00:00:27.4500000 Ah, there's my people. Great. All right, we'll give it just a minute to get everybody in here. We've got 40 people in already and we have almost 80 registered. So we'll give it just a minute or two for everybody to get from one zoom meeting to another. 00:00:51.3000000 It. 00:01:48.4400000 All right, everybody, we got about 45 people here and I just want to go ahead and get started and be respectful for everybody's time. So a little bit about me for those of you who haven't had ACLE with me recently. I work for the Success Partner that is my own consulting company, working with lawyers and law firms. I've been running that since 2018, previously practicing attorney, running a small law firm. And then about three years ago I started delving into AI, mainly talking about it, mainly learning about what was rolling out ethics wise and practice practice wise for attorneys. 00:02:24.6600000 And then two years ago I went ahead and became a certified chief AI officer. And this last year I started working for them as well. So I work part time as a fractional chief AI officer embedded in 20 million to $50 million regional construction companies. And so you can imagine that that's going to be a little bit of a different take on AI and the usage of AI and the adoption of AI than in my regular practice, my own consulting practice where I work with mainly with smaller size law firms, so solos up to mid size 30, 40 attorney law firms who have a much different approach to this. So I'll say one is more full speed ahead. 00:03:07.6500000 And then now after kind of adopting it and realizing what they want to do with it, now we're talking about some governance and safety. Of course as a cio we always talk about it first and make sure everything's set up and but as the, on the law firm market, we're looking at that like very, very first as in, first as in six or eight months before even deciding to adopt anything. And then there's a wide disparity on whether or not it's, whether or not it's being used, how it's being used, the shadow, shadow use of AI and, and such. But we'll, we'll get into that again. I'm Daya Naif. 00:03:43.5500000 If I didn't introduce myself fully, feel free to jump in. This can be very much a discussion conversation. I recently put on a presentation for the Louisiana Bar association. We had 200 people and I said let's have an intimate, let's have an intimate discussion about AI here over, over lunch at that point. So we were all munching on Sandwiches, I wasn't. 00:04:05.6000000 But they were all munching on sandwiches and we were talking, we were talking about AI and there was a good bit of audience participation. So it can be done. It can be done. So this, this program is to really give you an overview. So an overview of AI and again, it's been something I've been talking about for several years and I've been watching the audiences adopt and adapt and change in their perspective of it. 00:04:32.8700000 I'm going to make sure everybody can see my slides. Right. Let me just. If you can't see my slides, hop on and tell me because that would be a bummer. 00:04:43.2000000 Okay, I got a thumbs up. All right, great. So sometimes. Because sometimes that happens. Right? 00:04:48.1600000 So we're going to spend the next 60, 55 minutes now together talking about ethical obligations, competency, responsibility, to navigate this properly. So our learning objectives, you will see are not, hey, let's look at a whole bunch of cool tools or let's look, dive deep into the sanctions and responsibilities. But it's more going to be overview. So it's not science fiction. We're going to look at where lawyers are actually using AI, a little bit of ethics and guidance around the, what the American Bar association is saying about it because they'll set, they set rules and then the states kind of adopt those rules or not and use them as guidelines and eval. 00:05:29.1200000 How to evaluate tools. This is one of the biggest questions I get is what to choose and there isn't, there wasn't a one size fits all answer on that and then how to integrate AI without compromising quality. So that's what we are really dedicated to. Our practice, we spent a long time getting here and want to, want to keep this career until we, until we choose, we want to do something else or choose to retire. So social media has us completely hyped up about AI and that's not what we're going to do today. 00:05:58.8600000 We're going to be very grounded and have a very mature discussion about it. 00:06:06.7100000 So as I mentioned, this just kind of a brief flow of action here and also take, try to take some notes because there's going to be an action plan at the end. So if you think of things as they're coming up, where you have questions for me or you want to evaluate some tools, things pop into your head, definitely take a note and we can follow up at the end with an action plan. 00:06:30.2300000 So just kind of cutting through the hype. AI doesn't think and I know it says it's thinking or reasoning, but it actually is a prediction model. So it's a prediction model, but it's with a great amount of information. So when we're going about our normal work, having a normal day, or having a conversation with our kids or our spouse, we often finish their sentences. We have predicted what they are going to do or say based on what we know about them and conversations we've had in the past. 00:06:59.6500000 So if you look at it, we are also predicting, we're predicting patterns in conversation, but we are working with a very limited amount of data. We're working with, again, the conversations in the past, the context, you know, are they going out the door and forgetting their lunch, you know, what's actually going on. But it's. But it's pretty limited to our experience and our interaction. AI, on the other hand, has access to everything. 00:07:22.1200000 So it's you being able to predict based on having access to access to a very, very wide, inhuman amount of information or superhuman even amount of information, which can be incredibly useful. It can also be incredibly dangerous. So forgetting, if we forget what it is actually doing, it's very confident, even when it is wrong. So if you think about a law clerk or a junior associate, you, you will understand. You understand that. 00:07:51.1100000 So they've presented the best theory ever, and they're of course dead wrong. And then it's your job as a leader to decide how to teach them on that. So AI didn't arrive overnight. We feel like it came in this sort of big tsunami. Even now, even today. 00:08:07.3600000 I've been talking about it for three years, but even now, people are like, what is it? I'll get on. And it's like, what's a GPT? How do you make a GPT? How do I even access ChatGPT, which is the most sort of universally accepted model. 