're using this new tool and you're allowed to bill for the work that you've actually done, not the work that you saved.

29:15I mean, that's we've always looked at billing by the hour. You know, if you're smarter than your colleague, you get it done quicker. It doesn't mean that you could have or you have more expertise and they have to do more research. I know there's questions about research to catch up on a field of law and how you can practice area, but you know, you shouldn't be able to be like, well, I did it faster than them for any number of reasons. Therefore, I'm going to charge the same amount of hours that they would have taken this other person over there, this associate, instead of me as a partner who has experience.

29:43So I think it makes sense. And I also think that it was, I'm pretty sure New York's prior report said that. It may not have focused, honed in on it as much, but I'm pretty sure that it said that. Yeah, it could have been one of those things where it's like they felt like they had to memorialize it or they had or maybe it was like, you know, New York City said it and now the state is saying it. I think that's what it was. I think you're correct. I think it was New York. It might have been New York City. Yeah, or County, I guess. I guess Mark just posted it on in the chat. Yeah, I mean, I thought it was pretty straightforward.

30:16I thought it was. And to go to Steve's point, I mean, I would assume that if if if if the client and the lawyers were on the same page to start off with, like, OK, yeah, we'll. we'll cover this much of the time for you guys to learn how to use it properly. So you're not just like putzing around and like, and, and, and messing around with it without, without, without any, you know, clear objective in mind. We'll cover that. And then, and then, but then everything else after that was, well, you know, we're not going to, we're not going to do it. And I assume that would still be kosher. I mean, like, you know, if, if everyone's on the same page about it, I think this was more along the lines of like, you know, you know, you sending a bill to your client with this, this huge number on it being like, okay, well, we charge you for 10 hours, even though we only spent, you know, one hour actually doing the work, but we spent like six hours, you know, learning how to do it, then yeah, I mean, I could, I could see why a client would have a problem with that or not, you know, not wanting to pay that bill.

31:10And then I guess the other thing was interesting was that there was a appellate award in the 10th circuit, was it 10th circuit, where Brownstein Hyatt, they were awarded a$115 ,000 fee and they had used AI for some of their research and whatnot and the plaintiffs were challenging at saying that it was unreasonable but it got upheld and I think based on the opinion it was the reason why it got upheld was because the plaintiffs didn't get into specifics about what was used how it was used what time was saved or what they were being charged with that they were objecting to so I guess like a blanket objection.

31:50I mean, that's just the 10th circuit. I mean, other circuits are free to do what they want, I guess. But I guess a blanket objection without any specifics isn't going to fly. So you have to have some kind of knowledge as to how the AI was used, how the time was saved, and what the objection is as far as how it corresponds to the payment and whatnot. So I thought that was fair too. Because otherwise, they wouldn't just be, oh, well, you use AI. I'm not going to pay this bill. That wouldn't be fair either. Yeah, it's sort of like, just because you used AI, I should get a 10 % discount or 20%. I don't know.

32:24I mean, that wouldn't fly with it. I don't think a client could say, well, you know, Victor, your firm uses AI, so automatically I should get a discount. That wouldn't be very fair either. Although they may do that. I saw my firm's data. I saw my firm's data already. So yeah, you don't have to worry about that. Oh, so you got the 500 ,000. What are you doing on this show if you just made that? Well, I'm not getting a$5 ,000 check from the government now, so I had to make money somehow. Right. Anyway, yeah, so I guess that brings us to Nikki's Above the Law column this week. I couldn't think of a better segue than that.

33:04I think you're the only one left in the queue as far as stories, unless Julie had something. But yeah, can you talk about that one? Yeah. Yeah, I've just really enjoyed tracking the, it's interesting to me with technology how cases, you know, new technology comes out and then there'll be a wave of cases about one issue and then the new issue comes along and there's a wave of cases about a new issue. And a lot of the cases that are coming down right now relating to AI deal with the work product protection and the applicability of it in different contexts. And so I really enjoy, you know, tracking those cases and writing about them.

33:44And so for my ABA or above the law article for last month, I just wrote about a couple of different ones that have been handed down. And, you know, it's the cases I'll talk about, you know, there's a couple issues that the courts consider. is the person appearing pro se who was alleged to have put their some information into Claude or some other AI? Are they represented by counsel? If so, was that research conducted at the request of their attorney or not? Or did they ask someone else to do it? Because in one case, a party asked his girlfriend to do some research in Claude. I guess he couldn't be bothered to do it himself, and then handed that research over to the lawyer.

