Research — April 2026

G7 AI Architecture Divergence

Six of seven G7 nations are investing in sovereign AI infrastructure, domestic foundation models, and architectural independence. One is consolidating toward four vendors.

6
Building sovereign
France, Germany, Japan, Canada, UK, Italy are investing in domestic AI infrastructure, open-weight models, and architectural independence from US vendors.
1
Consolidating centralized
The United States is preempting state regulation, taking equity stakes in vendors, and framing deregulation of four companies as national AI strategy.
France
Sovereign + Open-Weight
€109B AI infrastructure · Mistral €2.8B raised
Open-weight models (Apache 2.0) deployed on French-controlled servers. Military eliminated US cloud dependencies. Mistral building 200MW sovereign data center capacity by 2027. Framework agreements with Germany for public administration AI.
Key signal
“This is our fight for sovereignty, for strategic autonomy.” — Macron
Germany
Sovereign + EU-Aligned
100,000 GPUs via Nvidia / Deutsche Telekom by 2027
Joint Digital Sovereignty Summit with France (Nov 2025). Partnered with SAP and Mistral to build sovereign AI stack for public administration. EU AI Act compliance as architectural driver.
Key signal
Chancellor Merz: GPU deployment is “an important step toward digital sovereignty.”
Japan
Sovereign + Domestic Models
¥1T ($6.4B) five-year plan · $5.5B AI infra market 2026
First National AI Basic Plan (Dec 2025). Fujitsu manufacturing “Made in Japan” sovereign AI servers. METI joint venture with 10+ companies for domestic foundation models. Focus on “Physical AI” for robotics and manufacturing.
Key signal
Domestic foundation models, sovereign hardware manufacturing, AI as national security infrastructure.
Canada
Sovereign Compute + Open Source
C$2B Sovereign AI Compute Strategy · C$925.6M Budget 2025
Sovereign compute infrastructure program. AI Compute Access Fund for domestic firms. Open source identified as “third path” between building walls and accepting subordination. New national AI strategy expected 2026.
Key signal
PM Carney at Davos: “Shared standards reduce fragmentation.”
United Kingdom
Trending Sovereign
£500M Sovereign AI Unit · £18B infrastructure program
Sovereign AI Unit launched April 2026. Innovate UK funding distributed and decentralized systems. £250M for AI Research Resource. Isambard-AI supercomputer operational. Hedging with concurrent OpenAI/Stargate UK partnership.
Key signal
“An AI maker, not an AI taker.” Distributed systems funded as sovereign research priority.
Italy
Institutional Capacity
AI4Industry €20M/yr · 8 Competence Centers
G7 2024 Presidency launched AI Code of Conduct Reporting Framework. AI HUB for Sustainable Development. National AI institute in Turin. Focus on SME adoption and institutional capacity-building rather than frontier model development.
Key signal
Institutional infrastructure over model competition. Standards and governance as national contribution.
United States
Centralized + Vendor Consolidation
$500B Stargate (announced) · $8.9B Intel equity stake
Federal preemption of state AI laws (Dec 2025 EO). DOJ mobilized against state regulation. Commerce Dept took 9.9% equity in Intel. $10B+ in equity positions across semiconductors. AI strategy routes through four companies. Flagship Stargate campus failed first winter stress test.
Key signal
“To win, United States AI companies must be free to innovate without cumbersome regulation.” — Dec 2025 EO

The divergence is not about policy preferences. It's about architectural choices with structural consequences. Sovereign architectures reduce latency, eliminate vendor lock-in, enable auditability, and distribute strategic risk. Centralized architectures concentrate all four. Every G7 nation except the United States has concluded that the structural risks of centralized AI dependency outweigh the convenience of incumbent vendor ecosystems.