00:08:24.3200000 Now, of course, you can get Gemini with Google. So if you're a Google user, they've sort of already plugged in Gemini. So you're using a lot of AI without even knowing it, without even knowing what you're doing with it. And you're not using it as an AI tool, you're just using it as a search engine like you always did. So it's creeping in, it's creeping into everything. 00:08:43.4500000 And that's another reason I harp on people having policies, because people are already using it. Whether you know they know they're using it or you know they're using it, it's happening. So the training and the policies are really, really paramount to success in our profession now. 00:09:02.5700000 So what is AI? It's really, it's making, it's, it's mimicking, it's mimicking being a human based on everything that it's learned. And it's a, it's a, not a rule based software. A rule based software is like if the input equals output and there's just a certain amount of constricted framework that it goes to. You know, AI teaches itself, it starts to learn, it iterates it. 00:09:27.5700000 The more conversations you have with it, and by conversations I mean like prompt at prompt and response and iteration. The more you work with it and the more guidelines you set or the more context you provide, the more accurate it becomes. 00:09:42.9500000 So here's a brief timeline. We started out early in the mid to 2010s, mid 2010s using it for ediscovery. So your big ediscovery platforms were already using a version of AI. We didn't call it AI back then, we called it Technology assisted review. And the more data we were getting in, the more data had to be sorted. 00:10:08.0600000 And so these systems had to be built in order to sort the mass amount of data. And then the more data you could sort, the more data was requested. So we saw those times when it's like produce everything you have. Well, you couldn't, you'd fill up a whole data, not really a data center, but a whole data area with a large discovery case. And now they're saying okay, these, they're targeting it more so that we don't have to have that huge expense of space, doc space that often comes with doc review providers. 00:10:39.0500000 And then we moved into a lot of research, a lot of contract analysis. The large firms started buying and developing software in the early 2020s to build in house machines, build in house machines to review and to do processes and to move things along. So some of your larger carriers have been doing that. Then ChatGPT came into the OpenAI product came into the mainstream. It was inexpensive, it was marketed, people people had it. 00:11:11.3100000 So the general public was using ChatGPT before many companies were using it or rolling it out to their employees. They were kind of still going through that. We're still kind of going through the fits and starts of what companies are using and what they're ruling out, what they're using to and what they're allowing in companies. But even as early as 2023, 2024, we saw employees using, using AI at work without any policies or guide rails, guardrails and what they were doing, they were just taking information from the company site, putting it on their Phones translating it back and, and using it. There. 00:11:50.6800000 There was a, there's a famous case where Samsung employees actually was, you were using I believe it was chat GBT to corre and in so doing released corporate secrets. You know, the secret sauce on a, on a particular Samsung product was no longer protected by intellectual property because their own people and in house put it out into the, into the world. And again it's not wholesale going out into the world. It's not like you could now Google it. It's now that you that the pattern recognition is there so often give the example of like peanut butter and jelly. 00:12:27.1600000 So if you were the first person to put peanut butter and jelly together and you were writing a recipe for peanut butter and jelly and put it in there a couple of times or however many times it took to create a pattern, the models would then know, the data would then be trained that peanut butter and jelly go together. So it didn't wholesale steal your idea of peanut butter and jelly going together. But since it was in there and it recognized it as a pattern and say that had never happened before, then your trade secret would be lost. Last year in 2024-2025 we started looking at legal specific AI tools. And I don't know about you, but people are getting inundated with the tools that are out there. 00:13:12 We had Tech week in New York last week. I didn't get to make it to that, but I'm going to the ABA Tech show in Chicago next week for the first time, I'm ashamed to say the very first time. And so I expect to be just completely inundated with what kind of legal tools are being marketed and being pitched. Being that I'm already on a daily basis being sent emails and seeing stuff. Now of course this is my job so I probably get more than the average attorney. 00:13:39.4500000 But still there are quite a few that they get hit up with and it can be mind boggling as to which one to choose. Especially since they all say they're, you know, they all say they're the best, they all say they're the newest thing. They say often no one else has this. And really there is quite a bit, there's quite a bit of competition. 00:14:02.7100000 Large firms as you can imagine were the early adopters. It was AI was originally very expensive to have and to create on and host on your own servers. Then things have rolled out again to consumers. Most subscriptions are in Your, in your $20 a month range for a regular type subscription like Claude Claude Grok. I don't Use Grok I don't know how much is no, no disappearing I just don't use it but I know gra. 00:14:30.0300000 I know Claude is about can starts about $20 a month. Chat GPT starts about $20 a month. There are free versions but I do not recommend using free versions of anything for work ever because if you're not paying for a product then you are the product and of course you need to go in and select do not improve the model so so that it's not feeding on your data to make them make the model better. So you want to set that, set that initial safeguard to now is it 100% for sure we know what they're doing with your data? No, we would, we would never know. 00:15:04.9200000 We would never know. If I don't know I'm being maybe being watched by satellite right now. I don't know even if I opted out and said don't watch me by satellite, how would I actually know? So it's something that you do the best you can and you work with it. And that's one of those things that you just go in and check all the toggle off everything that's toggled on so that you are safeguarding your data and your client's data also have different policies and different clauses that you can put in your engagement letters to to cover notification on that. 00:15:40.1800000 So this I get a lot I won't replace lawyers but lawyers with AI will. I give you a second to to think about that. 00:15:55.9100000 Some other speakers I'm working on projects with disagree with me. They think that lawyers will will be maybe I would pay but some lawyers will be or some tasks of lawyers will be 100% agree. The CEO of Microsoft AI so what's his name? Mustafa Sullivan. Sullivan. 