en.Wedoany.com Reported - According to the "2026 Sovereign AI Large Language Model (LLM) Report" released by Counterpoint Research, the "Falcon H1" developed by the United Arab Emirates has been rated as the world's most mature sovereign AI model, with Russia's "GigaChat 3.1 Ultra" ranking second. The report notes that Middle Eastern countries, leveraging government-led investment and a model development ecosystem optimized for local languages, have taken a leading position in the global sovereign AI race.
Falcon H1 scored highly in the dimensions of government-led ownership structure and self-developed foundation models, with its local language support and actual deployment scope also regarded as key competitive advantages. Saudi Arabia's "Allam" as well as the UAE's "Zai 2" and "K2 Sync V2" were also rated as leading models. The analysis suggests that the UAE and Saudi Arabia, supported by Arabic dialect support and broad accessibility, are spearheading the development of the Middle East's sovereign AI ecosystem.
In South Korea, SK Telecom's self-developed "A.X 3.1" has earned recognition for differentiated competitiveness thanks to its extensive deployment rate. The model is built on 34 billion parameters, and SK Telecom open-sourced it on the global AI platform Hugging Face in 2025. Counterpoint Research noted that South Korea is rapidly expanding its related ecosystem based on sovereign AI investment plans of up to $1 trillion. South Korea, Japan, and India were classified as leading countries in sovereign AI development in the Asia-Pacific region, excluding China.
The report covers more than 170 models developed across over 80 countries, including the Middle East, Central Europe, Eastern Europe, Western Europe, Asia-Pacific, Central Asia, Africa, Latin America, and Canada, assessing the maturity of sovereign AI in each country across dimensions such as model ownership structure, access to foundational technology, local language support, and deployment level.
Counterpoint Research analyst Mark Einstein stated that countries are emphasizing AI sovereignty from the perspectives of data security, language diversity, and technological independence. He also pointed out that following the ban on Fable 5 in Claude, the political risks of foreign LLMs have risen, further reinforcing the necessity of sovereign models.
