Kala before we start h perhaps you can uh say something about uh um the center of digital society what you're doing and what are your main goals >> sure happy to do so can you hear me well >> hey >> not yet >> hello >> I have a loud voice but I'm not sure if it's enough >> hello >> without technical assistance >> wait a second we we're going to fix this quickly there's not a problem there going to manage it please. The second um microphone on. Thank you so much. We appreciate it. So now for in stereo and in quadruple sound. Can you tell us uh what your center is doing? >> Sure. Does it it does work now?
Amazing. So the center for digital society at the Macato Foundation. Oh maybe I should start with what is the Macato Foundation? It's a private foundation. We fund um projects from academia, think tanks and civil society in four thematic areas, climate change, social cohesion, participation, Europe in the world and I'm leading the fourth one which is digital society. So currently we're funding around 40 projects on issues such as the fight against misinformation, safer digital public sphere, the digital transformation of the state or a topic of today digital sovereignty. Mhm. And what is what what did you achieve for example in the in the last year? What was kind of project that you were doing? >> I mean I have 40 different projects that my team is managing and I love them all obviously.
So I always feel uh like it's a very hard job to take one out in particular but uh I need to do if I need to do so I'd love to talk about the Agora digital transformation because it's a new think tank that we actually built around four years ago and their focus so far has been very much on the digital transformation of the state. So they've um consulted the ministry of digital transformation of questions of like impact um focused work or they've they're the first ones who actually calculated how much the German government has been spending in the past years on the digital transformation. Quite interestingly this database the most common user are the ministries themselves because they didn't know um yeah and they're trying to create a space where different types of stakeholders can meet and engage on these issues.
So across the political spectrum, but also from the private sector, civil society and academia. And I think that's that's this part is what I really like about it because it's also what we're going to do in the panel today. >> And you're exactly referring to the right one that this is the panel with the most theatrical title, competing with giants, Europe's own way forward. Please KA introduce your dragon slayers. Thanks. >> Yes. So I might actually start with our digital participant. Can you you cannot see him yet. Oh, there he is. Hi, Kai. >> He's joining. >> So, Kai is joining us from Brussels today and he has a very good reason to do so because he's there to participate in the omnibus debate on the AI Act and I hope we will hear some gossip later.
Um, but who is Kai? Kai is truly a political animal. He's been working in Brussels for more than 10 years on digital policy issues as the policy adviser to uh MP Axel Fos and I think there's few people in Europe probably that know AI regulation and related uh digital regulation as well as him. He's also currently a fellow at the Teo Munich and one of the brains behind the one of the Euro stack movements. Welcome Kai. And then I want to welcome Yeah. I want to welcome my two panelists. I think you can already both come to sta come to the stage who are actually here in person. Leoni and Alina. Um I will start with Leoni.
Choose your seat. I will sit over there. So take one of those. So Leoni Noihoza is the is leading research teams at I know that was also mentioned in the last panel as one of the few German AI unicorns and she's responsible in leading uh the research teams in their post LLM training and reinforcement methods but she also brings um to this group experience from the energy sector and from academia because she actually did research at aha on questions of diversity university and fairness and automated decision-making. And next to her is Alina Blancat. Um Alina actually started her career as an economic consultant in Berlin and in Oxford for Oxa Consulting where she advised various um industry stakeholders like media industry and manufacturing on questions of competition in digital markets and then she stayed within the theme but changed to the think tank world and uh kind of built the data economy project at a think tank back then called SNV and she's now building a new organization called Rebalance Now that has the goal
of fighting monopolies like the ones of big tech. So I'm really excited that we have such a truly diverse panel in terms of sector experience but also disciplinary backgrounds. Um and I thought that I'd start the discussion um actually with you Kai because you are today's gender minority in this panel and also uh since you're joining digitally I wanted to ask you about um the state of the debate the feeling in Brussels um with regards to digital sovereignty I feel like this term and in general the awareness of the problems of European dependence of big tech is like higher than ever. But at the same time, my feeling has been now for a while that we are kind of stuck between admiration and fear of what is going on in the US.
