Should states provide sovereign compute, national models, public datasets, shared safety infrastructure, or publicly funded alternatives?
Sovereign-AI programs answer a dependency worry: nearly all frontier capability sits with a handful of US and Chinese firms. The objection is cost realism — states buying relevance in a game priced in tens of billions.
View on the map → · Open in Browse →
Sovereign compute got a price tag and a ceiling: the EU opened tenders for up to seven publicly financed AI gigafactories targeting operation by 2028, while the sharpest analyses argue sovereignty programs cannot insulate against the real dependency — a US ability to switch off allies' access and reach their data — and that full-stack autonomy is unaffordable for most states anyway.
Washington can't promise allies it will never cut off their access to American AI or reach their data—but it can set, and eventually codify, clear rules for both.
Sharpens the dependency worry behind sovereign-AI programs: EU sovereignty initiatives cannot fully insulate against US cut-off and data-reach powers — only US self-binding rules can. Reframes what sovereign compute can and cannot buy.
The European Commission has launched a call for tenders to publicly finance up to seven AI gigafactories in Europe
The concrete milestone in Europe's sovereign-compute bet: the formal tender for up to seven publicly financed AI gigafactories targeting mid-2028 — the cost-realism question now has a price tag and a timeline.
AI sovereignty poses a tougher question: how much control is economically and institutionally feasible?
The cost-realism counterargument: full-stack AI autonomy is unaffordable for most states, so sovereignty should be reframed as selective control over strategically chosen layers rather than national replication of the frontier.
Public compute initiatives are the first step, aimed at securing the AI development capacity needed.
Articulates the public-AI agenda beyond compute subsidies: public compute, datasets, and models as democratic infrastructure rather than entries in a national AI race.