Sovereign AI needs more than national models

Singaporean Senior Minister of State Tan Kiat How recently announced that all government agencies are using artificial intelligence to improve public services, according to a report in the Bangkok Post. This development highlights a broader trend across ASEAN, where governments are looking to AI for efficiency gains and to deliver more personalized and transparent services to citizens. While a significant step, the focus on AI adoption within public agencies brings a deeper, more structural set of questions into view regarding national sovereignty in the digital age.

The Infrastructure of Sovereignty

The AI conversation is often dominated by the development and deployment of large language models. However, the true foundation of a nation's digital sovereignty lies in the layers of infrastructure that support these applications. As the book "ASEAN Rising" notes, the real test for sovereign AI in the region is "not by model launches but by who controls the compute, the data layer and the digital identity rails that sit underneath everyday economic life." For ASEAN governments, including Singapore, securing these foundational elements is a more pressing concern than simply deploying AI services. Without control over the computational power needed to train and run models, or the data that fuels them, any national AI strategy rests on a precarious base dependent on foreign providers.

This dependency creates long-term strategic vulnerabilities. If a nation's public services and economic activity become reliant on AI systems, the foreign entities that control the underlying cloud computing and data infrastructure hold significant leverage. This extends to digital identity, which is the system that verifies who is who in the digital economy. A robust, sovereign digital identity system is essential for secure and reliable public services, from tax collection to healthcare. When these rails are owned and operated by external players, a country risks ceding control over a core component of its national infrastructure.

Capital and Execution

Building this sovereign infrastructure requires immense capital investment and a clear-eyed view of execution. The cost of constructing and maintaining data centers, securing high-performance computing resources, and developing secure digital identity platforms is substantial. This is a domain where private capital and global technology firms have a natural advantage due to scale and specialization. For ASEAN states, the challenge is to create policy and investment environments that encourage the development of local or regional infrastructure, preventing a complete reliance on a few global providers. This may involve public-private partnerships, regional collaborations, and strategic investments in domestic technology companies.

Beyond capital, effective execution is paramount. Singapore