So how is the feeling in Brussels? Is it hopeful? Is it energetic? Cynical and kind of what policy goals are leading the debate on this issue right now? >> Yeah, very good question and I fully agree actually with your short summary. Um it's actually quite similar here in Brussels um as it has been in the past as many of you known there was a sovereignity discussion uh with Ptor in 2019 2020 when I came to Brussels in 1415 we had another sovereignity debate and yes there is now definitely as you said a lot of momentum in in pushing um certain ideas. It's also um true that we are much deeper in this discussion. It's um yeah certain areas or certain items are now being discussed that so far we're not really on the agenda um of let's say the highest political level.
But at the same time you still have those two let's say extreme positions. First one is we need to kick out um all non-European um tech players especially those that are not in line with um our European values. Um it's often yeah combined with at least some idea of protectionism of fostering national champions and so on. And then we have the second camp um that is saying no we cannot do it. it's not in line with actually European values. We are here for a free market. We want to stick to um WTO rules and so on and so on. And despite let's say this progress in in the debate so that we are now discussing also details and it's not only about let's say pro and cons in terms of those two general positions.
we actually are still always coming back and uh to those extremes and um when now in two months the uh European Commission is making this proposal of the so-called um sovereignity or taxity package. It's still not really clear which of those camps uh will succeed um to to convince the commission um yeah where the majority line among the different commissioners and um the president of the Lafonda line is um which was why and with that I would close my short statement on it um which is why um myself and many others we are pushing for a kind of let's call it third um with a paper called European way where we were saying okay look it should not be about outer key about protectionism and so on.
We cannot do everything um alone but we should definitely concentrate on reducing our excessive unitary dependencies and yeah check in which areas of the tech st we are leading or we have the potential to lead. of course they are in particular focus with our investments with our policy activities. Then however also check where yeah it's necessary that we um have certain minimal domestic alternatives but then also accept that there are some areas where at least right now we can only um yeah focus on maintaining credible fallback capacities. Um we think that with this approach we are much closer to the reality which is again that Europe cannot succeed in our opinion to lead everywhere and to to kick out everyone that is not from Europe but at the same time we would significantly reduce those excessive dependencies that um I have mentioned.
Um in the end again it's not really clear if this middle approach will make it. if one of those more radical approaches um are making it. But um in the end of May, we should know more because then the commission as I said will make this proposal um called tax reality package which will include um an update on the public procurement directives um the so-called Kaida cloud AI development act and other items. >> Thank you Kai. Um, Alina, I feel like the digital sovereignty bait sometimes is also unproductive because people have a lot of different understandings of what the term means coming from different like political goals, but they don't make this conflict um explicit.
And you're an economist, but you work on societal issues on strengthening democracy. So with those um kind of two backgrounds, which perspectives are currently missing in the digital sovereignty debate and how is your perspective on what Kai just described? Well, so from this societal [snorts] perspective, I think what we really also need to bear in mind are questions of what I would call democratic and regulatory sovereignty. So where democratic sovereignty means that we should think about how technology actually serves goals that are democratically legitimate and I know of course tech firms also need to be profitable and what we see right now is in many areas really cutthroat competition so like there's not a lot of room for maneuver but at the same time like that means we should still think about the impacts that technologies create and well the tools we have to steer technology in a certain way still is through laws and regulation but also through investment for example that can be channeled in a socially productive direction and this is where
then uh regulatory sovereignty comes in because we need to make sure that firms still adhere to law to laws so I mean I guess we all have an idea or can think of pieces of regulation that we consider useless or unproductive, but that doesn't mean that we can just ignore them. And what we see right now is that democratic institutions at at the EU and national levels have regulation in place, but certain companies can afford to deliberately break them. And that is of course a huge concern because it doesn't just uh call into question kind of questions of like okay how much societal benefit do we get but like are we actually risking principles of rules of rule of law and um sovereignty of the state.
