Ep.25 | Protecting law firm and client data in the age of AI (Guest: Dr. Ilia Kolochenko)

Show Transcript

[00:00:00] Hello and welcome to episode 25 of Privacy Lawls where I Donata Stroink-Skillrud speak with amazing privacy professionals, and we have some laughs along the way as well. Today I’ll be speaking with attorney Dr. Ilia Kolochenko about how to protect your law firms and your client’s data in the age of ai. Ilia is a Swiss cybersecurity and cyber crime investigation expert, privacy law specialist, and a lawyer admitted to the DC bar in Washington dc.

He started his career as a cybersecurity professional and has over 15 years of practice in information security, auditing and digital forensics. Ilia is a fellow at the British Computer Society Fellow at the European Law Institute and Vice Chair of the Information Security Committee at the American Bar Association.

He is also the Chief architect and CEO of ImmuniWeb, which uses AI to simplify and cut costs of application security for over a thousand Cubs customers in over 50 [00:01:00] countries. So Ilia, thank you so much for joining me today. Thank you for having me. And it was one of the most flattering introductions I’ve ever had in my life, so thank you so much.

Happy to do it. Um, so kind of to continue the introduction a little bit, um, you have a PhD in computer science for Master’s degrees and you’re currently pursuing your second doctorate doctor of forensic scientists of sciences. What impact has career, uh, education had on your career and your personal life?

So I think that continuous education while you work full-time in parallel, is challenging for your personal life, to be honest. Uh, but I think it is certainly worth it, in my opinion. Talk about professional experience. I would say that continuous, uh. Education is very important, and I’m not talking about diplomas or degrees because I know a lot of [00:02:00] great people, very intelligent people who have no degrees at all, and they’re absolutely smart.

Likewise, I know a lot of people who have numerous degrees. But they’re not that smart as some people who, who have no bachelor’s degree. Mm-hmm. So I’m rather pursuing my continuous education to ensure that I have a certain system in place because when I was young and I decided to study law I portrayed all kind of books like in a international criminal law or international.

Uh, law in relation to human rights and, um. I’ve got completely lost. And, um, I was thinking that, you know, law is very complex and complicated. Then when I started my first online degree in law, I was like illuminating, saying, oh, we actually need to start with like, foundations of legal [00:03:00] system understanding, you know, how it, how it works, their contracts down.

Civil procedure than many other things. So when I do my studies, it’s rather to have a system behind to make sure that I’m not starting at the wrong way and I properly follow, uh, you know, the, the, the, the course to make sure that everything that I should learn, I will learn it. And, um, I also enjoy the uh.

Knowledge that obtained, because it’s not about, you know, diplomas or titles, it’s about knowledge that you can apply. For example, now I’m studying a lot of psychology, and to be honest with you, like two years ago. I would never believe that I would ever study psychology. Think, well I’m a cybersecurity professional.

I’m also a lawyer, so why do I need psychology? But when I had in a class on psychology was my second doctorate degree, [00:04:00] I was like, wow, this seems to be super interesting because. It, it is actually very relevant for legal profession. Mm-hmm. And for cybersecurity as well. When you know how to better deal with people, how to be more respectful and, and more convincing at the same time when you know how to.

Structure your negotiations or pleadings in court. This really helps. So I continually discover new, new bodies of knowledge that are super exciting. I’m 38 and I think I’ll be studying till, you know, like 50, maybe a little bit longer, but I believe that, uh. In view of the, you know looming announcement that AI will replace everybody very soon.

I think, you know, having diversified knowledge never hurts. And I’d like to emphasize it’s about having knowledge and skills, not about having degrees or diplomas. [00:05:00] I, um, my bachelor’s degree was actually in psychology, so Oh wow. Bachelor’s of Science and that. And it was very interesting and it’s definitely something that.

I catch myself using a lot, um, even just subconsciously. Um, you know, I don’t necessarily necessarily think, okay, this person has this personality type or something like that, but I do kind of pick up on those things along the way as well, which is, uh, it’s interesting for sure. It’s a very interesting field.

