Five countries operationalize sovereign AI strategies and infrastructure

The article analyzes how five jurisdictions (EU, India, Japan, South Korea, Brazil) are operationalizing a policy of sovereign AI, shifting from regulation toward capability building and infrastructure investment.

• Main announcement/action: Governments are building sovereign AI capacity via compute subsidies, targeted industrial policies, and regulatory frameworks: the EU is launching AI Factories and Gigafactories; India’s IndiaAI Mission funds over 10,000 GPUs and targets a trillion-dollar economic impact by 2035; Japan designates cloud programs as critical products eligible for billions in subsidies; South Korea pursues an “AI G3” strategy relying on National Champion firms (e.g., Naver, LG); Brazil upgrades HPC (anchored by the Santos Dumont supercomputer) and advances Bill 2338/2023/PBIA for digital sovereignty.
• Background and specifics: Approaches sit on a spectrum from protectionism to partnership: the EU and Brazil favor risk-based, comprehensive legislation (EU AI Act; Brazil’s Bill 2338/2023); Japan and India emphasize innovation-friendly or soft-law approaches respectively; cultural and demographic drivers shape priorities (e.g., Society 5.0 in Japan; Portuguese-language models under Brazil’s PBIA). Implementation details include GPU provisioning (India), HPC upgrades (Brazil), subsidy programs (Japan), and industrial partnerships with firms like OpenAI and Microsoft in Japan.