00:16:17.1400000 Sullivan said it then 18 months our profession is pretty much kaput. So he said all professional services will no longer be necessary within 18 months of the start of 2026. So coming up pretty quick 2027, 2028 time frame some people are predicting that AI will reach human. 00:16:41.5400000 So you're talking about lawyers, doctors, accountants, other types of knowledge working professionals. And I think, I don't think it can roll out that quickly specifically from the human the human pushback on it now whether it will replace like direct to consumer type things. We've already had an inching away in the industry via technology via DIY forms. You know we used to you can get Zoom Zoom Lawyer to make forms for you and contracts for you and set up your llc. So we're already seeing these or there's several companies that will fix your tickets for you or fight your tickets for you. 00:17:22.4200000 Some will do insurance claims for you. So we've already been seeing a chipping away at the legal practice through technology and through other, and through like claim, like the ones that file insurance claims for you. They call themselves claim administrators or claim processors as opposed to, as opposed to lawyers. And you know, they have with or without legal supervision. So that's kind of part of a whole industrial shift that's been happening over the last generation or two of lawyers I'm seeing. 00:17:52.7900000 And so this is part, part of it all right, so we're going to look into some practical use cases because in my work as a consultant, a coach and working with lawyers, I want our lives to be better, I want our practices to be better. And at the risk of being a little Pollyanna, I think better lawyers doing better work makes communities overall. And so there we have it. So if we can do things better, faster and more accurately, saving time, then we can spend more time where things are important for, for lawyers, the strategy, the client management, the actual time in, in trial or settlement negotiations and then also learning our craft, you know, still becoming better lawyers without with taking some out. Some of the tactical day to day that can be automated. 00:18:44.2900000 Research has been one of the first things that changed, at least for the private practitioner. And this is where also things can get messy. This is where things can, they can mis. Cite, they can hallucinate, which is basically jumbling the information to make the answer, to create the right answer. Again, it's trying to please you, sort of like that first year associate. 00:19:07.8100000 And it can grab information and put it together and create sites and sources that don't exist with the end goal of matching up with the facts that you gave it. So some of the common softwares out there, of course we've got Westlaw and Lexis, which you know, they sold to us while we were in law school with those pens and notebooks and highlighters. And so we always will remember that brand recognition. So myself as a small firm lawyer, I always just used Fastcase, it's called Velex now, which was provided by the Bar association as my, my research tool. And then of course Google and then also asking other attorneys and also looking in any databases that I had compiled or the people I worked with had compiled. 00:19:54.4400000 An interesting predictive analytics tool that has come out. Trellis Law is very interesting. You know, again, as an attorney I would get a case in front of a judge or with a different opposing counsel. That I hadn't experienced before. And so I would ask the other attorneys in my, in my office who's, who's got some intel? 00:20:12.9800000 You know, how does this judge usually go on these type of cases? Or what do I need to watch? You know, what kind of pitfalls do I need to look out for? Or which ways can I, which ways do they like the information presented? And then same way with opposing counsel, how long, how likely is this opposing counsel going to, is going to meet with me and work something out, or do they always want to go to trial or always wait to settle at the last minute? 00:20:36.6300000 So getting, getting that intel now. All right, so I have a little bit of a demo for co counsel. This is one of my attorneys and my friends in Maryland. So we'll play a little quick demo because some people haven't really even seen the research tool. The first thing you do is type in a question. 00:20:57.5900000 And I will go through ways to ask questions particularly. But the one I chose here is can a case be removed to federal court based on diversity jurisdiction after a complaint is amended? You can either hit the little paper airplane icon or just hit return. Now what it's going to do is it's going to ask you a prompt. When you type this in and the prompt is right here under AI Assisted Research US it's asking what sources do you want to answer or use. 00:21:29.9900000 There are two primary sources you can drill down from these. I suggest that you don't and I'll explain why. But the first is AI Assisted Research. And that is going to research all of your databases, all cases, statute, everything, so you don't have to select like normally when you do research, you say all states or you say all Feds or you do statutes. This will search everything for you. 00:21:55.7500000 Or you can select search and summarize Practical law. If you're doing a particular pleading and you're looking for drafting suggestions, Practical law is useful because it will give you the practical law database of all of your secondary sources forms Westlaw's practice toolkit. Historically, personally, I haven't found these to be all that useful to me. I'm so used to drafting things and preparing them myself. But on occasion, if you have a case that you're not entirely the subject matter you're not entirely familiar with, this can be useful. 