AI in Europe: Building Bridges Between Policy, Business, and Healthcare
A conversation with Jan Kavalírek and Pavlina Walter
There are moments when policy debates stop being abstract and become defining choices for competitiveness. Artificial Intelligence in Europe is now at that point.
The European Union has chosen a regulatory-first approach with the AI Act. The ambition is clear: protect citizens from harmful or manipulative systems, establish global standards for ethical AI, and maintain public trust. But ambition alone is not enough. Execution determines whether this becomes a platform for leadership or a drag on innovation.
In Prague, I sat down with my business partner, Pavlina Walte,r and Jan Kavalírek, Deputy Minister of Industry and Trade of the Czech Republic and Government Envoy for Artificial Intelligence. Jan brings a rare dual perspective: he has built medtech companies and now sits at the EU policy table. Our conversation cut to the heart of Europe’s dilemma. How do we protect safety and ethics without suffocating the very innovators who must deliver AI’s potential?
Where regulation protects and where it paralyses
The AI Act is not wrong in principle. Few would argue against banning manipulative AI, social scoring, or high-risk deployments that could harm citizens. These prohibitions matter. They define Europe’s ethical brand in technology.
But the Act also contains provisions that add administrative complexity without adding safety. Documentation for the sake of documentation. Reporting loops that increase cost without reducing risk. Requirements that even large corporations will struggle with, and that could bury startups before they ship their first product.
Jan explained how the Czech Republic was the first to call this out. On June 6th, they formally proposed simplifying the Act and postponing certain inactive parts to give industry, researchers, and society more time to adapt. That step was more than a procedural note. It set a precedent. It said aloud what others were thinking but not voicing: Europe must regulate with pragmatism, or risk regulating itself out of the global race.
The MDR lesson: regulation without pragmatism kills innovation
For those of us in healthcare, the parallel is obvious. Europe’s Medical Device Regulation (MDR) was designed to raise safety and quality standards. In practice, it buried many smaller medtech companies under layers of bureaucracy they could not afford.
Firms with good science failed not because their devices were unsafe, but because they lacked the resources to navigate endless conformity assessments and procedural hurdles. Innovation slowed. Patients waited longer. Competitiveness suffered.
AI risks repeating this cycle. The intent is noble. But if rules outpace operational reality, they will throttle the very innovators who are supposed to deliver Europe’s AI advantage.
Europe cannot afford another MDR moment.
Global competitiveness: why speed matters
The US is moving fast, with lighter-touch regulation and heavy private capital. China is moving even faster, combining state-backed investment with a willingness to deploy at scale. Singapore, though much smaller, has already positioned itself as a hub by being precise: aligning government, capital, research, and operators in tight loops that shorten the distance between policy and practice.
Europe has an ethical advantage. Public trust is higher. Safeguards are valued. But ethics without competitiveness is not leadership. It is spectatorship.
As Jan put it, “The race is not over. If Europe acts now, with simplification and investment, we can still close the gap.”
The EU does not need to outscale the US or outspend China. However, it must align with their velocity in translating research into deployment. Otherwise, Europe will remain a rule-maker without being a market-maker.
Startups and SMEs: the front line of impact
This is not just about big tech or multinationals. Startups and SMEs are the true front line of AI adoption. They are the ones translating algorithms into products, embedding them into workflows, and carrying the commercial risk.
Burden them with excessive reporting and certification costs, and many will exit Europe entirely. Support them with clear, pragmatic rules, and they can anchor Europe’s competitive edge.
The Czech Republic’s approach is a case in point. Instead of adding layers, they drafted a ten-page national adaptation of the AI framework. Ten pages. Not hundreds. And they did it by sitting researchers, associations, and businesses at the same table with the government. That is governance that listens. That is a policy that works.
The healthcare lens: why this matters for patients
AI is not a separate vertical. It is already embedded in healthcare. From clinical trials to diagnostics to hospital back offices, AI is reshaping workflows.
In trials, AI is supporting:
Patient identification and recruitment.
Eligibility screening and adaptive design.
Safety signal detection.
Data monitoring and quality control.
In hospitals, AI is assisting:
Documentation and coding.
Scheduling and capacity planning.
Discharge instructions and follow-up adherence.
Referral routing and triage protocols.
Each use case shortens timelines, reduces errors, or saves costs. But each also depends on a regulatory environment that encourages adoption rather than paralyzes it.
Excess bureaucracy in AI does not just slow down startups. It delays patient access to better therapies. It extends trial timelines. It inflates costs in already-strained health systems.
The link is direct: if Europe mishandles AI regulation, patients will wait longer for outcomes that could already be delivered.