34:26So I just wrote about a couple of different cases. One of them was where the court determined that, and Morgan v2x, the court determined even though the publicly available AI tools often rely on user input to train their models, the pro se plaintiff had a reasonable expectation to assert work product protection regarding the AI use. um another one was another pro se defendant um and the court concluded that because the responses were prepared in preparation for litigation they were protected and then on a more recent one had to do with a represented defendant who conducted research of his own volition so not at his attorney's request and um the court agreed that using the ai tools didn't necessarily act as a work product waiver and said it could protect some but not all of the AI assisted research and you know went through and looked at the different categories of research but it's I think it just raises the point that and actually yesterday when I spoke with the New York City Bar Associations it raises the point and it solves the problem so I spoke the New York City Bar Association's annual solo small firm conference and one of the things that a whole bunch of the lawyers were complaining about and a lot of their questions centered on this because it was a session on AI ethics and billing and how AI was impacting billing and ethical issues to consider.

35:56And it was how do we bill for the AI slop that the clients keep giving us? You know, they give us, you know, stacks and stacks of AI created documents. They haven't even read them. Then they expect us to read them and respond to them. And then they just run it through AI and then they respond to it again. And it's taken up so much of our time. And what are we supposed to do? How can we bill for our time or how do we set expectations ahead of time so there's a lot of focus on that and the recommendations that come out of these cases in some ways solve that problem because what they say is you need to talk to your clients at the outset of every case about ai usage and how you inputting information or someone that you ask to do it um how it can impact your case in terms of attorney-client privilege and work product protection waivers and you have to set these expectations up front make sure you're on the same page about what research is being done how it's being conducted what they're putting in there and i think that that category of cases in this expectation that you need to talk to your client helps to solve that other problem and this is what i suggested yesterday you have these discussions up front but you don't just say stop giving me all this crap i don't have time for it or i'm going to charge you all this money because you're taking up too much of my time which feels kind of confrontational with your client and instead you talk about it around the concepts of protecting privilege and waiving work product protections.

37:20And then you're talking to them about it in a way that helps them understand how they're going to protect their rights in their case and why it's so important for them to communicate with you about these things and talk about anything that they may potentially input into these tools. And so it helps you have some control around it, but have some discussion about it in a that feels a little less confrontational with your clients and a little more productive. So I think this set of cases is both helpful in terms of analyzing those issues and also giving lawyers some construct in which to have these conversations with their clients and deal with this emerging issue that's a big pain point for them.

37:54Yeah, I thought your article was really well done, Nikki, and it's, you know, it highlighted to me that it's kind of an unsettled area of the law. I think the more legitimate decisions are geared toward more protection of the privilege rather than less. I think the Heppner case, at least on the work product rounds, was not correct. But it's still out there. And so, you know, you're sort of gambling that, oh, you'll get a good judge that understands all this versus some state court judge that doesn't have time and nor the proficiency to do it. One thing that I was going to note, I know a lawyer who primarily represents a plaintiff, so they're individuals.

38:37And he sets out in his engagement letter their commitment not to use AI tools about anything related to the representation or the case. And if they do, it could be grounds for termination of the relationship because he's just gotten so sick of having to deal with all this and try to clean up. And it's not just that he has to spend time reading the slob. He also has to clean up the mess. And now you've got all this stuff that could be discoverable. And if it is discoverable, now you've got to figure out how to deal with it. So he just said, I'm done. If you go out and look at ChatGPT, then find yourself somebody else, which you can't really fault him for.

39:21Yeah, but it's not really realistic, though. I mean, in this day and age, it's like everybody's going to use, you know, these tools in some manner, I guess. But, I mean, one thing I thought was interesting about, you know, what the article said was that, you know, it's good not to have blanket rules. And we talked about this on the show where, like, sometimes pro se clients get, pro se litigants, sorry, they get treated, you know, with the same, they get painted with the same brush like with like as like you know seasoned lawyers like you know courts have these have like these blanket rules about like you know what you can and can't do and that might not be fair uh for for especially pro se litigants who might not know any better and so i thought that you know the the articles that that you said nikki that i talked about how like you know for pro se pro se litigants that there's a little bit more of there's a little bit more um more leeway for them because you know they don't they might necessarily have you know the privacy matters you know at the top of mind and whatnot i thought i thought i thought that was good to see that and hopefully hopefully we see more of that.

