To bring more clarity to this, Europe is working on a framework for cloud and AI sovereignty: the CADA framework (Cloud and AI Development Act). It is still a draft, but the intent is clear. Technology and services are classified according to four assurance levels, depending on how much control stays within the Union. A bit like the Nutri-Score in the supermarket, but for technology.
In practice, it looks like this. Governments choose the appropriate level per situation based on a risk assessment, and providers are recognised per level after an audit.
LEVEL | WHAT IT MEANS |
Level 1 | Data is processed and stored on infrastructure within the European Union. |
Level 2 | The provider demonstrates independence from third countries, with transparency about its software supply chain. |
Level 3 | The provider is in EU hands and is controlled from the EU, with extra criteria such as the nationality of the staff. The Commission can also recognise providers from third countries. |
Level 4 | The provider has full transparency and control over its software supply chain, without interference from a third country. |
*Based on the CADA proposal (European Commission, June 2026).
If you are a vendor wanting to show how European your solution really is, questions like these come up:
where the infrastructure is located
who owns the technology
where the staff are based
who has control over the software
which subcontractors are involved
how cybersecurity and governance are arranged
And of course, this is validated through audits. You can already feel it coming: an entirely new market for certification and auditing around digital sovereignty is likely to emerge.