So and we and that is also economically relevant because it gives those companies who can afford to break laws an unfair competitive advantage visav those companies who do adhere to laws. So there are various reasons for making sure that we have this regulatory sovereignty in place. So um Kai talked about the different camps and you just said that the companies can afford to deliberately break laws because of their size and because there's few alternatives. Um would you say that from this derives the like your affiliation with the first camp that would say let's replace these things with European companies because with those we can regulate them better. I would agree with Kai that it's not that binary actually because we need to think about dependencies and risks kind of beyond just kind of where companies sit because also European companies could become more powerful and wouldn't necessarily behave you know more nicely.
So that is why we strongly advocate for the use of competition law kind of to focus on the concentration of economic power because irrespective of where economic power is situated, we should make sure we have a more level playing field so that we can actually challenge existing uh power concentrations from where we stand. That isn't easy, but we should certainly start with enforcing the laws we have, which is currently not happening and that is kind of really the first step we need to go. I would love to I will come back to this later actually. I also want to hear the gossip about the omnibus, but I want to go to you first Leoni um and talk from like your perspective as a representative of a German tech company um about your understanding of digital sovereignty.
you are from my understanding working with your own uh foundation models. You are um you have been working and are still largely working with European infrastructure but DeepL also recently announced a new partnership with AWS and this is also gained some attention. Um, so maybe you can share a little bit of the thinking behind this decision, why you had to make this move and how you can still be digitally sovereign despite this swift sh despite this change. >> Yeah, absolutely. I think that's a that's a very important point to look at this like in a bit more of a diverse way than just talking about digital dis sovereignity in terms of where does your data sit.
So I think there's like much more to that topic and you already talked about a lot of ownership topics which I think play a lot into that. So like ownership of your models in order to be able to innovate on them. Also ownership and a very good understanding of the value chain that you have in general and the problem you're trying to solve. I think this is like a very much underestimated part of digital sovereignity as well that you don't just build something you don't know what to do with it but you actually understand the space in which you want to change something with that product um and I think this is something where we are particularly strong in because we are vertically integrated in that way like we we have our own data centers um and we also train our own models in these data centers but we also have like a very good understanding of the problem we're actually trying to solve which is translation in like a very nuanced way which is
also not a simple problem in that regard. Um however um I think at some point you have to also think about if you want to provide this model quality and this product quality to the whole world like we have customers for example in the US and Japan etc. Um just as a very practical example latency is a very important um problem and in the use cases we have like if you have real-time translation voice to voice it matters a lot that the model can like inference fast um if you have customers in Japan you have to provide that model in Japan in order to provide that quality which also is about reliability and trust and that etc.
So this is something that plays into this hybrid cloud strategy that we're approaching right now to actually be able to deliver the quality and reliability of our um of our um products to customers worldwide which then also I think in some part is an aspect of digital sovereignity to be able to scale that to um to these countries. So I think in that sense it's not an eitheror and you don't um while I must say that of course we would love to do that on European infrastructure it's just not there right now um I think this doesn't put into question the general the general aspect of digital sovereignity and be in terms of being virtually vertically integrated in your model development and value pro provision >> and do you think that this is um kind of a necessary step in between to a goal where you could um work entirely on European infrastructure or is this already the end goal?
>> So I think that's a very good question. I think as soon as we would have scalable European cloud infrastructure available that would definitely be a very good goal to go towards too. Um so I think there's um more >> political decision-making in that space necessary to go towards that. Um, so I mean I wouldn't speak for the future that we would stick with that strategy in that regard, but this is like a hybrid strategy that we're going forward for now. >> Mhm. And Kai, what do you think is the prospect of us getting this European scalable cloud infrastructure? How hopeful are you? >> Well, on that topic, it's not something AX and I were really focusing on.
it was always a side topic but um yeah going back to this kind of uh clash between the two main camps I think there is a huge risk that in the end policy makers will look for a very broad compromise that doesn't really help in the end so I'm particularly afraid that there is a very vague definition of what is uh being or can be classified as European for example which companies are European and um this question is also not easy to to solve as most of you know um the docom has I think 70% um of uh their global income coming from the US so if you would use this criteria suddenly telecom would not be classified as Europe or European There is of course also an option for example company like Microsoft to um just split and create Microsoft Europe um um company to again undermine or um go beyond a certain criteria that was not well thought through.
