How AI is Transforming Law: Interview with Legal Tech Expert Harry Borovick

[Music] what's up guys and welcome back to my channnel today in this video we're going to be interviewing our wonderful guest Mr Harry barvick who is a lawyer who is an expert in legal Ai and who wrote a book called Ai and the law which also talks about that he also teaches at the famous King's College and Queen Mary University which are two universities which are also in London so let's get right into this amazing video where we're going to be talking about well legally ad let's go AI works on one of two basis you either pay and there's different costs to different uh systems and that work differently but you overpay and you typically get more security and more control of what your data or you don't pay you get lots of free service but in exchange the AO company keeps some of your data we'll see a future where very basic legal premises um what what's the basic law of X subject um I want to sell my house what do I need to broadly be careful about um that kind of thing um I think that will become increasingly available to ordinary people at a cheaper price than is currently the case if I have a subscription for a technology product like luminance um I know what that's going to cost and I know how much usage I can get out of it within my subscription whereas law fir fees have historically been quite unpredictable so here guys we have the man himself Mr Harry borovic and today we're going to be talking about the law of AI because as we know AI is an emerging technology that just came into place and that everyone is using nowadays and well here we have Mr Harry borovic who will teach us about the laws of AI and well the legal things that you can do in Ai and what are the restrictions because as we know it's a very powerful tool so hi hello welcome hi thank you for having me so first of all let's go to three main questions to know you better so what inspired you to do what you're doing nowadays what inspired you to be a lawyer in AI um well I wasn't always in the space I was previously um a lawyer in emerging technology of different kinds in gaming in advertising Tech uh in fintech various different types of uh emerging and rapidly regulated technology uh basically technology where the impacts of that technology are potentially significant enough that governments or Regulators who maybe report to governments or are funded by governments have to take a Clos look at the technology to protect consumers and businesses and I slowly expanded my work in that over the past decade and now work almost exclusively in the AI space uh at a company called luminance which is a legal technology company with its own proprietary uh I type of AI that they own and that they make um exclusively for lawyers and people who interact with lawyers all right but what inspired you to get to being a lawyer in the first place that's a good question um I had a when there's a bit of a myth that you need to know that you want to be a lawyer or you know you want to be in any profession that wasn't the case for me I did a politics degree University didn't know what I wanted to do with my career and I had a friend who um said hey I don't know what I want to do for work yet but so I want to stay in University for a bit longer maybe I'll do a LW conversion because that will allow me to um take a bit more time but also give me an additional academic experience that might be useful in many careers and I thought that sounded quite sensible um whether I became a lawyer or not and then one thing led to another and I actually quite liked being a so you just found out that it was a great opportunity so I can get into many domains and you had many chances and then from that you eventually evolved to being a lawyer for AI yeah exactly that's exactly how happens yeah it wasn't um a master plan um and so how were you able to start all of this so you were able to start it from high school right you start so you started learning all about this from high school then after that you directly started getting into a job into all that right um so I didn't really get into technology and uh law you became a lawyer and then you evolved to death yeah exactly so I wasn't particularly technical before then and uh the legal skill set came first technical skill set in along the way all right and so what made you involved in in the legal text base what got you involved in that well when um AI started really taking off and actually just before then so most people would be familiar with things like cat GPD and demini or anthropic and those are sort of consumer facing AI companies that also have business sides to them and I got involved in AI as we currently understand it or currently discuss it around that time just before but um for me the reason I got into the legal technology space which uh is much older than the AI industry per se but the AI industry sort of has integrated itself um into the legal technology space um I saw that legal technology was one um right for updating because lots of the technology that was involved in it was quite outdated had been around for a long time and normally that's a good way of understanding what technologies what sectors of Technology are likely to move when there is a technological innovation because they have been static they haven't moved or changed a long time so as soon as something that is genuinely different or genuinely moves the needle as the expression goes um they're likely to change a lot um so frankly I was in the right place at the right time but um it was um the