Sovereignty Decided – While Dependency Continues
clarus.news | Analysis | June 11, 2026
by Andreas Binggeli, Thierry Leserf, Ernst Anker, with Claude Opus | June 11, 2026 short version, facts only
On June 9, 2026, the Council of States adopted Motion Z'graggen for an "Impulse Program to Strengthen Digital Sovereignty" by 30 to 7 votes – against the express wishes of the Federal Council. This is a political success for the Parliamentary Group Digital Sustainability, Parldigi. However, what was decided for now is a mandate, not independence. The Swiss Government Cloud, the Army's New Digitalization Platform, and health data programs remain dependent on foreign technologies and providers. The decisive question is therefore no longer whether Switzerland wants to become more sovereign, but who will implement this by when.
The Council of States Contradicts the Federal Council's Reassuring Response
With 30 to 7 votes and one abstention, the Council of States adopted Motion 26.3221 by Center Party Councilor of States Heidi Z'graggen on June 9, 2026. It demands an impulse program through which the state, economy, and science jointly implement pilot projects in the areas of digital infrastructure, open source, cybersecurity, and artificial intelligence.
Remarkable is not only the clear result, but also the resistance against which it was achieved. Federal President Guy Parmelin recommended rejection. Numerous works were already underway, existing funding instruments were sufficient, and the budget situation was tight.
This is the Federal Council's standard response to strategic gaps: it declares ongoing activities as strategy and the missing budget as virtue. The Council of States was not convinced by this approach this time.
The Parliamentary Group Digital Sustainability, Parldigi played a significant role in the decision. Its core team – Min Li Marti, Franz Grüter, Gerhard Andrey, Dominik Blunschy, and Nik Gugger – had provided Council of States members with precise arguments. It doesn't happen daily that representatives from SP to SVP unanimously advocate for the same cause. In digital sovereignty, political reality seems to be faster than party boundaries for once.
Parldigi has been pursuing this topic for 17 years. The decision is therefore not a spontaneous mood-based success, but the result of long and often unspectacular foundational work. The group described the vote as a courageous decision and as a mandate to the administration to finally move forward. This is precisely where the more difficult part now begins.
Sovereignty Is Not a Product in the Shopping Cart
A day after the vote, Martin Andenmatten provided the appropriate assessment in Netzwoche. His central statement reads: "There is no sovereign technology. There are only sovereign decisions."
This shifts the discussion to where it belongs: to leadership. Digital sovereignty does not mean developing every software oneself or excluding all foreign providers. It means knowing dependencies, maintaining alternatives, and being able to actually decide in crisis situations.
Sovereignty is therefore not a quality seal that one sticks on a cloud. It manifests in contracts, exit options, open standards, exchangeable components, own competencies, and clear responsibilities.
This is exactly what the federal government must be measured against. Not against strategy papers, but against the architecture of those projects it builds today and finances for years.
Swiss Government Cloud: Sovereignty in Name, Hyperscalers in the Engine Room
The Swiss Government Cloud (SGC) is supposed to become the cloud foundation of the federal administration between 2025 and 2032. It is often presented politically as an answer to growing requirements for security, availability, and sovereignty.
However, previous planning paints a different picture. According to estimates based on the project message, approximately 68 percent of total SGC usage should run through foreign hyperscalers by 2032. Around 22 percent would be allocated to federal data centers, another 10 percent to Public Cloud on Premise.
This would make foreign public cloud not a targeted supplement, but the dominant part of the architecture. Sovereignty sits in the passenger seat, so to speak, while hyperscalers hold the steering wheel.
The finding is exacerbated by procurement practices. In September 2025, the Federal Chancellery extended public cloud framework contracts without tender for five years. AWS, IBM, Microsoft, and Oracle, as well as Chinese provider Alibaba, were beneficiaries. The justification: without extension, a supply gap threatened before the SGC would be ready.
Western Swiss provider Infomaniak filed a complaint with the Federal Administrative Court but withdrew it in December 2025. Regardless of this procedure's outcome, the political question remains: Why is a dependency planned for years being extended while simultaneously demanding an impulse program for digital sovereignty?