Awesome. So apart from your professional career, you’re very involved in community projects like the Europol uh, data Protection Experts Network, the GIAC Advisory Board and the San CISO Network. What advice would you give to other attorneys who want to be involved in community projects but they’re just not sure where to start?

So I think we have a lot of great projects. It depends really on what attorneys are willing to do. For example, if you’re interested in privacy law, [00:06:00] I think the International Association of Privacy Professional is a perfect place to start. They have online webinars, they have conferences, they have different, uh, online certifications.

And, uh, I found them extremely useful. Okay. Mm-hmm. Maybe some experienced attorneys will not learn you know, something substantially new for them. But, uh, even they will probably get something interesting as potentially when you prepare for your. Exams, you’ll certainly have to learn a lot of new things that will help.

If you’re willing to do some cybersecurity. I probably think about American Bar Association groups dedicated to cybersecurity and data protection. If you’re looking for advanced, uh, cybersecurity experience, I’d probably consider Sense Institute. They offer a lot of online and, uh, in-person courses dedicated to cloud security, digital forensics, penetration testing.[00:07:00]

It may be a little bit too technical, but they, they also have interesting courses for incident response that are specifically designed for. Executives, maybe the one for legal professionals where they explain how to build a bridge between technical team and uh, uh, lawyers trying to explain a new complex technical things in a simple, all the whole language.

So this may be a place to start and, uh, I think that continuous, courses, um, that you can, uh, access for free on YouTube, even, for example. Okay. They can bring value. And I’m not talking about, you know, this TikTok style content, like how to become a privacy attorney in five minutes. Uh, no. But when you have BDS, you know, from professional conferences.

Universities, a lot of, uh, great law schools now make some of their lectures publicly accessible. I think [00:08:00] it’ll certainly help. Yeah, you could always, you know, you start small and slow, like at the A BA, I started as a newsletter editor and I did that for a while, and then you kind of go into planning webinars or you go into planning CLEs and then you get more into leadership.

You know, just start with one thing you can do and go from there. Absolutely. Absolutely. How does your company mwe, um, leverage ai? So we have machine learning technology that we use for many different, different things. In a nutshell, we accelerate and intelligently automate certain tasks. For example, vulnerability detection, uh, noise canceling on dark web.

When we do dark web monitoring for. Our customers. Um, so we try to, free app, valuable time of our people to make sure that they will only do those tasks that [00:09:00] truly deserve their valuable time. My favorite disclaimer, my favorite disclaimer is that machine learning and ai, even in August, 2025.

He’s no magic, he’s no silver bullet and is no a pan cell. No, we, it cannot replace a tos, no, it cannot replace qualified security analyst. It can be a lot of awesome things, but I believe that we are still extremely far from artificial intelligence that would be able to replace human beings. And, uh, I also have some doubts that it’ll ever happen because even if.

One day we have a model that is smarter than an average human being. That’s nice. But the problem is that if we have a smarter model, human beings will not be able to leverage it appropriately. Because let’s say you, you, you have an AI lawyer that is as smart as a human lawyer, that’s nice. But if you don’t know how, how law operates, how it works, [00:10:00] you’ll probably ask, you know, for, crazy things, you know? And the output will be the disastrous because, you know, hu human being, they have ethics, they have certain, you know, social experience, but AI will simply do what you ask for. So I don’t think that AI will ever fully replace human beings. Having said this, it can automate many things.

It can save a lot of time, but I think that many AI vendors currently exaggerate their claims about. Capacities of ai. I, I agree. I think there’s so many companies right now that use AI or that sell ai, and then you try to use it and it’s like. You know, the first question that you have is, where’s this coming from?

And they have no idea, or they don’t have a good answer for you. Like, where’s this data coming from? Or, you know, we’ve had, we’ve all run tests on ai, like our developers have run tests on AI for writing code. And it’s always like, well, you have to fix things because if you. Use this code, [00:11:00] it introduces a lot of bugs into the system.