00:22:32.6100000 I 9 times out of 10 I'm working in AI assisted research, you just click the black box and it will begin to search the system. You'll see the the spinning circle of death. While it's doing it, you'll also see down here grayed out, reading 10%, two to five minutes. It does take time. You have to be patient with the system. 00:22:58.9200000 So this is part of my podcast series, so the whole video is on my YouTube channel. But that was just a little snippet. Some people are not, have not seen one of these tools in action. 00:23:13.8800000 All right, so document drafting and review. The AI Document drafting is going to be your first pass. And I have some people that aren't going to use it. They're like, no, I'm not going to use that. I'll use it to read or review or summarize. 00:23:26.8800000 But I always want to start with, with my own document drafting. Something you could also do is, if you want it to sound more like you or be more like you, is create your own database of documents that it can draw from. So instead of sounding like generic AI language, even using these document drafting tools, then you can, you can teach it, you can build, you can actually build your own library. And this is, I find this is amazing for training new associates and especially some of you who may be retiring soon or stepping into part time practice, then that would be something that would be very valuable to have. So an associate could go to the, the AI assisted, the bot bank, the mini me, actually I call mine my second brain. 00:24:16.7200000 But some people call it a mini me or clone or a digital twin. I've been hearing a lot of terms from that. Creating your own institutional wisdom in a AI assisted tool to share with people in your organization, certain people in your organization, so they have access to a version of you that's not your own human time spent. You know, they're grabbing you down the hall, hey, do you have a question? Or showing up in your, your doorway shadow, shadowing your doorway day in and day out. 00:24:47.7000000 You could actually say, did you use the tool or do you go through the tool and actually use it for training purposes as well as institutional knowledge. Caption so again, it's the idea that it's going to get you started so you can come in with some ideas on things. You can make your own GPT. I made it a law clerk GPT that I mainly just use for demonstrations. I don't practice a whole lot anymore. 00:25:12.4900000 But you want to look at accuracy like in a, in a live audience. I would do some call and response here about things, but what we're looking at mostly is accuracy. Looking for fake citations, sometimes tone. People are very attached to their authorship and their verbiage when they write. So I was never thrilled about writing. 00:25:37.1100000 So Creative writing maybe, but legal writing. Nah, just want to get my point across and so not, not as worried about that. Confidentiality, of course, is a big one. Where's this going? So you have to redact a lot of your documents before you put in, put them in there just as an additional safeguard. 00:25:55.7500000 Most of these system that you subscribe to do have protections. Do have protections and do have certificates, security certificates that it's not, they're not feeding into any models that they're using. But in order to double, double, double what I would have in your policy is the redacting information, any type of sensitive information before you put it in any software, even if it's a proprietary software. Unless like local. But that's way in the weeds. 00:26:24.5100000 We'll have to do a whole nother seminar on that one. Client perception. So this is, this is a big one. So perception of using AI, client's perception, opposing counsel's perception, the court's perception. You know, again, the jurisdictions are all over the charts on this. 00:26:42.2300000 The local rules are all over the charts on yes, use AI, don't use AI. Keep hearing, I keep hearing judges speak and say don't use it at all. Number one, that's not possible. Again, it's baked into, it's already in our email. So I don't know if you're using, I'm assuming most of you use Outlook Lawyers by and large, majority like 90% use Microsoft products as then the rest use Google products or Google Workspace products. 00:27:13.4400000 If you're practicing and you're just using a Gmail like smithlawyergmail.com you need to get a Google Workspace. So there's zero protections against an actual Gmail address. So your documents and things are all being fed into Google. But if you have a Google Workspace account, less of your stuff is being fed into Google. So definitely want to get that as a very, a very small upgrade. 00:27:40.2900000 If you're still, if that's still, if that's still you. If I'm speaking to anyone or someone, you know, asking for a friend. Right. So that, and then at some point there's going to be a shift from are you using AI and you're afraid that if you say yes, that's the wrong answer, there's going to be a shift to are you using AI? And no is going to be the wrong answer. 00:28:04.2400000 So that, that tide is, that tide is coming. All right, so we'll do another quick little demo here. 00:28:22.1700000 All right, so what we're demoing now is a product called Gavel and it is embedded into my Microsoft Word here in the computer. So I have it here as an add on gavel accept. I've created a sample commercial leaked just for a demo here and I'm going to ask it a question here, extract key terms and present as a summary table. 00:29:10.6700000 So as you can see, it did what I said, what I've asked it to do here, right in the sidebar of the document. So let's look at something that we want to maybe contest or change. 00:29:32.4600000 Nice rent reduce by 20% as we agreed in negotiation. 00:30:00.5500000 There we go. So what we could do is put in our whole list of things that we agree to in negotiation for it to update. So here's our apply all here it is in the document with redlining so that when we send it back over to the other party they see what we've done. Yeah. Now what's also useful, I did a longer demo where I created some little avatars and had them actually negotiating the contract on a mythological video conference where they could actually put in the transcript of the conference on a break I guess and then have it do the redlining, take the transcript, do the redlining and then literally instill the same exact meeting, go in and make some of those changes. 00:30:54.9400000 So just speeding things up quite a bit. 00:31:00.7400000 All right, so we already touched on discovery and E discovery. So how they really had to get this technology dialed in to stay competitive in the market. I used to as a young lawyer do outsource document review myself and it was bankers boxes and post IT notes and that would do my own summaries and things. And you know, it quickly became more of this, this at least for larger scale projects, you know, large claims, mass torts, toxic torts, things where there was a lot of data or technology, IP theft, things like that. There's just large amounts of technology held in data and not good old fashioned paper files. 