Then, suddenly, it happens. The US Federal Trade Commission (FTC) loses the independence from the executive branch on which the EU–US Data Privacy Framework depends. The legal basis underpinning much of the public sector’s use of cloud services is thereby called into question.

The legal position is unclear. What is clear, however, is that every public authority must now assess whether continued storage is compatible with the GDPR and, where necessary, migrate to European services. The question is what that would entail, and what it would cost. If the cost proves unreasonable, however, that will not be because of an American decision, but because of Swedish ones.

Imagine that the state had not taken overall responsibility for the electricity supply, but had instead left each municipality to build its own electricity grid. It sounds unthinkable, yet that is how our digital infrastructure has developed. While the state has taken overall responsibility for other areas critical to sovereignty, in practice it has left the design of the digital infrastructure to the procurement decisions of individual public authorities.

Technology has been procured as though each purchase were an isolated transaction, with cost-effectiveness taking precedence over national freedom of action. As locally rational decision-making has been allowed to govern thousands of procurement processes, the public sector is now highly dependent on a small number of suppliers, predominantly American. And, by extension, on the states under whose jurisdiction they fall. This dependence has reduced Sweden’s freedom of action and, consequently, its sovereignty.

Systems of governance create the behaviour they reward. The outcome is hardly surprising.

The vulnerability is not theoretical. US jurisdiction can restrict European users’ access to digital services, as Microsoft’s closure of accounts linked to the International Criminal Court has demonstrated. What a corresponding measure against Swedish public administration would entail hardly needs spelling out.

The state now emphasises the importance of reducing dependence, but appears to assume that the same restrained government direction that created it will now reduce it. The new cloud policy is advisory, using “should” rather than “must”. The National Agency for Public Procurement talks about digital sovereignty, but there is no requirement to take supplier dependence into account. eSam provides support for analysing cloud services, but offers limited guidance on procurement as a strategic instrument for digital sovereignty. There is little reason to believe that advice will make nationally necessary procurement decisions locally rational.

A state that cannot adapt to changing external conditions without unreasonable costs or widespread disruption does not possess digital sovereignty. If changing suppliers, migrating data or changes in legal conditions risk paralysing critical societal functions, freedom of action has already been lost.

Conditions change rapidly. Dependence comes at a price. If the state intends to reduce that dependence, procurement must begin to be used as a strategic instrument. Otherwise, Swedish taxpayers’ money will continue to finance the very dependencies the state says it wants to reduce.

If the EU–US Data Privacy Framework falls, it will be a litmus test of Sweden’s digital sovereignty. At a time when the distinction between real and digital sovereignty is, at most, lexical, it would be desirable for the state not to wait for the result before its governance begins to reflect that insight.

This article was first published in the Swedish publication Nyhetsmagasinet Fokus, under the Sticket banner, on 7 July 2026.

The post Digital sovereignty is built through procurement, not negotiation appeared first on Subversivt.

Principally right – but the wrong problem is identified

The Director of IT at the Swedish Public Employment Service has been suspended. He appeared in advertising for Capgemini, one of the agency’s IT suppliers.

That is, of course, both right and proper. In principle, at least.

Parul Sharma, Secretary-General of the Institute Against Bribery, says in an article:

“Even without financial compensation, appearing in advertising can mean that a private actor gains an undue advantage by using the civil servant’s position and the authority’s standing for marketing purposes.”

There is nothing, in principle, to object to in that statement either.

The Swedish Public Employment Service as a quality stamp? Hardly.

But in this particular case, it was the Swedish Public Employment Service that was involved.

It is therefore unreasonable to speak of using “the authority’s standing” or to suggest that the Director of IT’s participation in the advertising would have given Capgemini an “undue advantage” – at least not in the way Sharma intends.

No procurement officer or buyer saw the advert and thought: “The Public Employment Service, being so competent and discerning, must have made an excellent choice of supplier.” That simply did not happen.

Consequently, Capgemini has not attempted to use the Swedish Public Employment Service’s exceptionally weak brand to persuade procurement professionals to choose them as a supplier. The idea collapses under its own implausibility.

If the advert was not aimed at procurement officers – who was it for?

So why was advertising produced that could potentially even harm Capgemini’s brand? Marketing and undue advantage, is my qualified guess.

However, the target audience was not procurement officers in general, as Sharma suggests. The target audience was, more likely, a procurement officer in particular – namely, the Director of IT himself.

The Director of IT who is so pleased to appear in advertising that he does so without remuneration. The Director of IT who feels selected and special. Perhaps even grateful.

I do not begrudge the Director of IT the pleasure of enjoying his work or building relationships with suppliers; a certain amount of give-and-take is inevitable. But his experience and behaviour matter, because recognition, affirmation and symbolic inclusion create loyalty in people – a kind that is difficult to defend oneself against. The effect is often stronger than money, precisely because it does not feel like influence.

Advertising as a vector of status, not marketing

What we are talking about, then, is not marketing in the traditional sense. In this case, advertising functions as a vector of status – status as a carrier of influence.

The problem with this kind of influence, and the loyalty that follows, is that it can unduly benefit Capgemini. For example, because the Director of IT (like everyone else in the public sector) naturally knows how to design public procurements that favour – or even ensure – a desired supplier.

So yes, undue advantages – but of an entirely different kind from those described in the article.

Why symbolic influence is hard to see – and hard to regulate

The explicit nature of the advertisement is probably what caused it to be noticed. But since it was not a bribe in the traditional sense, it was not identified as symbolic influence on the civil servant in question, but rather as reputation exploitation – which is something else entirely.

While the system can relatively easily identify and regulate influence through explicit transactions such as money, gifts and remuneration, it has considerable difficulty dealing with the kind of influence that operates through status, recognition and social reciprocity.

This inability to identify and manage symbolic influence means that all the influence that occurred before the advert – and its potentially concrete effects – is overlooked.

Within complex organisations, it is often precisely these informal, loyalty-generating actions – visibility, affirmation, belonging – that constitute corruption. A form that is far more elusive and insidious than explicit, overt corruption.

A textbook example of corruption in complex systems

This is not an exception specific to the public sector or to this particular authority, and the decisive factor is therefore not who the Director of IT is. What matters is his role, and the incident as a textbook example of how symbolic influence and corruption function in complex systems – and, not least, how difficult they are both to resist and to regulate.

But it is clear that if we are to have any chance of addressing this, the first step must be to identify it correctly. We cannot have symbolic influence right in front of us and call it “reputation exploitation”.

The post The Most Dangerous Influence Feels Like Appreciation: Overlooked Corruption in the Public Sector appeared first on Subversivt.