Or I tested it for writing privacy policies and it like doesn’t know the laws or doesn’t know what the laws require or make stuff up. And you always have to have that human on the end fixing things. Um, and that’s why I think you’re right. Um, and this kind of segues us really well into a conversation about the use of AI by lawyers and law firms.

So I think there’s two types of use. So the first use is where it’s very clear to the lawyer that they’re using in ai. Can you give us some examples of the types of tools that, where it’s very clear that you’re using in AI and how those tools may be used by lawyers? Sure. So I think today we have many different tools for lawyers specifically.

I think where AI can really help is to summarize a long document because sometimes your client may send you a contract of 100 plus pages. Hmm. And you need just a quick summary [00:12:00] before you start doing your research. For example, you’ll probably still have to go through the contract manually, okay?

If you are willing to bring value and to make sure that, uh, you’ll not overlook something. Okay? We have pretty good tools now that can do a lot of impressive things like find, you know, clauses that are poorly written or that may be dangerous, but you know. All this is very subjective, to be honest. So you can get summaries, you can get, you know, uh, things like, you know, top 10 most problematic clauses or clauses that may deserve your attention according to ai.

But it doesn’t mean that they, that AI will do the entire job for you. So I think that. Summarizing is good and, uh, I’m happy to say that the current state of AI is, uh, good enough to produce high quality summaries. That’s like one of the things that [00:13:00] AI can do. Okay. Then talking about legal research I think AI can accelerate certain things, but um, you still have to do some research yourself because first yes, we do have AI that may access, uh, intranet in real time to enhance its answers and to make sure that it will always provide you with like current and up to date content.

But in my experience, when you have this and, uh. Uh, you start doing your research entirely with ai, it’ll certainly provide you with some irrelevant, uh, uh, cases coming from different states, for example, or where facts are different. And you’ll say, oh, that’s very interesting case because it’s. Fits perfectly in, but you fail to read it yourself and then you epically fail in court.

So it’s, it’s, it’s a, it’s dangerous tool. So it can be interesting to accelerate your research and, um, in my [00:14:00] experience, it’s particularly valuable because when you ask broad question like, what is the current legislation in California in relation to, Hmm. Data protection. For example, it may give you a couple of, you know, state laws that nobody even, you know, remembered.

And they say, wow, that’s interesting. So it’s not just about, uh, CCPA, uh, we also have this and that. Okay. I will also check this. So for, for me, AI for legal research, it’s like advanced Google that is quicker, that sometimes is smarter. But do not expect as lawyers that AI will do legal research for you at least.

Uh, you know, in 2025, I think it’s a little bit too early to entrust this important task to, um. And technology. Yeah. Uh, Hannah probably say that AI can also do a lot of interesting things [00:15:00] like reviewing your contracts, saying you have a mistake here. And, uh, to be honest, you know, when I’ve been drafting legal documents, AI found interest.

Things that I was like, aha, that’s, that’s a valid viewpoint. Okay. It’ll certainly not make a revolution. Okay. So it’ll not bring you a lot of, uh, substantial value. But sometimes I say, oh, that’s, that’s that, that’s a good point. Because in this jurisdiction, best reasonable efforts and best commercial efforts are two very different things.

So thank you for this hand. Okay. So I would say AI is. Is very good technology to accelerate certain processes, but it cannot replace lawyers for the time being. It’s interesting that you mentioned summarization. Um, so this is not legal work, but this is a way I’ve used AI in the past. So I wrote an article.

About something or other. [00:16:00] And the person who asked me to write the article asked me to translate that article into Lithuanian because I’m Lithuanian, so I’m fluent in English and Lithuanian. But on that particular day, I was just like, I don’t have time to translate this. Let me put this into an AI and have it translated for me.

