Brussels decisions will do “little” and be “too late” for startups

0:00Mike Butcher:Welcome to Path Founders with me, Mike Butcher, where we like to delve into the code, the capital and the consequences inside the tech startup ecosystem. Despite talking a lot about AI sovereignty, industrial strength and digital independence, European countries have gradually realized they don't actually own much of the technical stack underpinning the businesses and societies. Two things have happened to wake it up from its slumber, geopolitics and, of course, Donald Trump. Europe keeps trying to regulate big tech and keeps trying to buy from it, just like a drug addict attempting to regulate their own drug dealer.

0:43Mike Butcher:What is to be done? Well, joining us today is competition economist Christina Cafara, who is chair of the Eurostack Industry Foundation and co-founder and vice chair of the Competition Research Policy Network at the Centre for Economic Policy Research in London, but she's generally based on a plane. Thank you so much for joining us, Christina. Thank you for having me, my pleasure. You are an absolute expert on all of these issues, but obviously you know that PathFounders likes to cover what's going on in the tech startup land and what the investors are doing. When you think about that audience and these issues, what are the kinds of things you think that we should think about in that context?

1:33The main message I would like to give is that Europe is building and Europe isn't over by any means. There is a narrative that has developed over time that Europe, of course, is lagging behind. We know very well the diagnosis. We have been admiring the problem for a very long time. But as you said in your introduction, there is a big change in the geopolitics and, you know, the other factor you mentioned, the Trump administration, which is related to the geopolitics. I think this has been a massive wake-up call for Europe.

2:14Europe asking itself why are the Americans mean to us but there is a very big shift in posture right now as of the last few months there is a sense that we got it you're telling us that we need to grow up and develop our own capabilities and we must and we will. Europe as you said has fundamentally regulated itself into the ground for the last 15-20 years as a competition economist I was very deep in that tent. And I'm not suggesting that regulation should be scrapped or we should do away with it, but we should also realize that we had a monorail for too long in which we somehow deluded ourselves that regulation was going to build us an industry.

3:00Regulation can do things if it is done properly, perhaps it can contain some of the excesses of market power of the main players, but it cannot build us an industry. Building a business requires, as we know, addressable market, funders, a business model, KPIs, and none of this is brought about automatically by regulation. The mistake Europe has made in the last 15 years has been the magical thinking that if we regulate the tech giants, then some magical thing will happen and we will create our own industry. How does regulating Google search create me a cloud? How does regulating the Apple App Store create me a software capability?

3:43It doesn't. It does at the margin benefit developers who can get a better deal when they go on to the App Store, but it isn't creating an industry. So this message is sinking in. The notion that we over-regulated and we need to do something about it, particularly we need to just move the effort somewhere else and build, is now very clear. What can you review on the next episode?

4:10Mike Butcher:Right, well, but surely you don't argue that all regulation was bad, but that it regulated the big tech players and put too much, maybe, but overburdened companies that were trying to get going in the same kind of marketplace. Yeah, but this is all in the past in the sense that we have created this regulatory machine. Europe has diluted itself through incredibly dumb narratives. We went for a while with this incredibly dumb narrative of the Brussels effect. A professor at the University of Columbia, Anu Bradford, was getting a lot of likes from the bureaucrats themselves because she was out there saying, oh you're sexy you're doing brussels effect kind of things and you are ruling the world becoming the world's prime power in regulation big deal it has delivered nothing to europe tell me how is europe winning this is a debate that i would like to have with people who are pursuers and supporters of regulation tell me how we have won we haven't won a single battle i can go religiously through every one of the cases pursued by the European Commission since 2010.