The federal government's private cloud is also not free from strategic risks. It relies substantially on technology from Broadcom/VMware. Thus, a central layer also lies in the hands of a US provider there.
According to current status, the SGC does not untangle existing dependencies. It organizes and expands them. The Council of States' decision binds the project neither to a lower hyperscaler quota nor to binding exit scenarios.
NDP: Protected in the Bunker, Dependent on the Licensing Model
The contradiction becomes even clearer with the Army's New Digitalization Platform, NDP. It is supposed to go into operation in protected data centers on July 1, 2026 – just a few weeks after the Council of States' sovereignty decision.
The platform is based on virtualized industry standard products. Virtualization technology from Broadcom/VMware plays a central role. The financially effective expenditure amounts to 477 million francs, with expansion reaching into the 2030s.
The NDP is supposed to be protected against physical attacks, cyberattacks, and power outages and be practically operationally autonomous. This is militarily understandable. Technologically, however, a peculiar construction emerges: the infrastructure is in the protected data center, but strategic dependency travels in with it.
After acquiring VMware in 2023, Broadcom restructured licensing models, raised prices, and cut partner programs. Such changes show how quickly a technical standard solution can become an economic cluster risk.
A platform is not sovereign just because nobody can physically access the server. If an external corporation can unilaterally change licensing conditions, costs, or product strategies, autonomy remains conditional. The bunker apparently protects against much – just not against the next price list.
Particularly explosive is that Army Chief Thomas Süssli himself warned against US Cloud Act risks in September 2025. This warning gave additional momentum to the political sovereignty debate. All the more questionable is why the army's central platform is built on a US virtualization layer.
The Federal Audit Office has also critically assessed the NDP. It noted that financial resources were not sufficiently available and overall governance was lacking. Additionally, numerous other projects depend on the Swiss Command Network and the new platform.
The Council of States decision changes nothing about this for now. The NDP goes into operation as planned. A binding roadmap for replacing or diversifying the central virtualization layer is not publicly visible.
Digisanté and EPD: Here the Federal Government Can Still Decide in Time
The sovereignty question is particularly sensitive regarding health data. The electronic patient record, EPD, is supposed to be developed into an electronic health record by 2030. Simultaneously, the Swiss Health Data Space, SwissHDS is being created in the Digisanté program.
This is not about ordinary administrative data, but about medical information, diagnoses, treatments, and personal health histories. In hardly any data category do loss of control, unclear responsibilities, and foreign legal access weigh more heavily.
The starting situation is sobering. By the end of April 2026, only around 136,000 EPDs had been opened. Adoption remained significantly behind expectations. Development continues through the Post, with the new system supposed to go into operation in 2030 at the earliest.
The central architectural questions are therefore not yet conclusively decided: Where will the SwissHDS be operated? Which components must be controlled in Switzerland? Which providers are considered? Which exit options are required? And who is politically liable if sovereignty goals later fail due to costs or schedule pressure?
This is precisely where the impulse program could have an impact. Not all technical and contractual facts have been established yet.
However, the motion does not legally bind Digisanté. It creates political pressure, but no concrete hosting requirement, no budget, and no binding timeline. At the same time, according to parliamentarians, Digisanté is among those programs affected by austerity measures.
The opportunity is therefore real, but not automatic. Without clear responsibility, sovereignty could also be examined here until the system is procured and any fundamental change is allegedly too expensive.
Braking at the Top, Building at the Bottom
Is the federal government leading by example? At the political top, hardly so far. At the operational level, however, concrete alternatives are already emerging.
The SDS Network – Sovereign Digital Switzerland and the SDS Center emerging from it show how an abstract concept can become an actionable structure. The network was initiated by Prof. Dr. Matthias Stürmer, head of the Institute Public Sector Transformation at Bern University of Applied Sciences.