And it’s really good at translation. But when I read the article that it translated, it actually skipped multiple paragraphs. So it basically just like pretended that those paragraphs didn’t exist. And I’m wondering, you know, what would happen if you put in a contract to have it summarized, but then it skips entire sections and just doesn’t summarize those, you know that that seems kind of scary.

I think that’s a great point and uh, I also think that it’s not a question of if, but rather when this will happen and, uh, this is exactly why. I’m saying that AI can help and when you just need, you know, to. Get an idea about the contract [00:17:00] once when your client sends you a very long and complicated contract.

Mm-hmm. You say, all right, let me, let me just, you know, get my first impression, you know what it is. Then you’ll still have to go through this contract manually. Otherwise, you know, you’ll, you’ll probably miss a lot of important things. But importantly, you will probably have, you know, better understanding of, of the structure because some tools they provide you with a specific I would say.

Location of paragraphs and sections in, in the contract. So you already know that liability clauses allocated on pages 25 to 26 and 71, so you will probably be able to navigate through the contract in a faster mm-hmm. Uh, but it’ll certainly not replace you as a lawyer. Yeah. As kind of a starting point.

Um, so what kinds of data, whether belonging to the law firm or to the [00:18:00] clients, could lawyers input into these tools, um, whether, you know, knowingly or, you know, maybe they weren’t paying too much attention. Um, what could they put in there? I think that’s a million dollar question, uh, because, uh, we have several different interconnected problems here.

So first, if you copy paste confidential information, uh, that belongs to your client in chat, GPT for example, you may be in breach of your rules of professional conduct. Okay? Because, i’m not talking about any vendors specifically, but many chatbots or SaaS applications that are powered by ai. They have this small clause, you know, in their terms of service, uh, that were manually written.

Of course, they don’t trust ai. Uh, that says everything that you, you input will be used to improve our technology and to [00:19:00] deliver better user experience. So when you read it, it’s kind of, you know, inspiring. Realize that everything will be utilized somewhere somehow by companies and people who you don’t know any, and they’ll probably never tell you who and how and what for will be utilizing your data.

So prior to copy pasting anything that may be confidential or. Protected by a privilege. Make sure that you have clear understanding how these data may or will be used. This is like a very important thing. Uh, second point is that I think. It really depends on the use case, but you may even use AI to summarize core decisions because when I was in law school I think you also remember this, uh, good old days when, when we had like 50 cases to be read by next [00:20:00] Friday.

Yeah. So, so it was an interesting challenge. Uh, but now if you don’t really need to or. Learn everything about a specific case because it’s like pivotal for your own case. It may be a good idea to summarize a case, and in my experience, you know, it can, it can, it can work pretty well because when you just need to get, you know, like the rationale facts, uh.

The outcome, AI can be pretty handy at this. Okay? But once again, if you need to clearly understand the specific ruling. Uh, don’t rely entirely on ai, and I think this is what, you know, many state bars have already said that you may use AI at your own risk, but make sure that you control the process because ultimately you as a lawyer will be accountable and liable for any hallucinations or [00:21:00] whatsoever.

We don’t know what, yeah. Quite a few lawyers have been, um, sanctioned for using ai. Um, whether for uploading client data and destroying confidentiality or for submitting core documents, um, where the cases were hallucinated. You know, and not just, you know, because of the, how embarrassing something like that would be.

But also, you know, you could potentially lose your license, which is devastating to a lawyer. So you said, you know, you input this data into the AI and it can become part of the training set. So for data becoming part of the training set, does that mean that it could come up as an answer to somebody else’s prompt in the future?

In my experience and I’m not data scientist, okay, but in, in my experience that if you, for example, upload the contract and, um, it’ll be utilized for fine tuning of a model. It, it is quite [00:22:00] unlikely that the entire contract will be provided as an output. Okay. But in case of under fitting, for example, when you have a very small volume of specific data about the specific laws specific contractual clause, and you uploaded exactly this chunk of data, it is possible that several.