The investor’s lens: clarity reduces risk premiums
Investors are not afraid of rules. They are afraid of ambiguity.
A clear framework allows capital to price risk accurately. An ambiguous or overly burdensome framework raises risk premiums, slows deal flow, and reduces capital efficiency.
For investors, three diligence questions emerge:
Is the regulatory pathway clear? If not, costs will spiral.
Are founders capable of navigating compliance? Operational maturity is as critical as technical brilliance.
Is the market predictable? Without predictability, valuation multiples compress.
When clarity exists, investors fund earlier and more confidently. When clarity is absent, capital flees to jurisdictions with fewer uncertainties.
This is why policy matters for private equity, venture capital, and family offices. Simplification is not about ideology. It is about capital flow.
A practical playbook: what Europe must do next
For policymakers:
Simplify requirements that do not add measurable safety.
Stage rollouts to give SMEs time to adapt.
Draft frameworks with operators, not just officials.
Learn from MDR, do not repeat its mistakes.
For startups:
Engage early with regulators. Build compliance in from the start.
Focus on use cases with measurable quarterly outcomes.
Partner with associations that amplify your voice.
For investors:
Back founders with regulatory literacy, not just technical talent.
Price risk differently in clear vs. unclear markets.
Seek co-investment models that blend capital with operational expertise.
Jan Kavalírek’s vantage point: business and government combined
Jan’s perspective is unique. He grew up in a family business in medical devices. He knows what it means to fight bureaucracy as an entrepreneur. Today, as a deputy minister and AI envoy, he is applying that experience to policy.
That dual lens matters. It keeps regulation connected to operational reality. It ensures policy is not made in a vacuum.
His optimism is grounded. Europe can close the gap, but only if it moves now.
Closing reflections: the bridge we must build
Europe stands at a fork.
Path one: regulation that signals ethical leadership while empowering innovators.
Path two: bureaucracy that paralyses startups, drives talent abroad, and leaves Europe watching others set the pace.
The Czech Republic’s leadership has already shown a third way is possible: simplify without weakening safety, compete without abandoning ethics.
The future of AI in Europe will not be decided by documents alone. It will be decided by our courage to learn from past mistakes, act with urgency, and build bridges between policy, business, and healthcare.
That bridge is the difference between AI as a European headline and AI as a European system.
And in healthcare, especially, the cost of getting it wrong is not just economic. It is measured in years of delayed access for patients who cannot afford to wait.
Timecode:
00:00 Introduction to Ian Caval and His Role
00:52 Goals and Ambitions for AI in Czechia and EU
01:38 Simplifying AI Legislation in the EU
03:16 Challenges and Support for AI Providers
07:05 Collaboration and Competitiveness in AI
12:25 Personal Reflections and Experiences
18:10 Future Prospects and Conclusion
Links:
Jan Kavalírek : linkedin.com/in/jan-kavalirek-339a1a298
Pavlina Walter: https://www.linkedin.com/in/pavlinawalter/
Peter M. Kovacs LinkedIn: https://www.linkedin.com/in/petermkovacs/
Peter M. Kovacs Personal Website:https://www.petermkovacs.com/
PMK Group Website: https://www.pmk-group.com/
Transcript:
Pavlina: I would like to introduce Jan Kavalírek today who, uh, join us our podcast, who is actually the guy who is pushing the AI in Sugar Republic the most. So I will, I actually let you to introduce to our audience. What you are doing, what is your main scope and what is your interest right now?
Jan : Uh, thank you very much for the invitation. I'm delighted to be here with, uh, with both, both of you. So thank you very much for that. Uh, where basically right now I'm a deputy minister of, uh, industry and trade. Before that, uh, I had my own businesses regarding medical devices, but, but, um, also research involvement in, in MedTech and ai. So this is somehow my, my background, MedTech, and innovation.
And right now I'm a disposition of the deputy minister and also, um, I serve as a government envoy for, uh, artificial intelligence here for the Czech Republic. And, um, actually we've got plenty of goals, uh, we are pursuing right now. But if I should, um, tell it, like, generally our main goal is to create the best possible environment for artificial intelligence here in Czechia.
But, uh, maybe more broadly in the whole, uh, E level because of course we are focused on the European Union in, in many, many topics we're, we are dealing with right now. And we even have like a quite a slogan for that. Uh, check AI to top 10 and e AI to lead, which, uh, symbolize, uh, more than anything, uh, the ambition that, that, that we have.
That we want to, uh, what do we want to leverage our resources that we have here in Czechia, but in the whole eu and we want to excel and we want to be competitive globally.
