AI compute-cluster development becomes explicit national-security infrastructure in ≥5 countries by 2027, with sovereign compute-capacity targets (US Stargate model, EU sovereign compute, China domestic-chip strategy, UAE-Saudi sovereign compute)?

By late 2027, at least five governments have declared compute clusters critical infrastructure and set formal capacity targets. The US Stargate model spreads: Congress funds domestic GPU farms; the EU AI Factories program locks in 2026 procurement goals measured in exaflops; China's State Council publishes a Five-Year Plan annex with Huawei Ascend deployment quotas; Saudi Arabia's PIF and UAE's G42 both announce sovereign petaflop thresholds tied to oil revenue. Compute capacity becomes a line item in defense white papers alongside aircraft carriers and missile inventories.

Governments talk big on AI sovereignty but stop short of formal infrastructure designation. Stargate remains a one-off public-private hybrid without imitators. The EU AI Factories initiative stalls in procurement committees; China keeps cluster expansion implicit inside industrial policy rather than explicit targets; Gulf states pivot compute investments toward private partnerships when oil revenues tighten. By end of 2027, only two or three countries have both declared compute as critical infrastructure and published measurable capacity goals—below the five-country threshold.