5:28And yes, we've been pioneers, they've been an abject failure. Every one of the cases that are now being run under the DMA, DSA and so on are delayed and not getting anywhere and further delayed. The point is, forget that. Forget it. It's not going to be doing very much. Want to do it? Have at it. You have an army of lawyers and economists and bureaucrats who are happy to do that all day. not interested because what needs to happen is building every layer not as a single unified stack but we need to power up europe anywhere now we lost the lm battle fine let's face it we're not going to be saved by mistral but we need to build we need to create cloud that is sovereign at least in part we need to create software applications that are our own and allow so we need to create the environment we are doing it i think that things are happening but well we are we are it's

6:29Mike Butcher:quite clearly um i suppose maybe your message is being heard and i'd love to know what you are hearing from your sources but uh you know you've got uh france who's going to switch away from Microsoft Teams, you've got Belgium and other countries switching to messaging platforms which are not WhatsApp for instance for government institutions but that's not necessarily going to move the needle is it? There are two sides to the European demand for these services. There is the public sector and there is the private sector. The public sector is 20 % of the total. A loan is not going to create a massive change and shift in demand but But we are starting from a position where uniformly the European public sector is using AWS or Microsoft.

7:15This in itself is a scandal. That European taxpayer money is being spent uniquely and solidly to Google, AWS and Microsoft is a scandal. Why am I talking to the European Commission or I've been talking on Teams? Is there no collaboration software in Europe? Of course there is. We don't just buy it. it's because the procurement has been essentially funneled into these three suppliers for reasons that we know they're good they distribute well all of that is not in any way a critique of what they've done because they had a market and they filled it it is fine but this has left European vendors and builders behind so in the first place we need to certainly change and it is happening this one side preference for American tech in public procurement things are happening you quoted France Belgium no Germany the Dutch central bank the Dutch government have just left Microsoft and gone to stock it the Swiss are leaving Microsoft at the level of of course current contracts are still being written mainly to hyperscalers but there is a shift that is material.

8:28There is going to be, I think, more. We are seeing the tsunami is no longer a line at the end of the horizon. It's coming forward. So I think we will see much more of the public sector shifting to European solutions.

8:44Mike Butcher:Okay. If that's the case, then why is it that US hyperscalers are now selling sovereign cloud products in Europe and saying that they're European? Because what would you do to protect yourself? You are selling this fiction that you're sovereign. You are saying to your customers, no, we love you. We are going to be looking after you. We are not going to abandon you. We are going to be building. We are building. We have subsidiaries in Europe. We have data which is protected, which we have encryption, and we cannot let you down. The reality that is inescapable is that the Cloud Act of 2018 in the US puts all of this into the bin, because fundamentally, every one of these companies, which is an American company, is subject to jurisdiction of the United States, and therefore the Cloud Act, even if they have a subsidiary in Europe.

9:46The reach of the administration is profound and it goes way beyond, pierces the wall of every subsidiary you can create. And this is what the Microsoft CEO had to admit in front of the parliament and the rules in France, that ultimately if the administration requested him to sort of provide access to data, he would have no choice but to do it. So, yes, of course, you know, I would do the same if I were them. I would try to say we have located your data. Data location is important. Sovereignty is about more than the location of data. Sovereignty is about agency more generally on our infrastructure and investment in that infrastructure, which is European.

10:30If you have an infrastructure that is 80 % owned by non-European, the revenue and the profit gets sucked out of Europe. There is an estimate that says about$265 billion last year in revenues for cloud and such services went to the United States instead of to Europe. So they were sucked away to American companies. Imagine if that kind of, and I don't mean 100%, but even half of that, even a quarter of that stayed in Europe to support European endeavors. That is the point. We need to invest in our own stuff, not Americans invest in our stuff and profit somewhere else.

11:16Mike Butcher:So in five years time, all European countries are running their own European cloud, but what about compute? Because all governments are moving towards artificial intelligence in order to power government services and to make them cheaper and obviously to get rid of civil servants who are too expensive. And the pitch, of course, from US big tech is run open AI on a sovereign cloud in Europe or run, you know, misdraft, not misdraft, misperplexity on a sovereign cloud. it's still going to be the same thing, isn't it? You know, AI is where the rubber hits the road because the reality is that Europe has lost the LLM battle.