Pavlina: Okay. Today you had a post on your LinkedIn that you're actually fighting to decrease the bureaucracy, which is related to artificial intelligence in Europe.What does it mean, how you would explain it to the people in the USA?
Jan: That’s right. Thank you. Thank you for noticing it. Uh, I appreciate it. And, uh, actually it's, it's, uh, it's a very, very important step for Step Forward, uh, that we, that we have today, uh, regarding the, the, the, the legislation, uh, as a whole, uh.
Probably what definitely Czech out was, was the first one who, uh, uh, who uh, proposed some kind of simplification of, um, uh, AI rules and the postponement of some non-active parts of AI legislation. It was the 6th of June in U Council, uh, which I attended on behalf of the minister. Of course, our minister. Big, big supporter of these activities, obviously.
Uh, and definitely we are the, the, the, the, the loudest, I think, uh, voice in, in this. Uh, so this is the reason why we are so happy today, uh, that the European Commission, you know, uh, to, to, to be very precise, they open the call for. Uh, expression, which means that the member states can propose concretely what could be simplified in digital legislation, uh, and concretely, uh, also including, uh, ai.
But of course the, the story behind that is quite, quite long from, from the 6th of June. Uh, and what we proposed basically was, was, um, two things. The first one is that, of course we understand that we need to have some control of, of AI system. We totally agree with that. So we support AI act, uh, we understand the risk based approach behind it.
Uh, we also, um, support, for example, article five, where are some prohibited AI systems like manipulative, ai, social scoring, these, these bad, bad AI things, uh, which is, which is totally supported by us. Uh, but what we would like to, um, to mention. Uh, is that there are a lot of, uh, there are a lot of parts in AI Act, uh, that are just, uh, let me say bureaucratic, you know, which means that they will, uh, they will put a lot of burden on the providers of the AI systems, but, uh, they will not contribute to the safety of, of the AI systems.
So this, these, these are the parts, uh, we want to deal with and we want to simplify them. So, so this was our, our first message at the U Council. And, uh, second one, uh, was, uh, that we should postpone some, some parts of the AI Act. Simply to get time for the AI industry, but also for the researchers and for the society in general, uh, to prepare for, for these, uh, requirements.
You know, because we need ized standards, implementation as et cetera, et cetera. So the, the, the environment is not prepared, uh, yet, uh, and we need to provide them more, more time. Uh, so that was the start of all these events. And after that, uh, we, we tried to, uh, find some, uh. Friendly member state that will support us in this, in this initiative.
I don't want to be too proud. I don't want to say that we are the only ones, uh, that, that that's not the case for sure. But we, we definitely, we were the first that mentioned it out loud officially, you know, and, and this was important of course. Uh, and right now we've got quite a coalition of, of a relatively lot of member states supporting these simplification initiatives.
Uh, I myself, uh, had plenty of meetings on this topic with, with, uh, the leadership of European, uh, commission with the leadership of Gigi Connect, but also with the ministers from, from member states. Uh, so there are plenty of member states supporting right now, and this is the reason why European Commission said, okay, uh, we, we, we listen to you, we listen to your needs.
Uh, let me say it like that. So we will prepare the simplification omnibus, which is, uh, like quite a strange term, but what it means, uh, uh, basically is that there is a room for simplifying the rules of the AI to, to reduce the burden. This is exactly what we want from the beginning. Uh, and this is the reason why they today opened this, this course.
So, but, but we are a bit step ahead, uh, yet. Uh, we, uh, we have started to prepare our national concrete, um, proposals, what we want to, uh, what we want to simplify and all, because at the be beginning we were like quite in the minority. So, uh, it didn't make any sense to prepare something concrete. We just wanted to open the debate that we will simplify and to get the support, but now we have the support so we can, uh, we can move forward and we can come out with concrete measures.
And again, as, as previ as previously, uh, and it is the same thing right now, we are preparing it together with the AI ecosystem, uh, which means that a lot of associations, um, for example, a check chamber of, of commerce, you know, and, and confederation of industry, they're all involved. In preparing all of these comments, so, uh, keep fingers crossed and hopefully we will be able to present something meaningful and impactful.
Peter : I think it's really critical because you have to be competitive. European Union should be competitive with us, China, Singapore, all the other members or state. Um, they're not having so strict regulation, which is of course could be good or bad, but, but, uh, definitely it should be competitive because the European Union is a bit behind of, of the US and China.In this field.
Jan : Yeah, yeah, yeah. You are, you are totally right. And I think that, um, it's, uh, honorable to, to, to say it. We, we need to be, um, open in this. And right now there, there is obvious a gap between us and, and eu. Uh, but I'm very positive that we still can close it, you know, and in this, in this particular thing, I totally, uh, agree with, uh, president for the Lion and she stated it like a few weeks ago that the race is not over.