And again, there's also the question, do we actually want to to exclude certain non-European companies? Are we um capable to to fill up those um um those gaps that such an exclusion uh would lead to? Um my point would be again yes I I started co-started was a lot involved especially at the beginning in the Euro stack movement and I still um believe in all those things that we have said but um I really hope that we take enough time here in Brussels to clarify on all those details that I have just now a few of them mentioned in order to make sure that if we can agree on on a common for example, what the European preference is, that this doesn't backfire, that this doesn't lead to unintended consequences.
Um, for example, those companies um that we want to exclude uh are able to find loopholes and maybe other companies that we don't want to exclude um we would remove. Or another um potential um bad outcome would be that it's connected to a kind of external verification process that again in the end just makes um large law firms, large accounting firms even more rich. Um but in the end also doesn't really produce um a good outcome. um because with a lot of money you will always be able to go through the process. So with all those points um again I want to underline it's not really clear where we are landing and again like so often before there is this really big risk that we are creating something um that is not really effective on the ground and is not leading to the policy outcome that a majority of political groups um and member states are looking for.
Um, Alina, a lot of people in the room here are experiencing because they represent, for example, AI startups in Europe, the specific dynamics of digital markets. And if we come from the investor side, we see how investments that go to European startups, a lot of this actually ends up also fueling um, US tech and um, you are an economist and you've worked for a long time on like the specific specifics of digital markets. I think we're not all economists and even though we experienced this, it would be great if you could in short, not as a university lecture, but explain what is so special about digital markets that requires new forms of intervention. Um, and maybe you could connect it with a comment on the letter that you recently initiated together with I think 70 other civil society organizations addressed to Özila Fondonderline.
Um I think the Google ad techch case was the occasion but it has a more general message with regards to remedies. Yes. Um sure. Well, what's special about digital markets, and we've seen this already when it came to platforms and so on, is that they're not markets as we used to know them in competition analysis where you can look at it kind of market by market, but they um platforms uh to use it first as an example kind of shape markets that they give access to and AI does so at a kind of even higher level and the so kind of through AI um the big companies especially the cloud companies get access to what many people call chokepoint power and panopticon power.
So basically they can decide whether those markets work and they can kind of get a high level view of the entire economy and the markets that actually use cloud. And the challenge is that in the competition framework we still look only at individual markets and we don't really see power as it accumulates across markets. This has changed to some extent and we have some regulation in place like the digital markets act that actually a lot of startups support and still call for for its enforcement um because it actually gives them some more room to grow. So kind of the big gatekeepers cannot just occupy everything. But especially in AI, a big problem from a com competition perspective is that the big tech firms have been able to shape AI through certain paradigms that mean that a lot of those AI business models actually ultimately end up feeding big tech firms.
There is really interesting analysis that shows that for example academic AI research is heavily uh influenced by big tech firms and also big tech firms invest in many like just Google alone over 6,000 companies across the world. So they don't even need to own them. they can just invest maybe grant some cloud credits or just basically make sure that they only they remain limited to markets that are complimementaryary to Google and they Google just benefits when the cloud grows. So um the problem is that a lot of ecosystems are just dependent on them. Um so like and I think also a lot of AI companies understandably don't enter into competition with them um but rather hope for an exit a successful exit with with one of the big giants because that's how this works right um Leoni so uh actually in the round before I was surprised nobody mentioned to the question what's your favorite uh European AI I solutions/star.