What began as a loose network in summer 2025 now comprises over 200 authorities and companies. The center counts 31 paying member organizations. Cooperations exist with German ZenDiS, and an openDesk offering is being built for Switzerland. Participants include the FDJP, the City of Zurich, the Post, Cyber Command, and the cantons of Basel-Stadt, Bern, and Solothurn.
This work shows that digital sovereignty does not fail due to lack of technology. More often, a mandate, sustainable organization, and reliable financing are missing.
Other countries also provide examples. Germany is strengthening central open-source competencies with ZenDiS. Schleswig-Holstein is replacing a large part of its Microsoft Office installations. France is developing the state platform "La Suite" with DINUM. Austria's military is switching to LibreOffice.
The models differ, but the common denominator is clear: a central entity receives mandate, budget, and enforcement capability. This combination is precisely what Switzerland still lacks.
Three Questions That Cannot Be Answered with a Strategy Paper
1. Regarding the Swiss Government Cloud:
How does the Federal Council's statement that it is already doing much for digital sovereignty reconcile with SGC planning whereby around 68 percent of cloud usage should run through foreign hyperscalers by 2032? Why were framework contracts with AWS, IBM, Microsoft, Oracle, and Alibaba extended without tender for five years in September 2025, instead of setting binding reduction and exit targets?
2. Regarding the Army's New Digitalization Platform:
Why is the army building its platform, protected against cyberattacks, power outages, and physical attacks, on Broadcom/VMware virtualization technology? What technically proven and contractually secured exit option exists if prices, licensing models, or delivery conditions are fundamentally changed again?
3. Regarding Digisanté and EPD:
Who bears responsibility after the Council of States' decision for ensuring that the future electronic health record and Swiss Health Data Space are operated sovereignly? By when will hosting model, minimum requirements, budget, and exit scenarios be bindingly determined?
Conclusion: The Mandate Is Decided, the Dependency Far from Over
June 9, 2026, was a good day for political debate about digital sovereignty. Parldigi convinced across party lines. The Council of States corrected the Federal Council. Martin Andenmatten precisely named the leadership question. Matthias Stürmer and the SDS Center show that practical alternatives can be built.
But a motion decision is not yet implementation. It doesn't cancel framework contracts, replace virtualization layers, or decide hosting architecture.
As long as the Swiss Government Cloud primarily relies on foreign hyperscalers, the NDP is built on US technology, and the health data hosting question remains open, digital sovereignty is primarily a political promise.
Now binding roadmaps, measurable goals, responsible entities, financed alternatives, and technically proven exit plans are needed. Not sometime after the next procurement, but before today's dependencies are cemented for another ten years.
The Council of States has pressed the start button. Now the federal government must decide where the journey goes – and who continues to control the navigation system.
This contribution is based on Parldigi's open letter to Council of States members from June 1, 2026, inside-it.ch's reporting on the Council of States vote from June 9, 2026, Martin Andenmatten's opinion piece in Netzwoche from June 10, 2026, publicly accessible federal information on the Swiss Government Cloud and NDP, as well as sources on Digisanté, EPD, and the SDS Network/Center.
Sources:
- Parldigi: "Open Letter Regarding Motion Z'graggen," June 1/2, 2026
- inside-it.ch: "Council of States Wants to Strengthen Digital Sovereignty," June 10, 2026
- Netzwoche / Martin Andenmatten: "Digital Sovereignty Needs Leadership," June 10, 2026
- Motion 26.3221 "Impulse Program to Strengthen Switzerland's Digital Sovereignty"
- BIT: Swiss Government Cloud
- DDPS: NDP Program; FAO Audit Reports 23155 and 24127
- FOPH / eHealth Suisse: EPD, E-GD, Digisanté and SwissHDS
- BFH / SDS Network and Center
clarus.news | Andreas Binggeli, Thierry Leserf, Ernst Anker, with Claude Opus | June 11, 2026
Tags: #DigitalSovereignty #OpenSource #SwissGovernmentCloud #NDP #Digisanté #EPD #Parldigi #Zgraggen #Stürmer #Broadcom #Hyperscaler #FederalCouncil #FAO