00:31:48.8100000 And that just had to be done. And that's when this really got dialed in in the legal profession and really got to be household use, so to speak. And so we were using that technology assisted review back in like 2010. So now what is that, 15, 16 years ago. And so predictive coding they would train on sample, sample data, sample, millions of responses, privilege, they would put look out their things that were privileged. 00:32:16.8400000 You could really code it in. And even now in audio and video analysis you can actually search video. So for people who have volumes of say security footage, cctv. I don't know if anybody watches British, British crime stories, but my mom and I love Them and so always like review the cctv. Review the cctv. 00:32:36.6800000 So there's, we know there's like, like tons of stuff and then you see, the next thing you know you see an officer spending all night, you know, going through all the tape. So there's something that AI can, can do now. You can put in a set of parameters and have it. It's called visual, like visual AI, which is also the term that took out the technology for ocr. So ocr, when you're scanning documents, that was, that's the older technology. 00:33:02.0700000 There's now what we call visual AI. And so you can search minutes, you can search hours and hours of video in minutes. Now do you get everything? Maybe, maybe, maybe not. But you do get to what you're looking for based on your own interaction, your prompt, your search parameters. 00:33:21.0900000 A lot quicker than literally just sitting there and using the human eye, which gets tired and strained and sometimes it's not reliable. But the AI can really shorten that time, time frame. Compliance and risk management. So I use this a lot in the construction companies that I'm working with. Anybody who represents them knows that they work over a large geographic area in each county, municipality, type of, type of entity, Department of Transportation or Department of Health and hospitals or something like that. 00:34:00.0100000 Those are always changing. They're updating regulations all the time. And there could be conflicts between what the state said versus what the county said or what the municipality said. And so just putting in some alerts. It's been very helpful to kind of keep up to date with things happening in, in real time as far as risk management goes. 00:34:20.9600000 So any of you guys and gals out there who are general counsel or do consulting in that type of work, then that would be something to do. Also predictive analytics on financials. So if you're thinking about taking on a client and you want to run some due diligence on them, then you know AI has AI and AI deep research as well as data. AI data analysis programs are very good. They used to be, have to go out to a due diligence practitioner or evaluation expert and weeks would go by getting in their backlogs as a very, very specialized science. 00:34:57.0900000 Of course, if you want an opinion, an official opinion, of course it goes still through a human expert. But for you to, and if this is the type of work you do, for you to invest in a subscription based AI program would be, would be very helpful. You could, you could literally see the future. You kind of have your own business, crystal ball if you will. 00:35:19.9800000 So just pointing out a few things for different practical applications. So it's like the tool, the tools are the same or built on the same backbone products, but they can be manipulated or taken into consideration for different practice areas. So if you're in litigation, you can look at judge behavior, opposing counsel strategies using historic data. Kind of like watching the playbook. Like if you're watching. 00:35:46.7700000 I don't watch football, football per se, but I do like old movies with like the football team, the comeback, the comeback story or the Cinderella story. And you know, what's the coach always doing in the, in the room? He's watching their videos and he's watching the opposing team's videos and he's doing the X's no's on the whiteboard to really get honed in on, hey, how do we, how do we come at this team? What is there, what is their, what are their tactics? What have our tactics before used have worked and like why do they always get us on this particular play? 00:36:18.0900000 Well, because it works. Because that's the way they, that's the way it does. And so using that you just think that on a large scale of data that's able to be accessed and there's a couple of states, there's only two states that I know of on the top of my head, Alabama and I think either Utah or Ohio, they've not allowed these external data providers access to their records. But the rest of the states, there you go. So these, these subscription companies are getting access to real time data in, in the other whatever, 48 states that allow it. 00:36:54.3400000 There's only two, I just can't remember which ones. Awesome. So let's see. IP and trademarks. This can really help. 00:37:01.0600000 You're still going to want an expert to do your applications and your defensive applications are processed execution of applications. But to do some preliminary research that can be very helpful. Real estate doing title review. I used to do title title reviews constantly. That was a large part of my practice. 00:37:21.7800000 And just the ability to read the documents and then read, I'm using read in quotes, read the documents and catch errors and also read aloud. Another attorney I was talking to recently said she loves to have legal descriptions read aloud to her because she can catch the, catch the error, read it on the one set of documents and, and have it read from the typed in documents to catch that because that is one of the main sources of errors in titles. Just having that legal description not be, not be accurate. Just a couple of little things off here and there because it's so tedious and if you're using old records, like I was using a lot of old records and doing curative work. There'd be stuff like handwritten. 00:38:07.4400000 There would be white out that had kind of dissolved off. There would be something in italics that may not be in our regular fonts now. And I'm like, does that, does that mean a foot? Does that mean. Or in Louisiana we have French foot. 00:38:22.0300000 So that's two. That's FFT. So I'm like, is that two Fs? Is it one? Did they just press really hard on the keystroke? 