Sovereign AI – nations' pursuit to develop and maintain independent AI capabilities without reliance on foreign entities – has become a key strategic objective worldwide by 2025, with full independence achievable only for the United States and China. The concept began gaining prominence around 2020, particularly within European policy discussions, but was elevated in visibility by Nvidia CEO Jensen Huang starting in November 2023.
True AI sovereignty requires self-sufficiency across all levels of technological infrastructure, including chips, data centres, models, and talented workforce, which remains unattainable for most countries. US and Chinese companies – with Nvidia holding 80% global market share and Huawei commanding 75% of the Chinese market – dominate the AI chip market, while data centre infrastructure is controlled by US hyperscalers: Amazon Web Services (32% global market share in 2024), Microsoft Azure (23%), and Google Cloud (10%). The UK pledged £2 billion in funding for its AI strategy in early 2025, including £1 billion on building out AI compute by 2030, while the EU announced the InvestAI initiative in February 2025, aiming to mobilize €200 billion for AI development.
Full AI sovereignty means control over the entire value chain, but most nations can only achieve partial autonomy, such as domestic development of culturally and linguistically appropriate AI models. Significant efforts in region-specific models include OpenEuroLLM – a collaboration between 20 organizations – which launched a €37.4 million EU-backed initiative to develop open-source language models covering all European languages by 2028; in India, AI4Bharat raised $12 million in starter funding to develop India-centric LLMs; and the UAE's Technology Innovation Institute launched Falcon Arabic, a model optimized for the Arab world.
Sources:
Robert Dale
Vincent Caldeira