right company approached me which was luminance to be a lawyer for them and then it made sense to um join them in any event before AI went crazy and now um having that of foresight was quite lucky because obviously the industry has advanced enormously since then so it's been really great but the initial thing was I just saw an outdated industry that needed um a lot of updating and I thought the work that luminance was doing was really powerful to do that all right and so I'll turn that you made this book called Ai and law well based on this book uh you were talking about the future of of legal Tech and you also like from this book what do you think that every lawyer and every Law Firm should take away from it well the idea of the book wasn't really for um lawyers so much it was for non- lawyers so if you are you and you're going to school or university what can you learn from um Ai and what can you use AI to do um in a way that allows you to still stay within the right side of what is legally and ethically correct or um at least give you the right questions to ask about what you should be thinking about when deciding whether you should use AI whether that's as a lawyer a doctor um someone in design um or just uh a high school student writing an academic paper what are you doing and what the questions you need to ask yourself or the tool the companies you make tools that you want to use in order to understand whether you're doing something that's potentially right or potentially wrong all right so we're just doing it like not necessarily for lawyers or people who have their job in that but just so that people can can get themselves better informed inside of these topics which are like commonly used nowadays like AI it's everywhere so you just wanted to be able to overall uh give a better understanding of AI to Anyone who reads it exactly so you people are using AI Tools in all kinds of different ways whether that's to uh work within the academic context or to manage their money or to do legal work or to create new designs whatever it might be um the book's called Ai and the law but the more important bit is the subtitle which is a practical guide to using AI safely so it doesn't really matter what you are doing with it um it allow the idea of the book with simple strategies and simple questions to ask in order to understand how can you manage risk stay safe and use AI Tools in a responsible way all right and so how do you think that AI is shaping uh how do you think it's shaping the well the job as a lawyer nowadays because like you of course you you are doing a legal AI you're doing uh AI yeah legal stuff in AI how like how is it actually changing your workspace how's that changing how you're actually working yeah um so one of the most fundamental ways is it changes the dynamic between the legal customer and the law firm so even companies that have lawyers within their business they often use external lawyers I lawyers that are within law firms for additional support and and one of the big facts is a lot of the type of work especially at the more junior level that a lawyer like me would use an external Law Firm to do is now being capably done by AI tools that I can use myself and I can maintain greater control not only of what the outputs of that technology are but also the cost because it's more predictable because if I have a subscription for a technology product like luminance um I know what that's going to cost and I know how much usage I can get out of it within my subscription whereas law fund fees have historically been quite unpredictable and often charged on an hourly basis so the more you use them um the more expensive it becomes whereas legal technology allows you to use Tech to have a more predictable cost well what is the well what is the benefit of using AI to replace the role of external law firms because like what's what what makes AI better than external law firms in this case so a very good question question lots of lawyers ask me um one of the ratest advantages consistency so to start with we're not trying to use AI to replace lawyers the idea is we're using it to enhance lawyers now in the long run that may be that there are fewer um particularly junior level roles for lawyers or for tasks that are done by lawyers but we're not trying to replace the lawyer itself we're trying to replace some of the tasks that lawyers do which are repetitive and require consistency so one of the um more classic legal tasks is um large scale review of documents you know get millions of documents to do a deal and we need to go through them all and check we're happy with them doing that quickly and consistently having the same perspective of what's acceptable and not acceptable across all those documents is a really big Challenge and one that historically was very expensive to do now I can do that predictably in terms of cost but also in a way that is consistent in a way that uh a law firm just using lawyers without technology is not capable of doing so using AI as in like a a tool to help you task yeah um actually so if it's helping you do things what can AI do that like you can can really do because if the AI helps you do some of these things could the AI really just replace the job overall because like what are the things that you don't want AI to do uh that that you trust more lawyer to do than AI well it depends what you're trying to ask it to do so the one of the reasons I focus on tasks rather than the job is because one of the main reasons people use lawyers is because they want to be able to rely on the advice