Sovereign cloud stopped being a slogan when the contracts started specifying it.

For several years, European sovereign cloud requirements were satisfied by a region selector. The data sits in Frankfurt or Stockholm, the box is ticked, the deal proceeds.

That era is ending, and the reason is that buyers in regulated sectors started reading their own risk registers more carefully.

Residency, then control, then dependency

Data residency asks where bytes are stored. Operational control asks who can access the environment, under which legal compulsion, and whether the service continues to function if that access is withdrawn.

The third question is the one now appearing in bank and public sector tenders: if this provider became unavailable to us, how long until we run elsewhere, and what does that cost.

The answers being offered

Three models are in the market. Partner-operated regions, where a local company holds operational control under licence. Disconnected or air-gapped regions for the most sensitive workloads. And a build-your-own path on European providers or on-premise hardware.

Each trades capability for control in a different proportion. The partner-operated model preserves most of the service catalogue and the least independence. The on-premise path inverts both.

What buyers are actually doing

Almost nobody is migrating wholesale. The pattern is tiering: a small set of genuinely sovereign-critical workloads placed under strict control, and the remainder left on commodity hyperscale with improved exit documentation.

The exit documentation is the underrated part. Several CIOs interviewed described the exercise of writing a credible exit plan as more valuable than the migration they ultimately did not perform, because it exposed which dependencies were accidental.

The cost

Sovereign tiers carry a premium, typically described as somewhere between a quarter and double the equivalent commodity spend depending on model and service mix, plus an engineering cost in reduced managed service availability.

Organisations that priced this honestly made better decisions than those that treated sovereignty as a procurement checkbox with no budget line.