The illu­sion of digi­tal sovereignty

Can Switzerland have digital sovereignty? Min Li Marti and Franz Grüter, Co-Presidents of the Parliamentarian Group for Digital Sustainability, discuss dependencies, sovereignty and regulation.

The Philanthropist: Given the domi­nance of tech compa­nies such as Micro­soft and Alpha­bet, the ques­tion arises: what does digi­tal sove­reig­nty mean?

Min Li Marti: What’s inte­res­t­ing is that ever­yone under­stands sove­reig­nty differ­ently. Ever­yone adapts the defi­ni­tion to fit their own point of view. Some say you only have sove­reig­nty if you have parti­cu­larly good loca­tion condi­ti­ons for entre­pre­neurs; others say you need open-source programs or an indus­trial policy. The ques­tion of depen­dency has been on the agenda of the Parlia­men­ta­rian Group for Digi­tal Sustaina­bi­lity for some time. Howe­ver, the current geopo­li­ti­cal situa­tion has made this ques­tion very much more topical. 

Franz Grüter: In fact, the term ‘sove­reig­nty’ descri­bes an ideal state of being comple­tely self-suffi­ci­ent, comple­tely inde­pen­dent; having ever­y­thing under control as a nation. This means that no one would be able to exert pres­sure or turn off our digi­tal tools. Howe­ver, the ques­tion of depen­dency goes beyond these large inter­na­tio­nal corpo­ra­ti­ons. In the case of chip tech­no­logy, for exam­ple, there are major depen­den­cies. The whole world is curr­ently depen­dent on Taiwan, even though facto­ries are being built in Japan, the US and Europe, which will be able to reduce this depen­dency. But it will take five to eight years. We also have depen­den­cies when it comes to soft­ware, such as cloud solu­ti­ons. And there’s the physi­cal infra­struc­ture, the data centres. I work in that field myself, as my profes­sion. Switz­er­land plays a very domi­nant role in that.

TP: Does that mean that Switz­er­land is inde­pen­dent in that respect?

FG: There is an awful lot of infra­struc­ture in Switz­er­land. But when it comes to the actual hard­ware we use, we are comple­tely depen­dent. None of the compon­ents we need are manu­fac­tu­red in Switz­er­land or Europe. In short, we have very broad areas in which sove­reig­nty is desi­ra­ble. Howe­ver, it is simply unrea­li­stic to think that we will achieve this in the next few years, or even decades.

MLM: I see it the same way. It’s a shame. The fact is, we have a kind of globa­li­sed divi­sion of labour. This is not just about tech­no­logy. It affects a rela­tively large number of ever­y­day goods for which we are not at all self-suffi­ci­ent – and haven’t been for a long time. And perhaps complete self-suffi­ci­ency is not neces­s­a­rily desi­ra­ble. It may have always been an illusion.

TP: And that’s good?

MLM: What we need to think about is that there are things for which we should have a certain degree of inde­pen­dence for reasons of natio­nal secu­rity or because they are parti­cu­larly worthy of protec­tion. This includes our own cloud infra­struc­ture, which is now to be built with the Swiss Govern­ment Cloud. Some­ti­mes it’s enough not to be depen­dent on just one provi­der. This problem does not only affect the fede­ral admi­nis­tra­tion. It also applies to a large part of the economy. The high pene­tra­tion of Micro­soft solu­ti­ons is one exam­ple. I take a very criti­cal view of this. I wonder whether it is healthy and whether it promo­tes innovation.

TP: Is total sove­reig­nty at all achievable?

FG: If we wanted to achieve that, it would proba­bly take deca­des. If we want to become more inde­pen­dent and more sove­reign, Europe should also have tech compa­nies that can compete on a par with the US and China and are inno­va­tive. I don’t see that happe­ning in the fore­seeable future. That’s why we need to find a middle ground.

TP: What would that look like?

FG: The fede­ral govern­ment has deci­ded to triage all its data. In doing so, it deter­mi­nes which data is really sensi­tive, which it then does not trans­fer to a public cloud. Instead, it opera­tes its own data centre and physi­cal servers. Of course, these are purcha­sed from abroad. Nevert­hel­ess, this gets us closer to a certain degree of inde­pen­dence in terms of infra­struc­ture. That is a compro­mise that we can proba­bly achieve.

TP: But would abso­lute sove­reig­nty be desi­ra­ble in any case?

FG: After a large-scale hacker attack, Esto­nia has built a second natio­nal digi­tal infra­struc­ture as a redun­dant set-up outside the coun­try. That makes them a little less vulnerable. Such redun­dan­cies are proba­bly needed at diffe­rent loca­ti­ons and with seve­ral suppliers. 

