Online Signature Lawfulness for Healthcare in European Union

How to eSign a document: online signature lawfulness for Healthcare in European Union

in these exciting times of AI and Healthcare we probably need to stop and ask is is this legal technology is reshaping how we do a lot of Professional Services around the world it's impacting the legal system education and of course Health Care once the buzz wears off from all the hype how do we go about understanding the legislative requirements and the legal implications of emerging Technologies in healthcare for the vendors creating the technology and the providers using it well with me today is Mick cloth zorkoski a lawyer lecturer and advocate for legal Tech and health Tech Innovations and is based in Hungary in this episode we talk about the complexities of AI in healthcare and legal Tech how do you prepare for upcoming AI legislation in the EU and the challenges and opportunities for healthcare providers to uncover unbiased data in this new legislative landscape collaboration starts with the conversation team Health tech let's make it happen this is talking Health Tech with me Peter Burch featuring content and Community about technology in healthcare [Music] Nicholas how you going yeah that's fine good to have you here very good better now I'm talking to you buddy so it's great to be able to find a mutually okay time on the world clock for us to speak because I'm here in Sydney and you're all the way over in Hungary is that right yeah in the middle of Europe yeah yeah right and so before we dive into some of these discussion points because we just spent half the time talking about it all then because it's great to connect with you and learn about what's happening in health and Regulation and legal and everything else and commercialization in in Europe but for those that don't know you Mick loss or maybe they're not in the THD plus community and haven't seen your post there introduce yourself tell us what you do just before I start I like THC Community very much it's a so open community so uh you must join if you are not remember you're always welcome to uh to come back on the podcast if you start with a plug to join the community so uh that's a great way to start so back to me I'm a lawyer I'm working in Europe and I support league and health Tech Innovations and developments I learned some python coding just to have a better understanding how AI design should be ethic and trustworthy I do lecturing in a legal Tech Skill Center in Hungarian based University I meant to Legal Tech and health Tech projects in you know startup Academy also in Hungary so it's basically about me yeah and so you know that's that's interesting it's interesting speaking to people that you can tell the type of person someone is when they're like that that inquisitive Hands-On type person when they're trying to learn about a particular era so they become professionally proficient at it you know like you wanted to learn about Ai and and the technology side So You Learn Python coding so but tell me a bit more about that specifically what got you so interested about really kind of diving into that particular area because there's a lot of stuff you could cover on the legal side yeah absolutely uh you know AI is international so wherever in the world you work uh you are now a subject to the EI and also you need to deal with the AI problems and if you are a lawyer in Hungary and and working in a specific jurisdiction you still have international problems with the with the AI tools and especially in medical devices and also in Liga text tools so it's you know it's now it's very common it's very popular you can do work anywhere in the world with these Ai and test problems and legal aspects I was going to say actually you brought up a good point in that I've I'm not connected closely to it but that whole point of AI and legal Tech raises really interesting um where is it challenges is it opportunities what's the whole uh conversational Ai and and the speed in which you know that kind of stuff is entering the legal space um how's that having an impact yeah so the first program is we don't have a definition by the law so how do you treat AI if you don't have a single definition for it but there is a draft regulation in the pipeline in the EU so you know EU first we do legislation and then get some experience how it works in in real life but you know um it could be applicable within 18 months after the adoption of it and it will have a definition finally and it says that an AI system is a software which is good that is developed by a technique of machine learning approaches again something about the attack including wide variety of deep learning logic or knowledge based approaches expert systems or statistical approaches and can generate outputs under human based objectives to make content predictions recommendations decisions influencing the environments they interact with so how to comply with the uh with with the AI act and how how to export for example an AI solution uh from uh from your country that meets all the requirements so if you want to export the solution to the EU first there are community laws of the EU to comply with like as you know the general data protection regulation which was established five years ago uh at least we have data protection act and it was the first in the world oh this is gdpr right it took the GDP areas so back to the AI second there are laws of the member states to comply with the local so uh in in within the EU there are 27 different locals though and local jurisdiction which is which is not easy to comply with but you know the on the level of EU law if you get a certificate like like a CE certificate then you can do a commercial activity in the single European market so you need first to comply with the EU law then to the local and and then you can you can roll out your activity around the uh European market I want to come back to the ca thing in a second and commercializing in Europe because that that's an interesting conversation itself coming back to the AI and and this this this AI law AI legislation what so it's within Europe um it sounds like it's a pretty all-encompassing thing that's that's addressing areas of risk and I imagine Healthcare is one of those areas of risk that would be in have implications by this legislation yeah this this aea act will treat as a high risk activity uh the health connectivity will be treated as a high risk activity but what does it mean if a medical solution is treated as a high risk activity you need to have a risk management system and it should be implemented and maintained and maintained within the system you need to do training you need to have a validation and testing data sets it shall have appropriate data Governor data governance practices you need to have technical documentation which is obvious everywhere around the world you need documentation but it shall be drawn up before the system is traced and you need to submit it to authorities it shall be designed that record events automatically login files everything must be recorded just in case of trouble you need to be accountable another the meaning of transparency the user must be informed on the details of the provider and the system a human machine interface tool must be developed to have an appropriate human oversight to minimize risk to health safety and fundamental rights and also you need to comply with the cyber security um requirements the providers shall have quality management system a Conformity assessment shall be made before the registration made by the authority and also you need to report all the incidents that you have to the authority so the good news is that there will be an AI regulatory sandboxes for those startuppers and and startups who want to want to have a controlled environment for their facility for the facilities of the development and the testing and validation of innovative AI system before the placement on the market so in this