of the lawyer they want to and lawyers are insured normally and they most countries it's a legal obligation like doctors that if they do something wrong there is a safety net of insurance now the reason for that is because it is reasonable that if you or your family were using a lawyer services if they got it very very wrong that you should be able to challenge them legally um or take them to their regulator you know most law like in America the bar associations or in the UK it's called the SRM uh and challenge them for the quality of the work they've done so there's still a place for the lawyer there what is more likely as the technology advances and we move into what's known as a more agentic using a AI agents era is that agents are more likely to do more tasks autonomously I without me asking them at every stage to do every single task as a tool the tool is likely to be able to do more tasks predicting what I want more accurately and roll out so you are right that there is there is likely to be a future where more tasks are simply done by AI but there is no task that an AI is currently capable of doing which a lawyer isn't it's more about how they do it fast so at scale and efficiently so a lawyer basically uses AI as in like a tool like it's like so it's like calculator you are the one who is who is actually knowing what you're doing and you're the one that uh well a lawyer is supposed to be giving advice in the uh yeah in the legal domain so like lawyers they're always going to be here to be giving advice and computer's not really going to be able to give advice so use the AI as in like a calculator basically to help you do some things which would be repetitive and simple just to do it faster yeah somewhat the the calculator example is an interesting one because um a calculator can is probably more accurate with the actual mass once the right numbers are inut um than most people would be the point is is that the calculator doesn't know whether the source data you've asked are the numbers you put into the calculator are correct or whether there's any other questions you need to ask in order to make the end product of that equation correct that's what I mean the the human is very good at at um contextual understanding well like as you say AI just Min Chas and you it for for enhancing uh the job basically it's being a lawyer well how is this going to to shape the career of a new and uh and and like a lawyer who just got into the job and who just started and like a young lawyer who just started how do you think that's going to shape like their their go their growth path because if the AI is helping so much and if the AI helps to do so many simple tasks how is there room for improvement for new lawyers and for junior lawyers yeah that's a great it's a great question so the basic point is that at the moment if you're a junior lawyer or new to the career you're likely to be doing tasks which aren't necessary for the customer you know reviewing lots of documents for example but don't necessarily make you a better lawyer for the things that later in your career your clients will really value like the quality of your ability to analyze and understand complex problems and give them useful advice the advantage of using technology to replace not just AI but any kind of technology that can replace some of the more Junior repetitive tasks is that it allows Junior lawyers and Junior practitioners of many kinds to access and be involved in more complicated work earlier in their career um they'll just have more free time read up to do more complicated work so now the other wrong sorry go ahead no uh continue sorry because maybe it will answer my question the other points is that they will also have um the ability to understand earlier what matters to a client because they will actually often be client facing much earlier in their career because it will be necessary for them to have those client facing skills earlier because that would be an expectation now that AI comes in uh especially Junior lawyers use the uses this tool not only as a mentor kind of thing like they look at how it solves it and they look it like from reference basically uh but also it's going to help the junior lawyer U grow faster and hopefully faster going to be able to go into tasks that clients would usually want the lawyer to complete and more complicated things earlier because AI is going to do much like it's going to do much more of the work which would which a young lawyer would usually do at the beginning yeah exactly all right and so well AI can process a lot like a lot of data very quickly but how like how can you make sure that lawyers don't R too much on it and review the data carefully that is the million dooll question and I think that the big point in this is that law firms and universities who teach lawyers um will have to fundamentally rethink how they teach lawyers to focus on critical analysis critical evaluation of output of AI so you know a very simple example is instead of teaching lawyers how to draft something from scratch makes much more sense to get an AI system to do the rough first draft of a contract and then teach the lawyer how to critically Analyze That output and modify it to the exact purposes in a way that the client is expecting adding the context on top and I think that is a more efficient use of everyone's time but it's definitely quicker as an output in most cases or at least in many cases and it's to mean that universities do have to adapt their curriculum and law firms have to adapt their training systems that's inevitable