Phrases from your contract will appear somewhere as an answer to someone’s else request. Mm-hmm. But I don’t think that, um, there is a big risk that the entire contract will, will appear somewhere. Having said this. Now I think the biggest risk is not even that someone else will see no parts of your contract.

The biggest risk is that AI companies, they actively share data. They actively share training data. So your contract may end up at. 500 different places around the [00:23:00] globe and you’ll have a lot of interesting questions like what happens next? Because a small startup in Germany or in Australia or in China may be compromised and your contact will end up in wrong hands.

So I think that’s all about, you know, who will eventually have access to training data. Because now given, um, the avalanche of litigation that we have a against. I would say almost all AI vendors that are large enough to pay something. I think, I think now we’ll see that vendors, uh, will rather.

Collaborate saying, listen, we have this part of training data that that was lawfully licensed. We have, you know, full rights to utilize this specific data for training purposes. Uh, they’ll probably be cross sharing, cross licensing data to each other and eventually your. Input data, your contracts, your documents, whatever you upload to the system may end up somewhere [00:24:00] in a very strange place.

Do we know who these AI companies are sharing this data with? Um, or whether they’re selling it? Or is it just kind of impossible to tell at this point who it may be shared? I think it really depends on the company, but I believe that most vendors will say that this is protected by trade secret and they cannot give you this information.

Mm-hmm. Yeah. That’s, that’s nice. Good for them that they’re able to say that because the rest of us can’t. Um. When data becomes part of the training set, does that mean that you can no longer request to have it deleted? Um, you know, you, you just lose the ability to, to have that deleted. I mean, you can, but whether, uh, the vendor will comply, that’s the big question.

And how efficient, uh, their compliance with your request will be. That’s probably even bigger question because once you have your [00:25:00] model trained, um, you either need to retrain or fine tune it, but you know, you cannot adjust. Delete it as with a simple database where we say, okay, here we have records, you know, uh, 36 and 37.

Let just remove them. With ai, it’s significantly more time and resource consuming, so. I think we’ll have a lot of interesting battles in courts where people will be asking to remove their personal data or to correct their personal data. And the AI vendors will be probably, you know, they’ll find a lot of creative ways to, to deny this saying either it’s in a un burdensome on their side or some other things.

So I believe that, uh, deleting something from a trained model is a very. Complex and time consuming tasks. So probably we’ll see all AI vendors trying to find a way to say no, no, no. If it’s, uh, already in the model, [00:26:00] uh, we’ll do something, but, uh, this will happen in the next five years. Yeah, that, that’s gonna be very interesting to watch.

Um, so the, the A BA recently published, um, formal opinion 5 1 12 on the use of generative AI tools by attorneys. What does that opinion say about the preservation of confidentiality and the use of AI tools by lawyers? Uh, so that’s a very interesting document and, um, I obviously cannot speak on behalf on of the a, BA, but I, I found particularly useful that this document.

Included many different aspects of using ai. Uh, as I said, confidentiality, it’s a good reminder for all lawyers saying that, listen, if you use ai, that’s fine. Enjoy your AI journey, but make sure that you protect confidential information of your client. You cannot later say, oh, [00:27:00] listen, that’s not me. It’s actually AI debt.

Did something I have, I have no clue what happened. Right. Yeah. I also found, you know, one, uh, specific thing there that is probably underestimated now because if, uh, we read the news, we’ll probably say a lot of things about hallucinations, but I believe that state bars and courts will be. Stricter when dealing with attorney’s fees because of course it’s very interesting idea to say, oh, I will do half of this job with ai, but I will still bill for human hours.

Mm-hmm. And the a BA was crystal clear saying, just don’t. Just forget about this idea because it’ll probably be disbarred. Uh, they probably said it in a more politically correct manner, but, uh, I think, I think that courts will probably be [00:28:00] more lenient, you know to I’d say innocent mistakes.

I’m not talking about, you know. Citing fake cases before Supreme Court as this is like in a very, yeah, very, very problematic. But, uh, if, if, if you have a typo or you know, a phrase that is just inappropriate for a specific pleading due to ai, I don’t think that you will, uh, be severely sanctioned for this.