And yeah, it's not over. You know, we are at the beginning, uh, for sure. Uh, which means that if we will start now and we will do the proper steps and we will invest properly. Uh, we still can build the infrastructure, you know, and we, we can reduce the bureaucracy. We, we can have, uh, our language models. It doesn't mean that we shouldn't cooperate with us.
Uh, not at all. I think that we should cooperate, but, but as equal partners, you know? But of course, it's kind of a difficult question right now. There are a lot of political debates on that. I would like to, uh, stay, uh, stay in the maybe more professional area or in a political one. So let me say that I think that the collaboration is beneficial for both sides, but uh, we need to collaborate as equals.
Pavlina : I mean, it takes a lot of time. So where you actually see that it might be implemented, this simplification?
Jan : Great question. Uh, if now is, now the call is open. Uh, the, the next very important meeting will be 24th of October. AI board is like the, the, the, the, the highest level of, uh, of body, which coordinates the agenda on the U level.
I think that it will be quite a, a difficult debate, uh, on this, on this AI board. And then we will have next EU council devoted to this topic, but it will be in December. But meanwhile, I think that, uh, the European Commission will, uh, propose, uh, their, uh, idea about the simplification. But of course, it's, uh, it's not the end of the process, not at all.
Then it'll go to the trial log. Uh, between, uh, parliament and between the U Council. So we will give our comments again to that. So realistically speaking, uh. The first half of next year maybe could be the, the end of it. Uh, and I think that it's, it's probably also the deadline though, because, for example, when we are talking about the postponement of the non-active parts, uh, for, for the businesses of all the environment, uh, these parts will be active in October next year.
So we definitely need to, uh, agree before this term of course.
Pavlina: And, um, this simplification will have the biggest impact on which sector.
Jan: Probably on the providers of, of AI systems, but it, it could have impact on everybody. Uh, almost generally speaking, you know, because, uh, when we are talking about the AI providers, they, they are the ones that they have to follow the rules and fulfill the criteria.
But of course, if we will put a lot of burden, a lot of bureaucracy on them, it, it, it basically means that the cost of the air tools will be higher or they will not be available at all for the new market. So, uh, so at the very end, the impact will be on, on our, uh, on us, of course, so it can indicate influence almost everybody and also the AI researchers.
For example. Um, earlier today I had a, a great meeting with Professor Haj, and he's quite a well known, uh, AI researcher here in Czechia. He's responsible for developing new, uh, UI models. He is leading a European project in this. He's also very interested if, if we will be successful in this because he's, um, inventing a foundation model.
Uh, but it also, um, uh, it is also regulated, uh, under this, uh, AI act, you know? So I think that now it's an important topic for the whole ai, um, industrial e. Yeah.
Peter : But also I think it's important to highlight that the simplification do not affect the safety. 'cause European Union Yeah, is, is much stronger on the safety level than us, compared to US and China.
They are not care so much, that's why they are faster. Mm-hmm. Mm-hmm. But I, I think it's important to highlight that if I know well. That the simplification do not affect this, this safety.
Jan: Yeah. I, I think that definitely we can confirm that, uh, because, uh, it's not even our proposal of course. You know, we, we, we really stay behind these ethical rules and, and, uh, that's great that you're mentioning it, uh, or highlighting it.
I think it is important. And I think that it is a commodity of EU that we are like, like a SS, that we've got some rules, some ethic rules that we want to implement for using AI in, in the society and in, in our everyday lives. So definitely we don't want to question this. We are just aiming, uh, to the, uh, um, extensive bureaucracy and, uh.
Pavlina : You mentioned this AI things. The bureaucracy. You came from the medical device field. So now I will change a little bit topic to more easier or maybe easier. So which actually part of your life was more fascinating, more easier or more exciting. It was when you were developing the medical device or now on the government side fighting.
Like for the AI in Europe,
Jan: everything is exciting. I have to say, if I can be a bit personal, um, I think that the, the, the thing is that you should, um, try your best no matter where, where you are. And also when, when I was doing the, this kind of business and it was like a small business of course, of. My father started it previously.
Um, it's very difficult to, uh, to do any business. I, I think so. Uh, so I really support anyone who is doing business in the whole, uh, European Union Union because the environment is difficult. The bureaucracy is difficult in here, and it's a great experience, I think, to enter politics and to enter this level of, um, decision making.
Let me say it like that. Uh, because, uh, then you know what you are talking about and you can imagine what the business is, is struggling with, and you, you maybe even listen to them more or be better. I think that it's, it's an important background, uh, to have. Of course, I don't want to question any background, but, but I, I am, I'm happy that I, that I, uh, that, that, that we were doing business in medical devices and of course, uh, another great advantage.
