Strategic autonomy is the ability to act on your own — even when others no longer want you to.
The concept comes from the vocabulary of defence: a power is strategically autonomous when it can decide and act without depending on the goodwill of a third party. Applied to the digital sphere, it raises a simple question: if a foreign supplier cut off its deliveries, shut down its service or changed its terms tomorrow, would Europe still be able to:
It differs from digital sovereignty, which is a matter of law and control — “who decides?”. Autonomy is a matter of capabilities — “whom do we depend on in order to act?”. One can be sovereign in law without being autonomous in practice: the GDPR applies throughout Europe, but without a homegrown champion in cloud, GPUs or devices, that law is exercised over technologies the continent does not control.
Unlike sovereignty, autonomy is never all or nothing: it is measured dependency by dependency — this step in the semiconductor chain, that building block of the cloud, that segment of the space sector. Nor does it mean autarky: the Union claims an “open strategic autonomy”, which does not seek to produce everything itself but to reduce critical dependencies in order to regain freedom of choice — the ability to say no, to switch supplier, or to hold firm in a crisis.
That is the work this pillar recounts: Europe’s capability programmes — the Chips Act for semiconductors, Galileo, Copernicus and IRIS² for space, EuroHPC and the AI factories for computing, the Cloud and AI Development Act for data centres —, the industrial sectors with their strengths and their gaps (the GPU value chain, critical raw materials), and the successful switchovers, those migrations of entire organisations that prove the alternatives work at scale.
Strategic autonomy is the material condition that makes sovereignty effective: a right that cannot be enforced, for lack of alternatives, remains declaratory. Rebuilding capabilities means giving the law its strength back.