But if you start, uh. Over billing your clients. This can cost you a license. So this is like a big thing that many law firms, they tend to forget. In my opinion, they see AI as a huge opportunity to automate everything. That’s not bad. I don’t think that they will manage to automate everything, but many processes deserve automation because we still having, you know, certain tasks and processes that are [00:29:00] super slow and inefficient and you can automate them both with AI technologies and classic software.

You don’t always need AI to. Automate something. But the thing is that as soon as you automate your legal processes, you’ll probably have to revise your billing. And you’ll also be required, as pointed out by the. A BA to disclose to your clients then A, we’re using AI first. Is it acceptable on your side?

And second, this is how we will build ai. So we are fully transparent with you. So I think a lot of low firms will have this aha. Ethics saying, oh, we hired all associates. Uh, uh, we, we, we, we, we, we hired all associates. We have no more associates, only partners, and we’ll make a lot of money. And later on they’ll uh, realize that, [00:30:00] uh, listen, actually we cannot, uh, build all desires if we use ai.

So actually we need to hire them back. So I think, I think we’ll see a lot of cases. And I actually know, so several examples already this year, when. IT companies, they fired, you know, their first, uh, layer support team saying, oh, now we automate everything with ai. Very smart. But then once they started getting, you know, customer complaints about in inappropriate abuse or harassing behavior of AI chatbots and they started receiving, you know, summons to courts.

So saying your clients, assuming you, they, they were kind of, you know, saying, uh, listen. Yeah. In indeed we’re not paying salaries, but now we need to spend a lot of money on litigation and reputational damage is like huge. So let’s just stop, you know, this [00:31:00] ai, hype and rehire everybody back because yes, we’ll be paying salaries and maybe we’re not willing to pay salaries, but at least we’ll, uh, stay out of the court.

So I think this hype will soon start going down saying, listen, we prefer to have, you know. People because yes, we need to pay them salaries, but they’re more reliable and predictable. Yeah. I could also imagine, um, some clients that go to law firm that are told, okay, your work will be performed by the ai, would not be happy with that situation.

I think they, they would say, well, why am I hiring a law firm? Then I can input my facts into an ai. Get a similar output. Why am I paying you? Um, you know, so a lot of times when you go to a vendor, you’re hiring that person because of their profession, because of their knowledge, their expertise. You know, if I could just do the same thing with an ai, I wouldn’t hire a law firm at all then, you know?

Um, so I think there’s gonna be law firms that get pushback from the [00:32:00] use of AI by their clients. Excellent point. Excellent point. Yeah. Um, so AI could be used by the lawyers themselves. Um, it could also be used by the staff. So do lawyers have supervisory responsibilities for their staff using ai?

I think the answer is clear. Yes. So whatever happens, I mean, if you’re associate or a paralegal does something for a client, you will be eventually and ultimately liable for his or her performance or non-performance. Same with ai. It’s like no excuse, and I think many state bars and the. A BA, they crystallized this point saying, listen, uh, don’t come to us saying it’s not me, it’s ai.

Yeah. Or My dog ate my homework. This will not work. Yeah. Uh, yeah, you see that a lot. But for some reason my dogs have never [00:33:00] eaten my homework. They’ve eaten a lot of other stuff, but not the homework.

So one of your recommendations for lawyers and law firms is to create an inventory of technology technology vendors, um, who have access to data. How does having that list help? So I think this is a very important point for all law firms, and I think they should also consider reviewing, uh, uh, their common vendors.

For example, if they, uh, utilize Microsoft Office 365, it may be interesting to see what options you have, how your privacy is configured, because now virtually every single vendor started doing something with ai. Including Adobe, including Microsoft. Uh uh. So double check what’s going on. Okay? And uh, when you have this list of vendors and you have understanding [00:34:00] what they do with your data and.