00:38:28.1900000 I don't know. So really being able to speed up that process for the human review family law, generating settlements, analyzing documents, saying, hey, where are we actually off on this? Here's this huge amount of property, you know, movable property that they've put into the record saying that they own it, personal property. And then here's this person's sworn statement list. Where do they interact, where do they say, where are they conflicting? 00:39:00.6800000 And you can immediately say, okay, look, this is like, like 90% the same. Here are the things that we need to look at so we don't have to go through each little set of china or whatever leaf blower that's listed in the sworn statement. Immigration, which is very date sensitive, you need to know when people are eligible for certain things. And you can set up automatic reminders. You can set that up. 00:39:30.5600000 You can also monitor policy changes because we know that's, that's an area that has changed minute by minute. I was just talking to an immigration attorney earlier this week and he said my advice is only good for the day. He said, I used to could say within the next six months, my advice is good. And he says, my advice is only good for today. Things that change quickly, regulations that change quickly, you can track and monitor. 00:39:59.2600000 All right? And so of course, no good. CLE does not talk about ethics. So this is something that is, that is huge. So we have now a new tool at our disposal that is getting some people in a lot of hot water. 00:40:15.2600000 And it's something that we have to think about how we work. So we've had a duty of technical competence in our rules for over 10 years now. And this is something that would fall, that would arguably just fall under that, that, that duty of technical, technical competence. I also think it falls under duty of supervision because AI in and of itself is sort of both a tool, but it's also being used as an assistant or as a junior lawyer in some cases. And so we really have to double down on that. 00:40:51.3400000 Duty of that, duty of supervision. Like how, how we're using it, how our associates are using it. Yeah. So I hope by now everybody's read Opinion 512. I think it's been out two years now, and it really, it really boils down to that competency. 00:41:09.4000000 And I've got two more. I'm seeing how many? Too many. Two more video clips. So how much more I want to talk, how much more I want to share, but really be up to date on that ignorance of AI is no longer defensible. 00:41:21.9600000 So you really have to understand how it works, how it's can improve your practice and being aware of its risks and then communicating those with your clients. All right. This, I believe, is Dan Siegel, who worked on that ABA 512. So here, here's a quote from him. 00:41:46.7300000 Generative AI is a great tool for lawyers to use. If they're not using it, they're really making a mistake. Once you start to use it and find what it can do, it's remarkable. But that said, you don't go in and do it at full blast right away. You wouldn't, you know, you wouldn't drive a race car if you've never ridden in one before. 00:42:10.4700000 You would slowly go in and you have to do the same thing. You act reasonably, you learn what the tools are, what they're able to do, what they're not able to do. And you start with little projects. I use them for some reason of records, for preparation of all different kinds of things. But I know what the tools can do. 00:42:30.9700000 I also know when they really are not doing the right thing. And that happens. And ethically, you have to just be prepared, act reasonably and jump in at some point. Because if you think it's going to go away or it's not going to be around in a year or two, sadly mistake. 00:42:50.5400000 I think we've seen that. I think it's no longer completely ignorable. I get more and more people saying that they're using it, they're looking at it, they're trying to make a decision on what to do about it, versus two or three years ago when at least half the room would say, no way, not going to do it, not going to use it for anything, whatever, not even going to load it up on my machine or even learn more about it. Arms crossed. Now, pretty much everybody is at least realizing that it is, that it is here to stay and is causing a fundamental shift in actually all businesses. 00:43:29.8400000 So not just not just professional services. And I can't say like a fundamental Shift in just our industry because it's in our all industries and our clients, you know, industries. So they're, they're going to be coming to you for more and more guidance on it. And so being knowledgeable is really something you have to, to have to stay up on. All right, so we have confidentiality. 00:43:51.1600000 So rules, Rule 1.6 does not disappear because we now have cool tools. So you really have to look at data storage. You just recently we had two data centers bombed in Dubai and that was a real live data. So it was, we think of data being in the cloud. Data actually is stored somewhere. 00:44:09.5600000 So even if it's in the cloud, it doesn't mean it's not stored anywhere. And, and so you have to look and see is my information stored in a risky area? Where is it exactly? Where's the backup? Most, most data is stored in at least two places. 00:44:24.8300000 Make sure that you make sure you know that. What's it being used for? Is it being used for AI training? Is it being. What's the security and deletion policies where, what about data breaches? 00:44:36.6700000 Definitely look into cyber security insurance if you haven't. Recently, I, I know a lot of the bar associations are getting reasonable policies, basic cyber policies rolling out. I know, LA Bar just sent me my renewal for my Eno and said guess what, you can now get cyber for as, as little as a thousand dollar basic policy. So something to be looking into. Even if you want to just look and get the checklist and say what do I need to be doing to safeguard, safeguard my system? 00:45:10.2600000 Because law firms are low hanging fruit to cyber attackers and they literally just go through and just go system to system to system. Kind of like a thief in your neighborhood would just go door to door, car to car to see if something was unlocked. So if you think that you're too small or don't have any protected information, then that's not a, that's not a good, good excuse. They're literally just picking, picking pockets as they go and anything that's not protected they'll pick up and take or use as ransomware, which is, which is becoming quite, quite prevalent. I've heard from other people in that realm, like they're very nice, they, they take your money and they, they give you your stuff back, but your stuff has been ransomed and it could shut down your operations for days or weeks. 