MLM: Complete sove­reig­nty is neither reali­stic in any way nor desi­ra­ble. Cross-border coope­ra­tion is requi­red in many areas. It proba­bly makes sense to try to shift the risk. We shouldn’t focus on just one provi­der or one stra­tegy. In this respect, the current situa­tion is certainly posi­tive. It is a wake-up call for Europe to act now. Because in all these discus­sions, one thing that is often unde­re­sti­ma­ted is the fact that states usually have an active role in the inno­va­tion of tech­no­logy. Both the US and China have very active indus­trial poli­cies. The tech­no­lo­gi­cal domi­nance of these count­ries has to do with poli­ti­cal choice and not with mere dere­gu­la­tion. A lot of money has been inves­ted there. Europe has not done this. Like Switz­er­land, Europe has no indus­trial policy as a matter of prin­ci­ple. I’m not saying we have to do this. Nevert­hel­ess, if we consider the depen­dence of space travel on private compa­nies, this is funda­men­tally proble­ma­tic. We can’t be indif­fe­rent if Elon Musk wants to shut down or relo­cate Star­link. A sole proprie­tor­ship has a tremen­dous amount of power here. This was allo­wed simply because people wanted to save a little money. Or they didn’t consider it a prio­rity. And this is where the wrong poli­ti­cal decis­i­ons made in Europe and Switz­er­land take their toll.

FG: A good exam­ple of how to become more inde­pen­dent is Airbus. Europe did not want to be depen­dent on Boeing. At that time, it was deci­ded in Europe that we would set up our own aircraft manu­fac­tu­rer. Today, Airbus is inter­na­tio­nally compe­ti­tive. Of course, it would also be desi­ra­ble in the space indus­try to have a second or even third provi­der in addi­tion to Star­link. Why is inno­va­tion in Europe not succee­ding at this level and at this speed? There is a lack of venture capi­tal. Billi­ons of dollars of invest­ment are usually needed. The climate in the US is simply different.

In short, we have very broad areas in which sove­reig­nty is desirable.

Franz Grüter

MLM: Setting some­thing up takes a lot of time and money. Some­ti­mes it’s about conve­ni­ence versus secu­rity and grea­ter inde­pen­dence. Given the choice between an open-source solu­tion and Micro­soft 365, you’ll quickly choose the Micro­soft product, because ever­yone knows it. 

FG: Today, when I ask admi­nis­tra­tors or IT mana­gers in large compa­nies, they say: tech compa­nies offer a product port­fo­lio with a high level of inno­va­tion and with cyber­se­cu­rity in which they invest billi­ons. An open-source solu­tion can hardly keep up with this. That’s why I think it’s unrea­li­stic nowa­days for a fede­ral admi­nis­tra­tion to exclude all these tech companies.

MLM: But haven’t we missed the moment when open-source solu­ti­ons could have been deve­lo­ped further? Let us take the exam­ple of Airbus. Simi­larly, they were not on an equal footing with Boeing from the start. A lot was inves­ted in it at the start. With open-source soft­ware, this step has never been taken.

FG: The aircraft indus­try centres on a machine, on physi­cal hard­ware. There’s a big diffe­rence between that and soft­ware. This is demons­tra­ted by GAIA‑X, the Euro­pean cloud project. Although its deve­lo­p­ment has taken years, it is not nearly on a level that could be compared with Ameri­can tech provi­ders. The state is inca­pa­ble of being an entre­pre­neur and produ­cing inno­va­tion itself.

MLM: The history of tech­no­logy is full of failed inven­ti­ons. Even Facebook’s meta­verse has been a total flop. Only the state cannot fail so easily. We would not allow the state to fail; we would let it try again. That’s part of the problem.

FG: That’s right. I think it proba­bly has some­thing to do with our culture. I’ve been in the IT indus­try for 30 years. I have seen many projects in which money has been lost. But time and again there were indi­vi­dual success stories. I wouldn’t say it’s impos­si­ble to start again here. But you’re stig­ma­tised rela­tively quickly.

MLM: Size also plays a role. I used to work in the film indus­try. They have a formula: in Holly­wood, out of 100 scripts, 99 are usel­ess, but one of them might be a hit. In Switz­er­land, you don’t even have 100 writers writing screenplays.

FG: That’s a good compa­ri­son. In the IT indus­try, Ameri­cans have a gene­ral rule: I invest in 10 projects. Three report a total loss, four are more or less okay, and two or three explode. These are used to finance my losses. And over­all I’ve crea­ted added value.