To be honest. To be honest, I don’t think that we’ll ever manage to ca a crystal clear understanding of what’s going on because in many terms of service, you will find in a very ambiguous and, uh, over broad and vague definition saying we may occasionally, uh, utilize some of your data that is not confidential.

I have no clue. What does it mean? Okay. Yeah. Uh, uh, for, for, for some training purposes of certain AI models that will never be used for public this kind of thing. So you scratch your head and say, okay, let me check with their support. And each time you talk to their support, you have the impression that you are talking to an AI chat bot that will, you know, persistently say, uh, we do respect and value your privacy.

Uh, we absolutely appreciate you being our customer and we always do our very best to protect all your confidential information. [00:35:00] Might you need any additional information about our privacy, please click on this link or that will. Take you to our privacy policy page that you’ve already read. Yeah, exactly.

Exactly. Exactly, exactly. Very true. Yeah. Whenever you reach out to a company with a privacy question, you have to wait multiple weeks for a response, and then you get the non-response, which is. You know, if you have any questions, read our privacy policy. Well, I did and I told you I did and it wasn’t clear.

So I need clarification and then you never hear back. Should law firms have a policy on the use of AI and conduct training on the use of ai? Absolutely. Absolutely. And in my experience I would say this is similar to cybersecurity and data protection policies. So it’s not about creating a policy and, uh, be proud that you have a 50 page, uh, long policy, uh, because nobody will read it.

Okay. But it’s rather about, you know, enforcing the [00:36:00] policy and training as you. Said your employees that this policy exists and ensuring that your employees do understand what they are required to, to, to do, and monitoring that they actually comply with the policy. So what I sell as. Good examples of enforcing your AI or any other policies is to make sure that you have like a brief summary, you know, top 10 or top five, you know, key takeaways printed on a single uh, page and, uh, for example, placed in your kitchen if you have one in your company.

Hmm. So people, while you know, uh, having their. Morning coffees, they’ll say, oh yes, by the way, each time I click accept all terms and conditions, I need to first click on the terms and conditions, quickly go through in case of doubts, reach out to our legal, and only then, uh, [00:37:00] uh, I may actually accept them.

So I think, you know, this continuous reminder that, hey. We have an AI policy. AI is good. Uh, we endorse ai, but please make sure that you do 1, 2, 3, 4, 5. This will greatly help because if you just have a policy but nobody cares or it’s like too long and too complicated, people will probably say, oh, now we finally got a policy.

That’s, that’s great. Now we are all safe. Yeah, I like the reminders because people forget, you know, even if you do training couple months later, all, you know, everything happens so fast during our days. You’re getting emails, you’re getting calls, we’re getting pings, you know, we have to do this, we have to do that.

It’s so easy to just click accept or just, oh, whatever, we’ll just use this super fast, just this one time. But having those reminders, uh, kind of in your face really helps make sure that people take the time and say, oh, wait. This is one of the five issues and I, I should not [00:38:00] do this super fast. I should think about this.

So last question for you. If people wanna reach out to you or wanna learn more about your company, what’s the best way to do that? I think the best way to connect is using. LinkedIn and, uh, also know for non-European users of this amazing, uh, social platform. Make sure that you opt out if you would like to, of course, uh, from, uh, use of your, uh, LinkedIn posts for ai.

Training purposes because, uh, LinkedIn recently said that we will be using your content for AI training purposes, obviously to to to improve what we do and to, uh, make your user experience even better. That’s appreciated. But if you are unwilling to do so, you need to opt out. Stop opting people in by default, just stop doing that.

It’s still frustrating because then everybody has to spend time going through 50 different pages trying to find the right [00:39:00] optout page and it’s always like super hidden and, and super difficult. But yeah, we should definitely. Figure out how to do that and turn that off. Um, but Ilia, thanks so much for, um, speaking with me today about how to protect your law firms and your client’s data in the age of ai.

Um, and to our listeners, make sure to subscribe to Privacy Lawls so that you don’t miss our next episode. Thank you so much. It was a big pleasure.