The dark disc is national territory. Each glowing packet is a citizen's data request: it leaves home, is routed to a datacenter, gets processed, then returns. The platform inside the border is the country's own sovereign cloud — short trip, low latency, no foreign exposure. The three lit towers standing outside the border, floating above "the cloud," are foreign providers in different jurisdictions: a US hyperscaler, an EU cloud bound by GDPR-style rules, and a state-controlled cloud with weaker guarantees. Every request that crosses the border accrues dependency risk while it's away; a sovereignty-investment programme (local capacity, encryption, legal safeguards) burns that risk back down over time, but never travelling abroad in the first place is the surest way to keep it near zero.
route ∝ weight(node) · (1 ± localization_mandate)
risk(t+dt) = risk(t) + Σ_foreign(active) · risk_rate·dt − investment·decay·risk·dt
- Cloud provider toggles — take a jurisdiction out of the routing pool entirely, e.g. to model a foreign-cloud ban; disabled requests fall back to whichever providers remain, or entirely to the sovereign cloud if none are left.
- Localization mandate — bias routing toward the sovereign cloud (right) or toward whichever foreign providers are cheapest/fastest (left).
- Request pace — how quickly citizens issue new requests and how fast in-flight ones travel.
- Sovereignty investment — funding for local infrastructure, encryption and legal safeguards; higher values decay accumulated dependency risk faster.
Real-world relevance: this mirrors the actual trade-off behind data-localization law, GDPR-style transfer restrictions and "sovereign cloud" initiatives — keeping data local cuts foreign-dependency risk but each foreign hyperscaler taken off the table can raise latency and cost until domestic capacity catches up.