The distinction matters for Windows users, IT administrators, and policymakers. The new environment is intended to run in parallel with Microsoft 365, not replace it immediately. Federal officials have left open the possibility of a later, more complete sovereign workplace, but that remains a question for the program to examine and decide rather than an approved migration outcome.
What Switzerland has actually approved
The Federal Chancellery announced the follow-on program on September 2, 2026, after its proof-of-concept work under the BOSS project. The target is availability of a sovereign workplace solution for around 3,000 selected users by the end of 2027. The Chancellery estimates CHF 9 million for this first phase.
That is a platform-building and operational-testing commitment, not a final decision to displace Microsoft 365 throughout the federal administration. The official design is explicit: the sovereign workplace should be an independent solution operating alongside Microsoft 365.
This wording rules out several easy but misleading interpretations:
- It is not a confirmed removal of Microsoft software from 3,000 physical PCs.
- It is not a commitment to migrate all federal workers.
- It is not a declaration that the 3,000-user environment will become the default federal desktop in 2027.
- It is not yet evidence that the program will lower total IT costs.
The end-of-2027 date is an availability target for selected users, not a confirmed deadline for a broad migration. The Chancellery has also said that any large-scale introduction and the regular costs of operating such a service still need assessment.
A fallback strategy, not simply an office-suite swap
The clearest public explanation of the policy objective is dependency reduction. The federal digital-transformation leadership has characterized the primary goal as creating a genuine fallback option: an alternative that can be used if needed, rather than necessarily replacing Microsoft everywhere.
That makes the project more consequential than a limited software trial, even though its initial scale is modest. A workplace platform that can support ordinary work outside the incumbent environment could give an administration more room to respond to technical disruption, supplier constraints, or policy concerns. It could also make future procurement negotiations less one-sided if it becomes a credible operational alternative.
But a fallback has to work under real conditions to be valuable. It needs more than the ability to open documents. It must support identity and access management, communications, calendars, file handling, collaboration, governance, user support, and continuity during an actual disruption. Running a parallel environment can therefore be a rational resilience investment, but it can also introduce complexity: separate training needs, document-compatibility questions, integration work, duplicated support processes, and new security responsibilities.
The federal program should consequently be judged less by the headline number of users than by whether those users can perform their normal work reliably and whether the alternative can be operated when it is most needed.
What the BOSS proof of concept found
The new phase builds on the BOSS feasibility study, which used openDesk-based software and involved 172 participants from several departments. The testing scope was broad enough to resemble a modern workplace stack rather than a word processor replacement. It included document editing, email, calendars, contacts and tasks, file storage, project and knowledge management, chat, audio and video conferencing, plus identity and access management.
The Federal Chancellery reported positive assessments for key day-to-day functions, including document work, file storage, collaboration, email, and calendars. That is an important result: it suggests the tested environment could support a meaningful set of standard administrative workflows.
At the same time, the findings do not amount to a blanket production-readiness verdict. Large video conferences exposed limitations. This is particularly important because video meetings are one of the areas where organizations often discover the difference between a technically available feature and a service that performs predictably at organizational scale.
The test result therefore supports two conclusions at once. First, a non-Microsoft workplace environment appears viable for a defined range of federal work. Second, substantial implementation, scaling, and operational validation remain necessary before anyone can credibly claim an equivalent substitute for Microsoft 365 across an entire administration.
The Microsoft 365 installed-base context
The program is taking shape after the federal administration completed its Microsoft 365 rollout to approximately 54,000 workplaces in mid-December 2025, according to independent trade reporting. That context underscores why the “replacement” framing is inaccurate.
A 3,000-user parallel deployment is a comparatively bounded program beside that installed base. It may be strategically important without being numerically transformative. Nor should the 3,000-user figure be converted into a device count or a precise percentage of the federal workforce: the official announcement describes users and employees, while the available reporting on Microsoft 365 refers to workplaces. Those are not necessarily like-for-like measures.
For Windows readers, the most practical point is that workplace-software policy and operating-system policy are not the same thing. The announcement concerns a sovereign workplace environment with office functions. It does not establish that 3,000 Windows installations, 3,000 PCs, or the Windows operating system itself will be removed. The available material also does not say that every selected person will permanently abandon Microsoft 365.
That is not a semantic quibble. Device management, security tooling, local application needs, browser compatibility, and identity infrastructure can all shape the experience of using an alternative workplace platform. A parallel service can alter how documents, meetings, mail, and collaboration are handled without constituting an immediate desktop-OS migration.
Open source policy is relevant—but not a migration mandate
Switzerland’s legal framework gives the project a broader policy backdrop. The Federal Act on the Use of Electronic Means to Carry Out Official Tasks, known as EMOTA, entered into force on January 1, 2024. It requires federal authorities to publish the source code of software they develop or commission.
The rule has material exceptions. Code need not be published where third-party rights or security-related reasons prevent or restrict publication.
This law is relevant to public-sector software governance, but it should not be overstated. It does not require the government to replace commercial workplace products. It does not make every government IT component open source. And it does not itself mandate a Microsoft 365 exit.
Its direct scope is software developed or commissioned by federal authorities. The sovereign-workplace program is better understood as a separate operational and strategic initiative that exists within a government taking source-code publication and digital autonomy seriously, not as an automatic legal consequence of EMOTA.
The Cyber Command reports need caution
Separate reporting has linked openDesk to Switzerland’s Cyber Command, but the coverage is not consistent enough to support sweeping conclusions.
One report said that employees with particularly protected identities were expected to receive openDesk workstations by October 2026. A later syndicated account described a broader plan in which all Cyber Command employees would receive openDesk by that date. Because the latter account relies on attributed reporting rather than a located primary Cyber Command announcement, the scope remains unresolved.
The prudent conclusion is that Cyber Command may be pursuing a more focused or accelerated use of openDesk than the 3,000-user federal program. It is not yet safe to state as established fact that every Cyber Command employee is leaving Microsoft Office or Microsoft 365 by October 2026.
This distinction also prevents a specialized security-related deployment from being used as proof that the whole federal administration is about to follow the same timetable.
What success would look like by 2027
The first phase will be more informative than dramatic. Its value lies in testing whether a separate workplace can be made dependable for thousands of people while Microsoft 365 remains in place.
Practical measures of success would include whether staff can complete routine work without unacceptable friction; whether document, mail, calendar, file, and collaboration workflows hold up in normal operations; whether the video-conferencing limitations can be resolved; and whether the organization can support, secure, and govern the service at a predictable cost.
A successful result would strengthen the case for a durable fallback and give the federal government more evidence for the later decision on whether to expand the model. A difficult rollout would not necessarily invalidate the sovereignty goal, but it would demonstrate why office-suite features alone are not enough to replace an established productivity platform.
For now, Switzerland’s move is best described as a carefully scoped resilience experiment with real funding and a clear 2027 milestone. It is a notable challenge to the assumption that a government must depend on one workplace ecosystem. It is not, however, an announced Microsoft 365 replacement for 3,000 computers—or for Switzerland’s federal administration as a whole.