MLM: That’s what I mean. We’re too small. This is crucial when it comes to self-suffi­ci­ency. Just because of our size, we are depen­dent on others in certain ways.

TP: But wouldn’t foun­da­ti­ons be perfectly placed to enable at least smal­ler projects to go ahead with grea­ter risk?

MLM: I’ve also tried to get funding from foun­da­ti­ons for various projects. Lots of foun­da­ti­ons have a very narrow purpose. At most, they only support one project that is clearly defi­ned. Theo­re­ti­cally, there might be that willing­ness to take risks, and there are certainly examp­les of that. But I haven’t noti­ced a great willing­ness to take risks when it comes to inves­t­ing in innovations.

TP: So has Switz­er­land achie­ved anything yet?

FG: Over the past 10 years, Switz­er­land has earned an excel­lent repu­ta­tion for data hosting. This deve­lo­p­ment began in 2010. The reve­la­ti­ons of Edward Snow­den in the USA in 2013 gave the issue another huge boost. Compa­nies and govern­ments have become incre­asingly keen to move the physi­cal hosting of data to a secure and stable loca­tion. Today, Switz­er­land is one of the most important loca­ti­ons in Europe for data storage. We have the highest density of data centres per capita in Europe. This makes it all the more pain­ful for me that the Inter­na­tio­nal Commit­tee of the Red Cross (ICRC) has left Switz­er­land in terms of IT. The ICRC works with highly sensi­tive data in which other states have a great inte­rest. It has trans­fer­red the storage of this data to Luxem­bourg. With the E‑Embassy, Luxem­bourg has crea­ted a legal frame­work, an inter­na­tio­nal sove­reign zone. No state formally has access to data stored there.

TP: Can regu­la­tion streng­then sovereignty?

FG: In my view, Switz­er­land has so far opted for a really prag­ma­tic approach. We iden­tify parti­cu­larly sensi­tive data. We store that data in natio­nally owned data centres. What I would totally reject are any bans on indi­vi­dual manufacturers.

MLM: I think we need to be more open to alter­na­ti­ves than we are today. We should try to reduce our depen­dence on these big tech compa­nies. We should not simply demo­nise them all. But a certain amount of effort would be sensi­ble. Perhaps we have done too little here. We should also make tenders more open. To say cate­go­ri­cally that we will never need Micro­soft products again is proba­bly unrea­li­stic. But I think it makes sense to be more open to alter­na­ti­ves. We’re also still waiting for the Fede­ral Council’s report on Heidi Z’graggen’s postu­late on digi­tal sovereignty.

FG: We need regu­la­tion, but above all inno­va­tion. This issue is compa­ra­ble to the issue of free­dom and secu­rity. It is precis­ely here that demo­cra­cies face great chal­lenges. They must ensure that digi­ta­li­sa­tion – and this includes the issue of sove­reig­nty – does not incre­asingly rest­rict the free­dom of indi­vi­dual citi­zens. Topics over­lap with regard to the important ques­tion: who actually has access to the data?

MLM: The ques­tion of privacy is indeed important. If we regu­late social media, for exam­ple for the protec­tion of minors, this may rest­rict free­dom of expres­sion, but the rest­ric­tion may make sense for the protec­tion of minors. This means that we also have to weigh up the pros and cons. But I do not believe that regu­la­tion per se makes inno­va­tion impos­si­ble. Stan­dar­di­s­a­tion in parti­cu­lar can even be helpful. 

FG: I’m not arguing for no regu­la­tion either. But, for exam­ple, the EU’s AI (Arti­fi­cial Intel­li­gence) Regu­la­tion can lead to provi­ders of AI programs no longer coming to Europe because deal­ing with the regu­la­tion is too burden­some for them. I do not want us to stifle inno­va­tion with exces­sive regulation. 

Shif­ting respon­si­bi­lity onto civil society is too easy.

Min Li Marti

TP: And what role should civil society play?

MLM: It is a ques­tion of what the task of poli­tics and the state really is. How must it provide frame­work condi­ti­ons that enable alter­na­ti­ves? Shif­ting respon­si­bi­lity onto civil society is too easy. These are deba­tes that are only being held in certain circles, and very inten­si­vely. Howe­ver, there is still a great deal of unease amongst the gene­ral public, such as when it comes to protec­ting privacy. Civil society groups play an important role in enab­ling society at large to lead these discus­sions. Many people feel that they really don’t under­stand enough. They feel a little lost. Civil society has a very important role to play in this trans­la­tion and aware­ness-raising work, and perhaps also in lobby­ing amongst us parliamentarians.