00:45:55.9200000 And I don't know anybody who can shut down a business for that amount of time. 00:46:01.8400000 So you definitely want to look at that. Never assume AI tools are confidential. Very Carefully go over terms of services, actually create GPT that reads terms of services, terms of service because they're usually like 80, 90 pages. And I want to, I want to hit the red flags as quickly as possible. 00:46:21.4200000 So again, this kind of goes into that duty of supervision and I'm going to go ahead and play another clip for you. 00:46:33.3400000 Well, Council, again, the thing that I haven't never heard, I've heard of heard of AI creating a non existent case name and citation that simply don't exist. What I've never heard is that AI not only generates a case citation but then produces an actual false or fake or AI generated opinion, written opinion associated with that citation. So that, that's what I'm saying. I wanted to clarify because I don't want to be very different than simply not checking it and not seeing it. I don't want to double down on a technology that even that I understood not at all before I honestly thought it was kind of a super Google and that the danger was that you could cite something without properly attributing. 00:47:29.2300000 But I didn't understand it before. I understand a little better now. I have to take full responsibility that they that are artificial intelligence was used by my firm. I did not know that artificial intelligence was used by a firm or even really what artificial intelligence did. But I have to bear responsibility because I submitted this under my name. 00:47:46.5300000 I submitted cases, a brief that contain cases that do not exist. And I did not review the entire cases of the majority of those that were cited. Nope. 00:48:02.3000000 All right, there's more to it, but there's a little sticking point, so we'll move on. And I don't want to embarrass my fellow practitioner here, but this is a learning lesson, so don't be that person. You just misunderstood what it was and miscalculated that duty of supervision. And what I've been seeing in these cases is an associate prepares associate or paralegal prepares the petition and a very busy overworked partner reviews it. It looks really good. 00:48:36.3500000 AI is a very skillful writer. It writes very eloquently and very convincingly. And looked over it probably had a whole stack of things to review, signed it, filed it, and then the court or a clerk or even opposing counsel or someone in opposing counsel's office read it, checked it. Oftentimes they call, they do a nice courtesy call. There's civility, duty of civility in our practice. 00:49:02.4600000 They give them a nice courtesy call and invariably in these sanctioned cases they double down. They actually go ahead and produce additional AI generated materials and that's when you get in some serious hot water. So definitely keep, keep our professionalism in mind. Fall on your sword if you need to make corrections. Be gracious and thankful. 00:49:25.9900000 Sort of like if you, if you get pulled over for something and you and they let you off with a warning, be grace, grateful and thankful and appreciate it and go get whatever your expired tag or your tail light or whatever it is fixed that you need to have fixed fixed. You have that opportunity, so use it. And so that's what's not happening in these particular cases where sanctions are coming. And I'm seeing the amount of money being fined as being not very low. I mean not very high. 00:49:55.5700000 But you know, you have the embarrassment factor. You have being certified to cle. Again, when I give these in, in person, I'm like, okay, who's my person who got like sentenced to be here? But, but no. And it's, it's going to get worse. 00:50:08.4500000 So there have been suspensions, there have been multi year suspensions for using AI. Specifically in some of those cases that I mentioned where the attorney doubled down, there will be disbarment there. There are definitely some courts and you definitely check your local rules. There are some courts that are really taking a hard line with this and a hard stance. One of which wants you to produce prompt logs which I think has all kinds of issues, you know, with confidentiality and court, you know, where's the court going to serve? 00:50:41.0700000 You know, how's that and what are their data safety standards? So that's a whole nother, that's a whole nother issue that we can, we can chat about later. Disclosure, of course, disclose when it's better to disclose than to not disclose. I have some clauses for you for your engagement letters and how to discuss that. Some courts mandate a sign off on. 00:51:04.8300000 I've used AI. I haven't used AI. I've checked the AI again. It's, it's really redundant. You're under a blanket to tell the truth to the court and to check your work. 00:51:14.2300000 But now it's becoming a specific thing. So definitely disclose more than you have to now to jump into tools. I know we're coming up, we're coming in, coming in for a landing here. But this is something that I get questions all the time. What tool, what tool should I use day of like we're just getting inundated with it. 00:51:35.2800000 The mistake is thinking you're going to use one tool, you're going to use four or five tools, hopefully not too much more than that because the tools can clash. Clash. Kind of like, I don't know, the personality of your office. You might have people who clash. If you get too many people on board, then if you get too many products on board, you're going to get some clash and you're going to get some disparate information which you don't want, but you have to define what you want it for. 00:52:02.3100000 So people ask me, they say, I want the AI and I'll go to the office for the consultation. And they're like, okay, plug in the AI. And I'm like, no, no, no, no, no. We have to have a lot of conversations. We have to really dig in deep and roll up our sleeves to know for what do you want it to do? 00:52:18.7700000 How is this going to affect your practice? We have to do some pilots, we have to do some test runs, we have to improve your systems to get them AI ready. We have to look at the future. Like, what's AI going to be doing six months from now? That's going to change your entire process anyway. 