Middle East Leads Global Sovereign AI Race, UAE Tops Counterpoint Index
The UAE's Falcon models are helping cement the Middle East's leadership as countries across the region build sovereign AI capabilities.
News
- Mubadala Eyes Up to $6.3 Bn Investment in Japan AI Data Center
- Alibaba Plans Revenue-Sharing Model for Next Open-Weight AI Release
- AI Dispatch: Global Reorder and a Region on the Rise
- Ajman Targets 100 AI Initiatives Under Government-Wide Program
- JPMorgan CEO Leads 40-Company Effort to Tackle AI Risks
- Identity, More Than Software Flaws, Is Driving Today's Costliest Cyberattacks
[Image: Nomita Samaiyar/MITSMR Middle East]
The global race to develop sovereign AI is becoming increasingly regional, with countries pursuing distinct strategies to build and control their own large language models (LLMs), according to Counterpoint Research’s latest Sovereign AI LLM Index. The report ranks the Middle East as the world’s most mature sovereign AI region in the first half of 2026, led by the UAE.
The index credits the UAE’s position to the Technology Innovation Institute’s Falcon H1 model, while noting that the country’s broader sovereign AI ecosystem also includes contributions from Mohamed bin Zayed University of Artificial Intelligence (MBZUAI).
Across the wider Middle East, the report says Iran continues to rely primarily on adapted open-source models rather than homegrown foundation models.
Qatar is expected to launch Fanar 3.0 later this year, Iraq is developing a government-backed sovereign LLM, and Oman plans to expand its Mu’een model through the Ministry of Transport, Communications and Information Technology. Kuwait, Jordan, and Bahrain remain in the planning phase, according to the index.
In the Asia-Pacific region, the report identifies South Korea, Japan, and India as the region’s leading sovereign AI developers. It says India’s Sarvam 105B has strengthened the country’s position through a public-private partnership model, while South Korea accounts for 17% of the region’s foundation models and Japan for 11%.
Unlike several Middle Eastern initiatives, both countries rely primarily on government-backed private companies rather than state-owned programs. Japan’s ecosystem includes developers such as Preferred Networks, NTT, SoftBank, and Rakuten under the GENIAC program, while South Korea’s efforts are led by LG AI Research, SK Telecom, and Naver with support from the Ministry of Science and ICT.
Russia dominates the sovereign AI landscape in Central and Eastern Europe, accounting for about 30% of the region’s sovereign LLMs, according to the report. Its GigaChat 3.1 Ultra, a 702-billion-parameter model, is identified as the region’s largest. Turkey accounts for roughly 16% of sovereign models, while most other countries continue to build on adapted open-source models despite strong academic research capabilities.
Western Europe has developed one of the world’s most mature sovereign AI ecosystems through public investment, research institutions, and collaborative AI programs. The index says 59% of the region’s sovereign models are foundation models, with France’s Mistral AI emerging as the leading developer.
Elsewhere, sovereign AI efforts remain at an earlier stage. Central Asia and Africa rely largely on adapted open-source models, which account for nearly 80% of sovereign AI activity in those regions. Egypt and Kazakhstan lead government-backed initiatives, while South Africa and Egypt are the continent’s principal developers of foundation models.
In North America, Canada’s sovereign AI ecosystem is concentrated around Cohere and its Tiny Aya and Command A models. In Central and Latin America, sovereign AI development remains largely limited to Chile and Brazil, where current efforts continue to build on Meta’s Llama models rather than independently developed frontier-scale systems, the report says.