12:05And as I said, Mistral isn't going to save Europe. There is far too much investment in capabilities which are being created by non-Europeans and we are way behind. So the current position in AI, which is a race we cannot afford to lose, is not that we will create our own foundational models, nor the platforms on which they will run. The place we are in is that when it comes to AI, the best that Europe can hope is to create applications on top of these models. So it is true that we are not going, you know, there's a lot of talk about sovereign AI. This is a misrepresentation. Sovereign AI is going to be the application layer, which is going to be European.

12:51is going to be using the industrial data, it's going to be using our capabilities in mechanics and machinery to run and develop industrial applications, but it's not going to be sovereign in the sense that we control the full stack and we're still going to be exposed to that. This is a problem. When you're talking, nonetheless, what are we going to do? Sit on our hands and do nothing? I think that the position now is that there is a momentum for creating much more cloud infrastructure. Remember, AI is cloud. So we are not going to be seeing a new LLM being driving the charge in Europe. Yes, we have Mistral.

13:35Good luck to them. But it's not going to be the salvation of Europe.

13:41Mike Butcher:Well, it may be the case that new technologies might appear. You have Jan LeCun doing the army platform. You have the new announcement yesterday of ineffable intelligence out of London, which is one of the key people out of DeepMind. So who knows? We might see sovereign European AI. You know, I'm hopeful. But the reality is that we have an infrastructure which is incredibly well established and integrated, which now integrates cyber. So the narrative that you see in Europe, in Germany, for example, that we're going to create massively industrial AI, we're going to invest into industrial AI and it will be sovereign, is a partial kind of rendition of the truth.

14:27It does refer to sovereign applications, which are built still in the main on foundations that are not ours. That said, I think we are not in a position to just resign ourselves to it because there is an element of resilience. I see the whole time, not just the government, but the private sector coming to me and saying, we are concerned about having all this dependency with, and this is not just, you know, what can I tell you, businesses that are exposed. I'm talking to banks, I'm talking to insurance companies, I'm talking to pension funds, I'm talking to retailers. There is a concern in the private sector that there is too much exposure to American assets, but also that we need to create the conditions for the continent profitability and the productivity growth to improve.

15:21We are in a draggy world in which the simple fact is that Europe as an economic model is not viable in 20 years unless we reduce that productivity gap with the United States. And that productivity gap is entirely and only due to the US having thrown more capital at workers in tech. That's it. You go and break it down. It's nothing else. It's not that the US workers work harder or, yes, they take a little less holidays, whatever. It's entirely to do with the investment per worker in capital investment, which is tech. So unless you have something there, we are not catching up.

16:02Mike Butcher:Well, let's talk about that private sector that you just mentioned. You mentioned banks, insurance companies, etc. But at the tech startup level of the lovables of this world based out of Sweden, OpenAI was 11 person team in 2015. That's when the big two new technologies of the future are built in these small tech companies, these tech startups. And so what do you say to them about how European countries can support them rather than waiting until they're of the size that can assist a bank in their cloud stack or whatever? So, I mean, the reason I mention bank and pension funds is because it is important to not forget that Europe has got a huge body of demand which uses services like IT.

16:51And they are the actual ones who are spending a lot of money on these services. So I don't want to just, you know, there's a lot of focus on startups. Not enough in many ways, not enough money directed at them. But I'm also concerned with the idea that all our established businesses are effectively sort of uniformly using American technology and thereby not spending real money, real money, hundreds of millions a year buying European services, which are there, are viable and are in distress because their share is shrinking. So that has been my concern with Eurostack. Startups, of course, are the lifeblood because they are the future businesses that need to be pursued.

17:41And of course, we know that in Europe we have had lower appetite for risk. We know what the limitations structurally are to the growth of startup. But unless you have also a scale-up sector and you have pumping up the demand for these scale-ups who in turn can absorb and possibly grow with the startups, you are nowhere. We are continuing to be the colony.