40:22I agree it's unrealistic to expect people to not use AI at all because historically you know in the online era clients have researched their legal issues using Google right and they land on attorney blogs or other websites or you know cases and laws and to tell them that they can't use AI means they can't Google stuff anymore either because those are ai results as well so i don't know whether you're telling your um uh so i think it's unrealistic to tell your clients not to do that and i think that but i do think that setting um uh restrictions or creating understandings ahead of time and making sure that they understand the categories of information that they absolutely should not put in there and also making sure that they understand that you're the lawyer i mean this is what they pay you for it.

41:16And again, this is something that I just thought of. One of the worst parts about hiring a lawyer is hourly billing, right? That's why you do all this other research, so that you don't have to call your lawyer and ask a question and then get billed for it, or send an email and they respond and then you get billed for the email, right? And so maybe this is going to be yet another thing that pushes towards flat fees. And that was another thing that was discussed yesterday was both the ethics of flat fees and also how do you set reasonable flat fees and how do you set flat fees that are, you know, hybrid, you know, up to a certain point.

41:50And then if this trigger thing happens, then it's back to hourly billing or else you set a new flat fee to cover that aspect of the case. But I do think that it's going to start a lot of these pressure points, you know, trying to prevent the AI slot, but being able to be responsive to your clients without them feeling like the bill's going way up when they know if they use the AI, they'd at least accomplish something and they don't have to pay for it. So there's, there are all these weird pressure points and conflicts and it'll be interesting to see how this all kind of plays out over time in terms of billing and also you know protecting all of these rights and privileges that are available too yeah i'm just thinking of health care and how people will immediately google some kind of prognosis or word that the doctor or nurse uses after they leave the appointment it doesn't We cannot police.

42:41I just was talking to somebody from the UK about this who's been working in legal AI for decades. And she's like, you just can't police consumer behavior anymore, Julie. I'm tired of trying to do this. This is just like so completely unrealistic. Even with flat fees, they will still do it. Like, even if you tell them you're opening the door, they will still do it. They're still going to do it. I really feel like I do I know what the answer is and all the questions you just had about flat fees I've written about this and then there's subscription services there's so many different ways to bill for legal fees and charge for legal fees and there's so many cool interesting people who have been doing this for decades that work has already been done and if anyone's watching and it's like I want to learn more go learn more you you will go find those people but in terms of this, like people are going to continue to use Claw, Chachapiti, all that.

43:38And AI tools are in everything. Like we were making it sound like it's a specific place to go, but it's in everything everywhere. So I don't know what the solution is, but even as an attorney being like, okay, this is what you can put and not put, it doesn't, you can say all those things, just like you would go to see a doctor and they tell you what you have and they tell you not to Google it because they're going to call you back and do that and you do and you google it anyway like it's just going to happen so we're going to have to figure out from an ethical perspective and an evidentiary perspective i'm working on an article right now about ai and criminal justice there's a huge report that just came out in ontario this is why i don't have an article this week because i'm knee deep and reading a lot about evidentiary evidence in criminal justice and how ai is completely disrupting that whole thing.

44:29I think this is the same case. Like, you know, a controversial take is going to be, um, how important is it to keep all those things confidential or privileged that the things they put into chat GPT, just like I'm using the analogy of the doctor saying something to the patient and they go and Google it. It's just, it's going to happen anyway. So I think our rules kind of either need to catch up with this but this whole idea of the clients changing their behavior is not going to happen they're going to continue the behavior so as as lawyers you know and i agree with steve like like you know you say okay i'm not going to have clients who do that i don't know which clients don't even my parents use chat gbt like i just don't know who doesn't do that now so anyway now i'm rambling but i i liked your article was really good and really made me think like i said i'm working on this piece on ai and criminal i'll put the i'll put the 132 page report in the chat and you guys can read some really scary things that are happening in the uk us and canada related to criminal justice but yeah but i i i i worry that we're gonna get stuck on this point of putting privilege information in ai it's already there so we move on and then talk about like what we're gonna do about that in terms of protecting the client or protecting justice or protecting procedural fairness and all those things.

45:55You raise a good point, Julie. I have a friend who's a public defender, and I had this discussion with him because I said, are your clients doing what you've just described? He said, all the time. Drives me nuts. And he said, but what I figured out is I have to be able to tell them why what they put in there is A, dangerous, and B, probably not right. I mean, it's one thing if they bring you something and you say, oh, I hadn't thought of that. That's a good idea. Most of the time they bring you something and it's a terrible idea and you have to convince them why it's wrong. And so he says, I've sort of adopted the approach.