many law universities already are um I myself lecture at various universities um on exactly that kind of subject and I think we're going to see a lot more of that internationally as time goes on all right and so like uh what you're saying is that uh lawyers shouldn't like rely too much on it but like they should use it like like kind of at least as they can so like they should maybe review what they're doing but not completely only use it so it's like you use chat GPT to like do an equation do whatever but then you want to make sure that it's actually valid and you're not sure if you got it right so it's like you want to review it and you want to know if it's accurate but you don't want to rely too much on it at the same time exactly that's a a perfect something all right and well as we know a AI is really great for speeding up things but how do you protect your data that you're giving uh to the AI uh and how like how do you protect the sensitive data from your clients and from the people you're while working for well I did talk about that quite a lot in my book so if you want to know some of the secrets by all means it's very practical in there uh now available on Amazon but the most important point is no matter what system you're using matter how Specialized or secure it is whatever the system is the best means of protecting your data is being selective as to the data that you give the system so if you have millions of documents before giving the AI system all the millions documents to review might be worth considering actually do we need to give it all those documents or do we actually know um that only a certain percentage are really relevant to the system and therefore let's limit what we can do and only use that other thing is there are different qualities between B like companies that sell AI systems um and mostly AI works on one of two basis you either pay and there's different costs to different uh systems and that work differently but you overpay and you typically get more security and more control of your data or you don't pay you get lots of free service but in exchange the AO company keeps some of your data maybe all of your data depends on the company and can use it to improve their systems in that case you're probably not really protecting your data but Mo for most people you know non- lawyers non professionals who are using an AI system they're not really that worried about protecting their data because the data is normally not that sensitive it's not that important they're just trying to reword an email better or rewrite an essay in a way that makes more sense so they're trying to use um an AI system to help them as a creative partner in creating something um in that case you can be more relaxed with the data you put in but it's worth really thinking through those considerations before you start using so the more mature you get into uh into being lawyer into law um the more you need to rely on how uh protected and how a sensitive your data is because like the the the more experience you get you want your data to be more protected and more uh well better yeah better protected because like you said usually new people like young uh Juniors lawyers uh they usually want usually like it's not a really big deal if the data is not that protective but then uh eventually it's getting more sensitive more sensitive where it has to be more prot protected it is that what you're saying somewhat I mean for someone who's like a lawyer working with clients it doesn't really matter if it's um a big deal or a small deal the deal the deal data is always sensitive because they owe specific legal duties of confidentiality and various other things to their clients in the same way as a doctor but maybe if you are an ordinary business that is not regulated you know maybe it's a property company something of that nature um really you what you're trying to think about in terms of sensitivity is how personal that information is is it people's names and addresses is a is a good example people probably wouldn't want that being put into an AI system that's controlled by um a company they don't know um so you really have to assess the type of Industry you're in that's one of the reasons why my book is divided by industry purpose so I what you mean so it's like it's not necessarily about uh how experienced you are but it's more about the task and how how valuable and how like how fragile the information can be and how like how much impact the data could be uh on like how much impact it could give to a person if it's well misused basically exactly taking um what lawyers often call a risk-based approach if if the data is used in a negative way or in a way which is not intended by the people whose data it is how severe or uh potentially negative even if nothing actually negative happens but how potentially negative is the outcome of that data being used in a way that wasn't originally intended all right and so what do you think the role of the government should play in controlling the use of AI legally in law that's a big question uh I would say uh there are some nuances so I'm not going to get too technical because we could go very deep into a discussion about copyright and uh various things but I think the purposes of your audience I'd say the most important thing is understanding that consumers Ordinary People in members of the public expect governments to place reasonable Protections in uh around technology to support them they don't need to be an expert in every piece of emerging Tech technology or every iteration every new version