Speech by Paul de Bijl: towards digital autonomy in the cloud, one step at a time

On Thursday, April 9, 2026, the Dutch platform for the information society ECP organized a discussion meeting on ‘Taking back control in the cloud: from debate to decision-making power’ (the Dutch title was “Regie terugpakken in de cloud: van debat naar beslissingskracht”). Paul de Bijl, ACM’s Chief Economist, was one of the speakers. Below is an English translation of his speech. Please note that the original speech was given in Dutch, and that the Dutch spoken version is authentic in case of any discrepancies.

Speech by Paul de Bijl: towards digital autonomy in the cloud, one step at a time

The topic on the agenda is not so much an IT problem, although it is about IT of course. From a geopolitical perspective, it’s a power issue, and from an economic perspective, it’s a coordination problem.

Those who control access to the infrastructure on which our economy runs, for example communications, data storage, AI tools, and payment systems, hold power. And most of that power lies with actors outside of Europe. Competition and markets aren’t enough to solve this. Or rather, a lack of checks and balances on the formation of power on free markets is the root of the problem. A coordinated approach, coordinated by the government, is needed to overcome this.

The situation: the cloud market as an oligopoly

First, here are some facts.

• Three US hyperscalers (AWS, Microsoft Azure, and Google Cloud) account for 70-85 percent of the cloud market in Europe, with a one-stop shop for storage, monitoring, network services, automation, and security.
• The largest European competitors (Deutsche Telekom and SAP) have approximately two percent. European providers predominantly offer specific services.

Economists refer to such a market as an oligopoly, a market structure in which a small number of businesses hold substantial market power. Moreover, this is no ordinary market, as it involves critical infrastructure. In other words, public interests are at stake.

This dependence manifests itself not only in economic terms, but also in political and strategic terms:

• The CLOUD Act: US providers can be legally required to share data of European businesses and governments with US authorities, even if that violates European rules.
• US sanction laws, from which coercive measures can follow against countries, organizations, and individuals that threaten US foreign policy or national security, which can, for example, lead to supply bans targeting specific customers in certain countries. The threat posed by this puts pressure on national sovereignty.

Despite such risks, switching barriers are so high that buyers rarely change providers. Switching to another provider not only requires transferring data, but often also involves untangling business processes and existing IT infrastructures.

The conclusion is clear: market power also manifests itself in geopolitical terms. Markets do not solve this, and competition enforcement is inadequate. Active government coordination is needed.

Why aren’t we moving?

Then the question arises: if everyone is seeing this, why are we moving so slowly?

For years, we’ve been hearing that digital sovereignty is important. Yet if a CIO has to make a choice, that choice quickly falls on a hyperscaler. A hyperscaler offers the best functionality as part of a total package.

New buyers are often offered a competitive price. Yet once they are tied to a hyperscaler, and contracts need to be renewed or expanded, it’s no longer possible to procure at a competitive price. In that case, vendor lock-in keeps alternative providers out of the picture, even if they are lower-priced or offer better quality.

This is a classic coordination problem. European cloud providers are unable to scale without demand, and demand will not increase as long as European providers offer insufficient functionalities. In a market with winner-takes-all characteristics, it’s nearly impossible for new entrants to compete.

This justifies an active role for the government:

1. To solve the coordination problem and to get competition going; and
2. To put digital autonomy into practice.

This obviously shouldn’t even be up for debate, if only on the principle that, as a government, one shouldn’t relinquish control or authority over public data and critical infrastructures.

Key message: procure a share of cloud capacity needs from European providers

So where do you start, and how realistic is it? We don’t need to strive for complete independence, that’s not even possible. But that’s not an excuse for not doing anything. You can always take a first step, and then continue on the chosen path. It’ll take years, but you need to start somewhere.

The starting point is the observation that it is possible to have a share of demand for cloud services be met by European providers. This is feasible. Real-world examples of this exist in Europe, with clients from private and public sectors. Also, a group of seven IT companies (called the Open Cloud Alliance) recently presented a Dutch cloud offer for the Dutch government. So how do you take that first step? If there’s a will, it’s not that complicated: as a government, start with taking an easy and manageable step. As a launching customer, the government can commit itself to having at least, say, five percent of its annual cloud-capacity needs be met by European providers. In that context, demand open standards for file formats and communication protocols in order to reduce the dependence on proprietary standards.

This is obviously still quite complicated: it involves an IT transformation that requires flexibility and strong organizational control. However, a first step, even a modest one, is possible. Such a step has a dual impact:

1. Buying a share of our cloud services from European providers keeps the relevant data on European soil. That immediately increases our data sovereignty.
2. Buying from a wider range of providers as well as using European providers strengthens our outside options, including expertise and insight into security risks. That offers bargaining power vis-à-vis the major incumbents.

The percentage itself doesn’t matter. What matters is a feasible commitment. A small percentage immediately increases our bargaining power vis-à-vis the hyperscalers, and improves the business case for European providers. That instills confidence, and helps them in scaling up.

The Netherlands is unable to do this on its own. A collaborative approach with other European countries, or a broader European strategy, is essential. Unity increases security of demand for European providers, and helps governments collectively bear geopolitical risks.

At the same time, bundled contracts cannot be too large for European providers and consortiums. The key lies in smart procurement: with transparent criteria for autonomy and sovereignty, and opportunities for smaller European competitors to compete for contracts.

The government as trailblazer

The government often faces criticism about red tape, inertia and so on, but when it comes down to it, it’s up to governments to solve coordination problems, and they often can act more quickly than companies.

Moreover, there are external effects, as economists call them: the societal benefits of joint European procurement are larger than the private benefits for individual buyers. Purchasing decisions by businesses thus aren’t always in the public interest. In other words, the market fails. In that situation, the government should take the initiative.

With a launching customer in the market, demand increases more quickly, but, in that case, the government needs to lead the way, and get the snowball rolling. That’s a political choice, and susbequently it’s up to purchasing authorities to apply objective criteria regarding safety and sovereignty in their procurement processes. You often hear that the law supposedly prohibits this, but that’s a convenient excuse to take the easy way out.

In conclusion

The theme of this conference is: from debate to decision-making power. The challenge is enormous, but each step in the right direction counts. The question is: do we, as a government together with other governments in Europe, commit ourselves to yearly purchasing a modest yet meaningful share of our cloud services from European providers?

Think of five percent per year as a minimum, with clear criteria and transparent procurement processes. In that way, we will, one step at a time, increase demand, the scale of supply, our knowledge, and our bargaining power.

What does that mean for competition? Are coordination and cooperation not at odds with competition law? The point is that cooperation between providers actually improves competition in the market for sovereign cloud services. Smaller providers will be able to offer more-attractive propositions, which strengthens their ability to compete with hyperscalers. Bundling on both the supply and the demand side creates scale and competitiveness on the international market. This helps European competitors force a breakthrough in an oligopoly where three businesses call the shots. In short, a coordination problem is solved with coordination.

If you have any questions about potential conflicts with competition law, please let us know. We are happy to offer guidance and advice.