It is virtually impossible for us to become completely independent of American and Chinese technology. Let us therefore focus on how we can function if that technology is unavailable.
The (now lifted) export ban on Anthropic’s AI model Claude Fable 5 shows just how vulnerable Europe is to the whims of the US administration. More than 80 per cent of European spending on professional software and cloud infrastructure goes to US companies. Concerns about resilience, strategic autonomy and economic clout are therefore more than justified.
But there is also a great deal of activity, as shown by the list of European initiatives to reduce dependence on US big tech,compiled by Wired journalist Matt Burgess. From the city of Lyon, which is swapping Microsoft software for open-source alternatives, to the University of Groningen, which is starting to use Mistral AI, to DNS Belgium, which is migrating its infrastructure away from Amazon Web Services.
Nevertheless, the debate on technological sovereignty risks getting bogged down in a misguided discussion. All too often, it is presented as a choice between dependence and independence, between foreign technology and local alternatives. Such polarisation is tempting, especially with an unpredictable US president like Donald Trump. But it is based on a misconception: in a hyper-connected world, complete technological independence is simply not possible.
Take advanced chips, for example. The technological ecosystem required to enable modern AI systems spans several continents simultaneously. The most advanced lithography machines are built in the Netherlands, fundamental research is carried out in Belgium, amongst other places, the most powerful AI processors come from the United States, and the production of those chips takes place largely in Taiwan. No single country controls the entire value chain. No single region can completely sever its ties with the rest of the world without causing significant economic and technological damage.
Underestimated costs
We are, after all, interdependent. That is why Europe should not be asking itself so much how it can become completely self-sufficient, but rather how it can function if those dependencies are suddenly under pressure, as is the case with Claude Fable.
Replacing all foreign technology with European alternatives is financially unfeasible for most organisations and often technologically unwise as well. An organisation that decides to migrate cloud platforms, AI models or Enterprise Resource Planning (ERP) systems must invest in training, integrations, process adjustments and operational risks. The costs involved are often underestimated in the public debate.
Moreover, many local alternatives simply cannot compete with their American – and, increasingly, Chinese – counterparts. Of course, there are European success stories such as Mistral, Odoo, Collibra and Aikido, which prove that world-class solutions can be built here too. But not every player reaches that level. And as long as a local alternative offers less functionality, is less mature or evolves less innovatively than the international competition, there is no reason why companies would voluntarily choose it.
Instead of aiming for complete control, organisations should therefore strive for preparedness. What would happen if a crucial supplier were to disappear tomorrow? Which systems are critical? What alternatives exist? How quickly can data, processes and workloads be migrated? What contractual, technical and organisational barriers stand in the way of a switch?
These questions are rarely even asked in many companies. Organisations pay surprisingly little attention to their digital exit strategy. They know how to implement new systems, but not how to migrate away from them. They have roadmaps for growth, but no scenarios for disruption. Given the current geopolitical uncertainty, this vulnerability is becoming increasingly difficult to justify.
Björn De Vidts is the CEO of the Vlaams datanutsbedrijf Athumi