Nobody mentioned depot but we have you here now. Um and it's quite impressive. Grown from I think 22 people I found in 2017 to 900. Um you were in 2024. Uh the most highly valued AI startup in Germany. Um and you've done so in a market that is otherwise largely dominated by companies like Google. You are literally competing with the giants. So how do you do so? despite the market dynamics that Alina has just described. Is it better technology? Is it um the promise of safety, security, privacy? Is it um niche finding niche applications? What's your recipe? That's a that's a good question. I think it's a little bit of all of that to be honest.
So I think um DeepL has already been uh in this world of machine translation since 2017, right? So like before large language models even were a thing. So we have really I think found a problem that's worth solving and solved it with AI and continuously improved on the technology that we've been using for it and also maintained that's what I was talking about in terms of being vertically integrated. So we have powered our own data centers and trained our own models since 2017. So like we have like a very long history I would say of inter um in-house expertise both on like the technology but also on the problem like we have a lot of language experts we work together with both inter inhouse and and external we understand the um the space we're going in with value provision very well and I think all of these things come together with just having a worldleading product and I think this is something that's very often forgotten in this whole discussion about European sovereignity.
Um it it's not enough to be European. We want to be world leading and we are really world leading in this specific problem we are solving there. Um and yeah as I said I think there's a lot of talent that comes into it that we've been managing I think to to have internally and also to keep that have been building this for a long time already. um and we've understood uh where we want to go with this and what actually is value provision in this space very well and I think something that also comes into it is that um we're talking a lot sometimes about the contrast of do we need like a frontier model in Europe or like do we need foundation model training in the first place um the foundation models that we're training are very specific to this problem that we're solving which is translation now we're going to another space which is not textto text translation but voicetovoice translation but the USP is still translation and we understand this problem very
well also in this audio domain now and can train very specific models that are tailored to this problem that are not general purpose and provide well plausible solutions plausible translations but actually um coming back to the values you just mentioned reliable and exact solutions we build on like data trust this is all um also in terms of customers we we primarily work with customers who do care about these values not because they care about European values but because for example they in very regulated industries and they just have to work with these values because it's for example very important for a global aviation manufacturer if he wants to like translate all of his safety manuals so that they are like reliable and safe and you can trust that that translation.
So um I think this is like values that are European but that just provide the product value that we are solving at the same time. So um I think in that sense yes to all of these aspects. Um yeah and in the panel before and I think in many others um many other rooms generally where digital policy and AI is being discussed people become very wor worried because Europe doesn't have a competitive frontier AI general purpose model. Um, from what I'm hearing, are you are you worried about this as well? Do you think we don't need to because we will have more depots in the future? I think there's no eitheror to be honest. Uh, I wouldn't I wouldn't make a donutia out of this.
Um, first of all, I as I said, I think that foundation models can be very very tailored to certain problems and not just be a wrapper around a general purpose AI. Like I think there's two different sides of what application means here. Um I do think that it makes sense for Europe long term to be strategically independent to have a frontier model but I think where Europe's strength really lies is in applications and al also data in certain domains such as legal or or manufacturing industry etc. And focus about value provision in these areas and tailoring models to that I think is really where Europe strength lies and we should focus on. Kai, one of the um demands of the Euro stack movement is to use public procurement for the strengthening of the European tech ecosystem.
And I mean if we look at the US, a lot of the tech ecosystem there has grown because of money spent uh through public agencies. DAPA is one big example. yet when I talk to to policy makers uh I often hear this argument of European tech is not ready yet for for example German ministries. Um how would you respond to this? What's your assessment of the role the public sector can play in funding European tech? >> Yeah. So I would say I would completely agree with those people who are saying that it's an underused tool. Um but yeah also there similar to my previous um comments um I think like so often in life a middle ground solution is the best way forward.