18:12Mike Butcher:Isn't it the case, though, that, you know, the big tech platforms in the US have leveraged the very large single market of the United States of federal system? Is it not the case, though, then, therefore, that there's an argument that technologists should argue for a federal, a federated Europe, a federal Europe, a united federal Europe with a single digital market, a single proper single market and a very, very large federal market? It isn't just that there is a case. I mean, this has been the diagnosis for years. How many reports by a former Italian prime minister, Mario Draghi, Mario Monti, do we still need to see, or Enrico Letta, do we still need to see to tell us that Europe is fragmented and we would be doing much better if we were able to deepen the single market into a single one?

19:08you know, to use that very fine expression, no shit. But the point is who's doing it? So the commission, von der Leyen, has yet another blueprint that says by 2027, we will have made progress. This is the roadmap towards the single market. Will it happen in my lifetime? I hope, but we've seen all of this before. So I'm not very hopeful that the Brussels bureaucracy can deliver. That's why Eurostack is an effort which essentially gathers the private sector because yes what I want the regulators to do and the Commission and the public sector to do is just try and buy European please. This is the one thing you can do.

19:54Try and mandate that you as a public sector will buy European. Whether you can create me a single market, you can create me a capital market union you can create that will be lovely but when is it going to happen so it isn't i'm not selling hope that this is going to happen anytime soon another issue that uh often gets

20:17Mike Butcher:slightly missed is where is the lobbying power and we've seen that united states uh companies can lobby europe and the eu and the uk of course as well very effectively um with their lobbying power And often you see startups try and lobby. They've lobbied around something called the EU Inc, which is a Delaware style single company structure for Europe. And yet they don't seem to get anywhere. Is there a solution in terms of actually sort of joining forces on the lobbying side rather than having these disparate, fragmented approaches? So EU Inc. is so far a great disappointment. I was very, very much inspired by what Andy Klinger and his movement, which was a spontaneous movement, much like Eurostack tried to achieve.

21:15and they were let down the garden path and then the commission in typical commission style produced a fudge, a fudge which reflects the fact that we are fragmented and instead of doing the thing that would have been you think in a best scenario why can't you harmonize these rules? No, we can't harmonize them. We have to create yet another sort of entity that will be special in which have one set of rules. But no, we cannot do even that. We have to create that entity, but that entity then has got to follow the law of the 27 member states. That is insanity. And it is a classic outcome of the Commission's work.

22:00You know, a fudge. Am I hopeful that it can be fixed? Not very. because this is how Brussels work. Is it a matter of lobbying? It is a very difficult situation. That's why I'm very negative about Europe's descent into infantilization. We have become a continent that somehow whose first reaction is always, how do I get Brussels to do something for me? How do I get the government to do something for me? Haven't we yet understood that the reality is that they will do little and very late and it will not be for you. So I spend a lot of my time, you said I live on a plane, I spend a lot of my time in Texas.

22:45In Texas if I have a business idea I get on my truck and go sell it. I don't think as a first instinct I wonder what DC can do for me. This is a mindset that Europe has developed and I'm sorry to say startups and founders have been taken down the garden path themselves by hopeful lawyers and lobbyists. Look at the entire predicament of Digital Markets Act. I know countless challengers to the large platforms who were taken to Brussels to see the Commission by their lawyers spending vast amount of money hoping that they would succeed in getting the commission to do something that would help them.

23:33The position now, the posture is they're not going to get me to Brussels once again to waste my time. That was a waste of time and money. Nothing has been delivered. You mentioned one success of the DMA. You mentioned one business or two who have actually grown. Maybe there is one or two at the margin. So this is the position. We should, I think, ultimately stop thinking that the solution is going to come from some wisdom in the government or Brussels. There isn't any. All they can do is produce regulation, which is actually undermining Europe's posture and ability to stand up for itself.