Uh, concretely for this, uh, AI legislation process is the conformity assessment because it's. Almost completed the same thing, you know, and, uh, we both know it pretty well. Uh, and, uh, considering, um, how the MDR uh, was, uh, implemented, it was very difficult for a lot of companies in the whole eu. And especially for the small companies, of course, uh, and right now it, it's starting to look quite similarly.
And this is my, uh, this is what I'm a little bit afraid of, you know, that, that, that, that we will really burden the businesses under, uh, similar amount of bureaucracy. And I think that we should. Well, maybe it's too harsh, but I learn from the mistakes that we made previously and not to repeat them. Uh, again, with, with the ai, uh, systems.
Pavlina: When you mentioned that you were actually first one as a country fighting for the simplicity, why, why Czech Republic was nearly the first one. The others, they didn't see this issue or they were not so much, uh, progressive.
Jan: It’s a great, uh, question. I can, I can just assume it, it was my first u can, you know, maybe I, I was, uh, maybe it was, uh, it was the reason that, that I were new there, so, so I just said it.
Uh, I, I think that the, some, some states, for example, Poland, maybe, um, maybe even France was thinking about some kind of simplification, but, um. But, uh, it's true to say that we really were the first to mention it at the table officially in front of all the cameras, and it was somehow the starting point for the discussion that, that, that followed.
Yeah.
Peter: How do you see how optimistic or pessimistic you are in the next five years in the field of startups and SME companies in the EU in this field? They can be optimistic.
Jan: Uh, I, I'm generally very optimistic of ai, of course, startups and SMEs. Uh, it's, it's sometimes it's difficult for them. Uh, but this is also the reason why we are trying, uh, now I'm speaking for check only, uh, because I can speak on behalf of our government, of course, right now.
Uh, we are trying to, uh, to find every single way how we can support them, you know, and we've got a lot of initiatives for this. Uh, some of them are, are even come or are coming from the Ministry of industry and Trade, where I'm now like grant programs for research and development activities for, uh, implementing AI tools.
Uh, like innovation into, into real world, into practice, because this is what we have to achieve. You know, I think that we've got great area researchers in, in, in Czechia, definitely. And, uh, we've got, um, um, Czech ai, we've got, uh, that center of robotics, informatics. You know, we've got SAG and the University of Australia.
A lot of great academic institutions. We've got a lot of great, uh, researchers in here, but we have to support them that they are able really to, to, um, do the research and development here in checkout. And not coming to us, for example, let me say that. But, but I think that this is maybe problem of the whole eu, uh, the whole European, European Union.
So right now we are, uh, trying to explore how to support them and how to also support their startups and their small and medium, uh, medium businesses. And we are very open with that and we really, uh, communicate with them or listen to them. And for example, uh, right now we are drafting our national IO uh, we are quite proud that it will be like a really, a minimalistic.
Adaptation of the AI into the C legislation because we don't want to put there any more bureaucracy anymore requirements. So it is just 10 pages, you know, it's amazing. It's just 10 pages long document, and of course we prepare it together with the AI ecosystem. In here. We, we established like a national AI board, something like that.
And there are, of course, there are also the ministries involved, but, but also the associations and the companies and also the researchers. So we, we really try to consult all of these things together with them and this is also the reason I hope that they support the activities that we are doing right now.
Pavlina: Okay. And would you actually, after this experience, come back to the medical device sector? Would you again start and develop your own company? Definitely. Yeah. Yeah.
Jan: De definitely, of course. Uh, this work that I'm doing right now, uh, you can never know how long you will have a chance to, to do this. And this is also the reason why we are so focused and I'm very proud of my team on Ministry of Industry and Trade, and I, I really.
Uh, I'm really happy for the collaboration with the minister and I really think that we, um, we used every single day from of the time that I have been there so far, and I will work to the last day that I will have the chance. But of course, um, uh, then if for example, I, uh, there will be some changes in the government, whatever, I don't want to take this topic, um, too deeply.
Uh, then I will definitely come back to the private sector and I will continue with, with some of my own project, but experience, uh, or experience with the, with the government level activities.
Peter: Yeah, it's very nice. Thank you very much for sharing with us all, all this information. It was very useful for, uh, thank, thank you.
Pavlina : Thank you. And we really appreciate your hard work, which we're like doing, uh, on behalf of us for Czech Republic and, uh, in the. So thank you so much.
Jan: Thank you very much. It was, it was a lovely conversation. I appreciate it. Thank you.