of a technology in order to be able to use it safely they should have reasonable safeguards placed around them particularly against things like being defrauded or um um being misled on social media for example there's all kinds of ways that harm can come to individuals through the use of AI either by themselves or by other people towards them um whether that's deep fakes uh and imitating someone's voice on the phone whatever it might be these things are critically important for the protection of ordinary people in society um they're very when you get into the details of them actually making laws that protect people and having Regulators who have powers to enforce these things is actually very complicated but um having a visible Desire by government to implement practical protection that actually make sense for most consumers is really important first step and we are seeing that increasingly mhm all right and so what kind of legal tasks do you think should always be handled by humans and that AI should not really touch because as we were saying uh like a lot of like we use AI as a tool but like what kind of things do you want AI not to touch at all and usually the lawyers should do by themselves yeah I I basically think none I actually do you think the AI can touch any task I think the question is is it the right AI for that task and there's all kinds of different AI tools or tools that just use AI or have some kind of integration with AI but um I don't think we should be afraid of AI so there isn't any task that AI can't do it's more a question of should it right and if you are for example speaking to your doctor now maybe AI is a really good um initial diagnostic tool hey I've got a headache what's the right medication for me and if it has access to a medical history you know if it's a tool by the NHS or something of that nature then that would be really great because you can get access to Quick lowrisk Health Care in an efficient way but if you've got a really complicated diagnosis of something with a lot of risk you probably want to have that discussion with an actual Doctor Who can ask more contextual questions they can give you a physical um analysis and uh that doesn't mean they can't be assisted in that diagnostic process by system so it can touch it in that way but not being the sole um face of the process uh and also sensitive data can be do you see that sensitive data could also be touched by a ey it can be it depends how well managed and secure the system is and how well trained the users of the system are to understand what makes sense into the system if they're just putting every piece of data into the system no matter how sensitive it is that's probably likely to end that all right well like the biggest challenges with AI uh is like usually like uh ethical issues and like uh usually not very equal uh thoughts of AI so how do you think that that would be in the uh legal world like how do you think the legal word world in AI could prepare for this because it's not always that um that accurate you know yeah and I think that is always the thing is any tool that is consumer facing facing um ordinary members of the public that's providing legal advice of some kind needs to be very very careful and needs to be clearly marked so one of the um one of the biggest discussions in AI is about something called labeling and labeling is do I know I'm receiving advice from an AI system if you know it's from an AI system then lawyers are regulated right so it is illegal in many jurisdiction in many countries to give legal advice or medical advice for example without a regulatory license I there is a government body somewhere that says you are competent to provide this service whether that's as a lawyer or a doctor and no AI system should be doing that without very big warning saying this is not legal advice you should seek the advice of a professional as well but that's not to say that AI systems won't be able to provide some fundamental guidance in a very democratically available way because they'll be a much cheaper to provide basic guidance at scale than dealing directly with lawyers would it really does I do think we'll see a future where very basic legal premises um what what's the basic law of X subject um I want to sell my house what do I need to broadly be careful about um that kind of thing um I think that will become increasingly available to ordinary people at a cheaper price and is currently the case all right and and so our last question of the day um what is your me message to the people watching right now I think people shouldn't be afraid of AI people should use use AI Technologies and explore AI Technologies one of the best ways to learn is to play and playing in a lowrisk way by you being careful about what data you put into a system but exploring the capabilities of an AI system or multiple AI systems especially if they're lowc cost or free is a really good way to get familiar with the technology and when that technometer becomes so common that everyone's using it for everything or that's reasonable you be in a good position to understand how to use it safely and of course there are great academic resources and non-academic resources you know my book aims to do something like that which can help provide education and um guidance as to how to play in a safe way but also to gently integrate that into a more professional or academic setting right well thank you so much for this amazing interview and uh yeah well thank you thank you welome thank you so much for having [Music]

2024-11-05 23:36

Other news