24:12Mike Butcher:Right. OK, well, we all know the problems, but where do we go from here? What do you think are the top three or couple of solutions that everyone, certainly listening to this podcast and watching this, should concentrate on right now? My main focus now is to direct demand, to drive demand, to make different choices. Because there are three components to all of this. Of course, you need the supply side to be offering products people want to buy. and the startup and the starters of course they are producing new ideas and they need visibility they need to find partners this is very important but even at the level of scale-ups we need products that people want to buy sometimes you say to some of these companies why do you think you're more successful and the answer is oh you know customers don't really understand my product and my reaction is no it's like saying my wife doesn't understand No, you're boring.

25:14And so the point is, if your product isn't attractive to the customer, then you have to just do serious soul searching as to why that is. There is a question of visibility. Yes, OK, we have maps with products on the map. No one is going to buy that way. That's why I think the supply side needs to do serious effort into creating products that actually customers want to buy. And customers don't want to buy a bit here, a bit there, a bit there. Even a bank, even a pension fund will tell you, but what I get from AWS is wall-to-wall service. It's an integrated service. I'm not going to buy a bit of cloud from this guy, a bit of software from this guy, a bit of AI from that guy.

25:55So unless Europeans organize, and we are doing it as Eurostack, I'm personally kicking ass in the direction of create products that people want to buy as a bundle. Because if you're a Mercedes, you're not going to just buy a bit here, a bit there, a bit there. even if you are any company you need to have something which is is viable supply side demand side put your money where your mouth is or you know the point is you can't go around whining that you're worried about resilience and then continue to buy from AWS and Microsoft you you need to just explore the alternatives work with the supply side and create them

26:32Mike Butcher:so maybe you see a little more M &A in Europe where people companies try and create more end-to-end solutions uh it it sounds it sounds like a difficult place for more much smaller startups even if they're relatively well funded to play in and the software market is uh so much vaster internationally um do you do you see that um having any kind of uh play in the factors these factors at all i think i think what i see the third factor i was going to mention is funding What I can see is a major, major sentiment shift in funders. Funders are now coming to me saying, family offices, we want to create, I don't know, a Eurostack catalyst fund which invests in European tech.

27:18And there are other initiatives that are focused on Europe. I know some VC. We're not talking about necessarily VC class. VC is one type of asset class. I'm talking about an asset class that is somewhat intermediate between VC and BE, but something which is slightly more patient than VC, and there is appetite for that. Even the pension funds, even the institutional funders are saying we can see an investable window in European tech. So I think the sentiment is changing. Of course, if you compare it to the hundreds of billions that Americans are spending even for things like Stargate and so on, it's unfathomable.

27:57We don't have anything comparable. But you know, the answer cannot be let's all lobby the European Commission. We are in a different mindset. The European Commission is not going to design the future of the European digital industry ever. They're very nice people, but they are bureaucrats and lawyers how do we expect them to have anything meaningful to say on the future of the of the European industry ultimately the future of European industry is going to be decided by tech by industry and by funders that's it there isn't anything that we can hope coming from by magic the European Commission getting smarter all of a sudden and becoming an entrepreneur Why are we not seeing this?

28:41Forget all that side of things. We need to just absolutely be clear that Europe is a superpower because we have a size which actually is bigger than the US as a market. It is fragmented. But don't keep telling me that as an excuse because this idea that we are fragmented, therefore we cannot do business. Look, I was working at the time for Amazon when they came into Europe in 2005 and they created the marketplace. And the instruction from Seattle was go and create a headquarter in Germany and roll out UK, France, Italy, Spain. And they did. And there was no obstacle. They faced it. They did it.

29:20So, of course, they had bags of cash. But the point is this whole kind of floppy sort of attitude. We can't do it. How do we? I had somebody telling me, how do you overcome the linguistic fragmentation of Europe? Are you serious? This is a feature, not a bug. And with AI, it is irrelevant anyway. What do you want me to do? So, I mean, some of it is real learned helplessness. Get off your ass and do stuff.