Um if I remember correctly um there was an attempt in Munich um some years ago um to to um use a lot of opensource um products um in the public authorities. it was not really well prepared to what I understood or there were were some push backs and those things or those episodes of the past are often showing that if you make a radical shift um in for example using more European products and not preparing also your staff um increasing certain skills um in the deployment of uh tools they haven't used before and so on, there's a huge chance um of um yeah backfiring um um developments and therefore what I would say well it's a German word um the state as anund so as basically it's the key customers it's a customer is definitely an important um tool we should use more and I think it's for example very promising what um our um government um did now when it comes to open desk um together with the French government and um the the
yeah key example how they did it um even though there were some problems at the beginning um did exactly what I was saying they prepared it over months there was a lot of um actors being involved. They really tried to include as many um people also from the outside in order to to make it work. Then the the chances are quite high that it will be at least um um some kind of success. And another um example more theoretically but um there was I don't know if it's a public example therefore I'm trying to make it or to anonymize it a bit. There was one stream in let's call it um European public authorities that was actually thinking about going in a similar direction like with open desk but by still using basically the Microsoft operating system but having in parallel Linux um operating system and then depending on the programs uh you want to use or depending on the end user on the staffer basically everyone everyone would be able to switch and I
think this especially for a kind of transition period um is an extremely smart um way forward to yeah to conclude um my response to your question is um public procurement by public authorities um and using this tool by public authorities in my opinion is a great step forward and also the uristic movement was making this point um from the beginning. But again um going back to what I said before, it needs to be very well planned and probably it's smarter to not make a super radical step um removing everything um else that is not this objective or this new um services or products that you want to use. But by yeah going um an approach where you use at least for some time things in parallel.
M >> I kind of if I may jump in there for a second because I just feel quite strongly about the Munich example because I I mean if you look at the facts and what happened like they actually had a fairly successful transition to Linux and open source solutions but then actually Microsoft came back in and said like hey actually we would really love to uh like create some jobs around Munich. how about you reintroduce uh Microsoft into your public administration and that was just a very successful um intervention. So I think it's really important when we think about procurement to also make sure we still have the technical capabilities and understanding in the public administration um and to make sure people know what they are procuring.
I mean this sounds so obvious but I feel like especially with AI it's actually becoming harder for a lot of people because there is a lot of magical thinking about AI and I'm actually quite uh positively surprised that there are discussions here also about like cloud versus onrem uh and so on which in a public administration context just don't happen like everybody's like okay we need to like do as much AI as much cloud as quickly as possible without even getting the underlying data straight and that's a huge risk. So um being aware of what to procure and what the problems are that people should solve with tech and with AI is extremely relevant if we want I mean that's a kind of sovereignty that's kind of the the first thing we need to get right because if we like if if we uh mess it up at that stage it really doesn't matter whether we get a EU or US provider for that.
We have very little time left and I want to go a little bit into what I consider the f false dichotomy between regulation and innovation. At least if we discuss it in a way that it's often happening right now and I think that's probably something everybody on stage digitally as well as physically would agree on that we need to look at like how exactly does regulation look like. So I want to get your like practical perspective and understand better Leoni how it's affected things like the AI Act has affected your work and before the panel actually I tested depot the typical ethical question with translation software is like for example gender bias so I tested when you type in the doctor into German comes out dear arts so the male version when you type the parent it comes out duta but then it was quite interesting because I typed the German chancellor and when I got debund um might have made me a bit happy and I wanted to ask you how do you deal
with these kind of ethical questions and has it changed since Europe developed AI regulation? Uh that's a super interesting question because I think it very much ties together again with the quality of the product because in translation uh you want to like the quality is kind of not always but kind of clear right like you want to have an exact translation of what you put in. So a gender bias for example if you have like gender in one language and no gender in another one is like either not depend like not not specified it can just be ambivalent or it is specified like later in the sentence for example and then it changes the meaning like kind of hindsight and then the model should always get it right but it's kind of a very clearly defined problem if you can do that exact or not and coming back to how do we deal with regulation there um again I think this ties very much together with the quality we want to provide anyways like we
have very specific um evaluation sets and evaluation benchmarks inhouse that are curated together with language experts on specifically those kind of problems like taking gender bias there but there's also other things like for example how do you transcribe numbers like there's a lot of ambivalence in that and that can also go terribly wrong down the line. So we have like if if you have like a problem that you care very much about, you can like design a specific evaluation set that like tackles that problem and you kind of try to make your models as good as possible on that. And how that ties together with regulation I think is that in the end I think regulation should incentivize exactly that right like this is trying to provide a good quality of what you're what you're delivering and having an idea about how data for example plays into that how your model plays into that how evalation plays into that having that well documented in the end that's kind of clean working I think where regulation
maybe becomes a burden is if it if like these values that are actually trying to be incentive um just become a checklist which like in creates a lot of work but in the end it would actually just be what like people in my team for example developers would just want to want to achieve in the first place like creating a good model that does exactly what we're promising customers to do. So I think this is a little bit like I don't see um the contrast between regulation and innovation there in the first place because I think it is incentivizing what we want to do anyways. However, if it becomes a checklist I think this is a problem.