00:52:37.9800000 So, yeah, we could speed up the process you have now if you even have a process. And a lot of small law firms don't have the processes they think they have. They have the institutional wisdom. They have a couple of quirky things that they do. They may have a standard procedure written down sometimes on paper, or they have office manager who does things certain, a certain way. 00:52:58.7400000 And you know, come hell or high water, we're going to do them that way. But then most people don't. They just don't. They just slip it under the radar. So we also have to look at cost. 00:53:07.9000000 Again, these tools are reasonable subscriptions, Subscription per seat tools, but they add it quickly. So who all is going to be using it? Who needs a seat? Who needs a license? When do we need to invest in training? 00:53:21.0600000 Before we get. Before we start putting the money out. You don't want to be paying for seats and then they sit there unused for six months because then you're at a ton of money and you're going to come back and say, daiya, this was a total waste, a waste of time. I was like, well, it's because we didn't, we didn't get on it. We didn't do the training, we didn't do the integration properly. 00:53:38.6500000 There were a lot of things that set this up for failure. Oh, don't want, you, don't want you to see that. Okay, did not do that demo. So we're skipping ahead. Vendor due diligence. 00:53:49.6100000 Again, I have some, some, some GPTs that I can share with you on how to check your vendor due diligence. You know, if they're not going to give you the scope before, or, I'm sorry, the terms of service before you buy the product, then that's not somebody you want to be dealing with. Training your team is huge. So your team is only as good as the weakest link. So if we want, we want to make sure everyone knows the processes and who is the human that's reviewing this in your process and policy. 00:54:20.5800000 There's got to be a human. They've got to be there in certain checkpoints in the process to cover it, to validate it, to verify it, to do data checks. And reviewing these policies has gone up from every, from quarterly to monthly. So someone in your office has to make sure your policies are up to date. Quarterly, quarterly, at least, or maybe even monthly, depending upon what you're using and depending upon how quickly it changes. 00:54:49.5100000 But stuff is changing really quickly. And they just dropped another ChatGPT upgrade like this week, and then they had dropped one maybe three or four weeks ago. So being that the top dogs in this market are so competitive, they are updating stuff all the time. And if you have legacy models of products like ChatGPT4, which is what I built on my original GPTs on like it, it got sunset and so. Okay, great. 00:55:16.2300000 So obviously I wasn't using, I, I used some of them, but the ones I, I wasn't using, I, they didn't update and so they're gone. And that happens, that can happen in a blink of an eye. So no, no foul on me because I wasn't using them anyway. So, but, but if you're running them on your organization, or maybe your organization is only using them maybe quarterly and so they're not updated because somebody's not using them all the time. A quarter is 10 years now. 00:55:43.5100000 So that's the way we're moving so quickly. So having those policies to where you can update it and stay on top of it. Critical to have the rules established, the approved tools, have some approved tools. Again, we're seeing firms that say, don't use any AI. They're getting either shadow AI, which is unauthorized use of AI, or they're not getting any hirees. 00:56:09.6500000 The associates that would be a good associate for their firm are going elsewhere. So we are seeing some law schools start to teach this, getting that. So it's going to, going to catch up in the curriculum. I expect the majority of law firms to be teaching AI in the Fall. So if you're looking for a new associates graduating next year, then you need to have some AI use in your firm. 00:56:36.0900000 You know, you can't, can't take away something. A tool would be like taking away our car and saying you have to walk to work. You know, now we just, you can't have it anymore. It's like, yeah, but we already know how to do it. It's already part of us. 00:56:52.3900000 I can send you guys this. I'm going to go ahead and click through to the end so I can get you guys your. My contact information and your code for the clerr. So just noticed we hadn't had any interaction lately and I don't know if Jama put the code in yet. 00:57:22.4300000 Okay, Jama did put the code in. Good. All right. And I'm going to give you the code again here in a second. Get back to the PowerPoint workflow integration again. 00:57:35.8300000 You got to know where you're plugging it in. And here's your action plan moving forward. And if you want to reach out, I have a worksheet that goes along with this so you can start putting in what you need to do day by day by day. So here's how to get in contact with me. You can scan the code. 00:57:53.7900000 It is for a zoom coffee. I love my coffee. If anybody's known me more than five minutes and you've probably already seen me drinking coffee today, you know I love coffee. So we'll be having a zoom coffee to continue the conversation. It's not a sales pitch. 00:58:06.6700000 I just want you to have more information. And then if you do want to have, have me come into the firm or talk further or give you some advice, I'm happy to do it. Here's the code. It is 27pxd. That's 27pxd. 00:58:26.4900000 You do have to be logged into your LexFit account under the live videos and enter that course. So I'm gonna stop sharing and then also put my contact information in the chat. 00:58:55.8300000 My contact information. Ah, pretty good. 00:59:04.1600000 Here we go. 00:59:10.7200000 All right, there you go. So you can save the chat and meet up with me later again to have more coffee and conversation about. About AI. It is my pro again. I do a lot of law firm consulting, but AI has been the, as you can imagine, the hot topic for the last few years. 00:59:29.5300000 So I do stay up on that through my work with both law firms and now also construction companies. So I can see it from a couple of different angles. 00:59:47.7700000 All right, Thank you again, everybody, for coming. Thank you for viewing LexVid programs. I always have a good time doing this and. And stay tuned. I do. 01:00:08.1800000 I do a couple of them a year. Also free to connect on LinkedIn too. I sent you my. 01:00:23.5500000 Did everybody get the code that needed the code? Anybody else need a code? 01:00:29.4700000 Okay. 01:00:39.0700000 All right, it's one minute after. Looks like everybody's heading back to. Heading back to work or wherever they're going now. 01:00:57.3700000 All right, I'll go ahead and in the meeting. Thank you all again.