way EU and the legislation will have startup first uh to to empty to the market which is you know very very hard because you you need data for this AI is data and algorithms uh which uh and as a result of it uh you will have a software but in healthcare it's very very hard to get to data Docker data quality data yeah and and um distributed or non-bias you know as well as is a real issue the um that's an interesting point that you raise that you know it's a good opportunity for it creates an opportunity for startups uh and it's not often that you hear about uh additional regulation or anything that that creates more of an opportunity for startups as opposed to creating more kind of costs and barriers and delays and everything so perhaps by having definitions and a structure and resources available it's uh it um it makes it less of a wild ride Western and it creates more of a structure and gives some trust to the whole process as well I imagine absolutely and from my member states perspective so in Hungary we have a national he has infrastructure and this is a single one single system all the healthcare data even if it's a government a hospital of its private sector hospital it's recorded in one single uh Healthcare e-healthcare System so um it could be a huge advantage other member states or even larger countries which are divided into regions or smaller provider and bankroller units so in and as a National Health Insurance Fund and there's a national ehealth in such infrastructure the National Public Health Care are in one management data management and data governance so if you need data as a startuper you can just Join one single system and then it's it's interesting they are very happy if if they have solutions to improve you know improve the the healthcare activity and if I think a little bit strategically I'll look ahead and think about the global implications of this this regulation and legislation things coming through you know you touched on gdpr really quickly and that one it impacts it's for anyone in in Europe however as an organization from outside of Europe even in Australia where we're not bound necessarily by gdpr for data in Australia that that's just you know we have our own rules a lot of organizations even if they have plans eventually they're not even in Europe now but they have plans to move into Europe they'll they'll make sure they're compliant with gdpr and you hear about organizations in the US they they're like well we may as well be gdpr because you know it's a it's a globalized world and people jump onto our website from anywhere in Europe or they expect that legislation is going to go that way anyway in the US because you know it already has happened with gdpr so the reason I bring that up is that there could be a similar kind of trend too is that the the world might end up following and see the Lessons Learned out of this kind of structure that's being created at a very large institutional level around Ai and and potentially be something that's that's rolled out in other countries too absolutely so imagine that this gdpr has a principle that the data process must be transparent but uh without having such AI Acts there are the data console and there is a data process so it needs to be transparent but what does transparency mean so is it the kind of accountability or or what should we do with this transparency principle ing to the AI so um I would say the transparency means that an expert is able to translate let's say interpet from the input and output data what happens between the two endpoints and why it comes to this conclusion so is there any legislation that demonstrates uh what makes a code interpretable with an algorithm become transparent if the code was published is ethical design simply the economic interest of the developer or just to build just in his product so these are questions after questions we have you know the the current law you need to comply with and we will see what happens when these uh forthcoming AI law will be in place it has very very similar principles like transparency accountability explainability so this is not new for us yeah but you know we will see how it worked in practice interesting to watch um and then I think you know as as an organization from outside of Australia if I would say medical device manufacturer and I was looking at entering into Europe you know I've seen it around the traps there's the the the importance of the CE Mark uh is that so for organizations that go into Europe they need to be is it certified CA what's that about yeah this is a certificate you need to get before any commercial activity and it's given by an authority and this recording so any any commercial activity at all right like not a medical device thing it's basically any any activity any any yeah and once you have this certificate it's free uh to go to the EU Market you need the local authorization of course and in such medical devices you need to register your product into a EU registry called el domed you would you demand for European medical devices but first it starts with the local Authority uh to to get the license and then you can go out and when you say local that's like in that particular country yeah in in in one country yeah within the EU got you got you and so you know thinking about then the the future or or where you're spending some of your time whether it's with clients or with your your your you know within your lectures or what that in the activities you're performing on a day today what are you seeing are the common uh questions uncertainties or or um the excitement around all of these changes what's in the wings where do you see this kind of going and and what does the future look like yeah you know the king is steal the data how to get to the data so in in some uh other countries they believe that data is a common assets and not the individuals right to be protected in EU we protect data and in Western countries mostly the data protected by the law even you can do some commercials even you can sell it but if you are sick if you are patient you want to be cured and you can sell your data if you wanted to um but you know this must be must be free and up to your decision up to your consent by the law we protect data and therefore it's very very hard for a startup company for example to get to the data and it's very very important that the data is is without bias yes bias could be if there are not enough patience in the database to learn for a machine and it could be a database could be biased if there are there are just certain types of of uh of local communities and not for worldwide communities because it will be uh it will not be trained in a proper way yes so the key is the data how to get to the data to to train a system to uh so this is the key question I think around the world around the world right it's it's an interest it's it's funny how we have this conversation on on um opposite sides of the world and in the end it's quite often it comes back to the data it's a challenge that we're all facing so um look miklos I I'm I love that we're able to make the time to have a chat on the podcast and um we should definitely do it again soon and come and perhaps it's you know when the this uh law comes into you know being around the corner or in full effect and I'm sure there are many uh within the the broader talking Health Tech audience or of course in the tht plus Community that'd be interested to speak more so we'll put your details in the show notes of this episode and I'm sure others will be in touch and look forward to continuing the conversations within the community soon thanks so much I always like to speak to any of you who are very open and they like it very much to share honestly your experiences of course keeping the confidentiality but you know this is a very good Community I like to to speak to anytime Mama thanks Nicholas for more content and Community about technology and Healthcare is it talkinghealthtech.com [Music] foreign