And do you see any change in how you work since the I AI Act came into for into force? >> Actually not because we've been building with these values at core from from the beginning on. So I think this is really something we don't have to re-engineer um because like reliability for example of the translations we're providing has been one of the core selling points of the product we're creating from the beginning on. So a lot of things that play into that model reliability etc is just something that we've been very cleanly working on. Anyways, >> Alina, before I go to Kai to spill the beans, um I want to ask you what's your wish or your demand even?
You're an activist, so probably we can talk of demands towards European policy makers when it comes to uh you said regulatory sovereignty, better enforcement, um also in the context of AI and the dynamics that we've described before. >> Yeah, I'm sorry. I actually skipped your questions regarding the letter to Fundline and uh what we are demanding right now is really for enforcement of the existing laws and not bending to a US government that is trying to get the EU to subordinate. Obviously there are complex geopolitical questions about like how do we deal with export risks and kind of we are dependent on the US in many ways but we need to think about how to manage these dependencies and not just kind of continue with business as usual while somehow being like oh but something isn't like they're not as friendly as they used to be.
Well, then we just adjust our laws in a way that makes them actually meaningless because I mean as I as I said earlier this doesn't just un this undermines the rule of law but it also undermines the prospects of having tech that is built on different principles because if then for us companies we're very happy to go like oh yeah just you know abuse your dominance even within the current legal framework How can we make a case for a European or kind of a global level playing field? That is a huge concern. >> Kai, what's going on in Brussels? Are you hopeful with regards to what's happening in the context of the omnibus debate? Is there something we should pay particular attention to?
And maybe what's the timeline for this? I don't know if everybody's so informed, so give us some context. I cannot help myself to just comment very briefly on what my co-panelists said before because um I agree completely with Alen um that >> and we have one minute by the way. >> Okay. Okay. I make it very short that the main issues indeed when it comes to market concentrations where competition law that has not been used properly is the lack of enforcement um and so on and so on. However, yes, of course, there is also, at least in our eyes, certain overregulation maybe, but even more, it's in our view um laws that are not really coherent that have good intentions that are trying to do the right thing or to address uh concrete problems, but uh they're just written in a way um in policy silos and not really um put in a as I said in a coherent problem, in a coherent way.
the AI omnibus and this whole simplification push of the commission is trying to fix that to your question. Now my personal opinion shared by my boss and many others and all think tanks here in Brussels is however >> 10 seconds. >> This is not real simplification and the omnibus is not really helping uh that strongly because yeah it would require much more evidencebased work before and based on this evidence uh you could then do proper simplification. that this process is unfortunately again rushed through and therefore will not help that much. I could go uh on long and long >> unfortunately time is over. I would have loved to ask what gives you hope but we're going to have to answer that question while having coffee or maybe the next panel will.
Um thank you very much to the three of you for taking the time and for this interesting discussion for providing your very different perspectives on the issue and I look forward to yeah continuing on this issue. Thank you so much.