46:32OK, tell me what you ask. Now let me ask it this way. And here's what we're going to get. And you can see how wrong. And I thought that was a pretty interesting approach to it, which sort of puts some of the burden back on the lawyer to be knowledgeable about these tools and where they can go wrong. One of the most interesting conversations that organically happened, and it was last year at a conference, was an argument that a couple of people got into about access to justice and AI. And the takeaway was, you know, like the evangelist was, it's going to be great, like it's going to open up all these opportunities for these folks to learn what's going on.

47:16And somebody questioned it. And I thought their point was really good, which was pro se litigants now have been able to do this forever, much like Julie's example of the WebMD Googling your whatever. But they always felt like in the back of their head, they kind of knew that they might be wrong with their research. like they were finding things, but like, maybe I'm wrong. What these AI tools do is not be much better than that research, but say it in a way where it's conclusive, the mansplaining as a service moment. And what this woman argued was that indeed, we're going to see a backward track of the quality of legal self-help resources that you could get.

48:03It was interesting. And that kind of stuck with me now i think following that up though i will say i went to everlaw's show last week uh and jim sandman was there and uh talking about access to justice stuff and one thing he talked about was that it really does depend on what you're asking uh there are questions about how do i deal with my landlord turning off my heat where chat gpt is a way better resource than any kind of legal thing, because any kind of legal thing is going to jump you right to do the file this and do this, whatever. And chat GPD was like, first thing you should do is write a letter that looks like this.

48:46Here's a business looking letter that says this and like whatever. And he was like, there's there's moments where this interjects in a way that's actually useful, that takes advantage of its mansplaining as a service skills. And then there are other places is where it doesn't. And yeah, I just think about that being that as part of this conversation of just like, yeah, these folks are going to look things up and it's going to be problematic and it's going to make things a little bit worse. On the other hand, there are avenues where it can be better. And I think a lot of our task needs to be explaining to the general populace where it is a good tool for law and where it is terrible.

49:27you don't want to see that that that little bubble on his i was wondering about that too what happened i think you did a thumbs up yeah did you do that or uh i mean i have that yeah there we go yeah i have those turned on where you can do the whatever i did i did must have done it organically i i didn't think about it but yeah there there are with certain max or certain ages of computers. I think Windows does it too at a certain point. Zoom allows you to do that and where if you do certain symbols, it'll blow things up. I think this causes... Yeah, there. Oh, there you go.

50:09I didn't know if someone was giving you props or I was like, oh, okay. Joe was giving Joe some props. There you go. Yeah, that sounds about right. That's the only The only person who does. Yeah. So did you did you have any stories this week, Joe? I did not. No, I. I have been very much down the rabbit hole of it turns out there was an election this week and I might have been newsworthy. And it has it has precipitated a lot of legal stories for me. So that's all I, yeah, and a surge in the market followed by a crash, followed by some recovery. And that market crash was mostly on the AI tech side, which is going to be blamed on the fact that they think Democrats coming in will crack down on data centers.

51:06But in reality, it was probably eventually going to be the sort of thing that happens when you've got entities like Blackstone saying, Anthropic, we're rating as high junk. So when that's happening, maybe this was bound to happen eventually of market correction. Yeah, my tech stocks took a beating this week. yeah yeah yeah i would steer like other than like apple i i'm not i'm trying to get i'm in a bunch of funds because i started as a lawyer so i didn't do individual stocks because you can't do that as a we were told as lawyers and so i always had funds uh and i am really now working to try to find a way to get all my funds to be purely insulated from the inevitable ai crash um yeah Well, that's why you got to sell your own law firm so you can sell your data then.

52:01I definitely need to sell my own data. They are welcome to my data. There's not a lot of confidentiality to it. All right. Well, that's where you fall down, Joe. I mean, you said you're welcome to your data. Your mindset should be, my data is really valuable, and I'll give you a discount on a price. It's not$500 ,000. it's only$400 ,000 what a deal fair enough they are willing to get my Bing profile for$400 ,000 or whatever alright listen that's the hour probably back next week hopefully we won't have any more technical difficulties yeah so sorry to all the viewers sorry to panelists there was a mix up with you know permissions and things like that so we'll do that we'll be back next week I hope everyone has a good weekend thanks everyone bye