By: Staff Writer
January 06, Colombo (LNW): Sri Lanka’s bold plan to create a network of national and potentially international AI data centres has generated excitement in policymaking circles, but the budget allocation of just Rs. 3 billion for 2026 has triggered intense debate among economists, technologists, and industry analysts who argue that the funding bears little resemblance to the scale of investment required to build and operate competitive AI infrastructure.
In the recently unveiled 2026 national budget, Rs. 3 billion was earmarked for AI development, national data platforms, and government-managed data centres as part of the country’s strategy to become a digital hub in South Asia. This includes Rs. 500 million specifically set aside to attract global and local investors to build data centres through subsidised land, lower initial electricity tariffs, and green energy incentives, alongside Rs. 750 million for AI research, cloud innovation, and workforce development programmes.
However, critics argue that these figures are insufficient even for preliminary engineering studies, let alone actual construction of data centres. Industry benchmarks suggest that a single mid-scale data centre without significant AI capabilities—can easily cost between Rs. 20 billion and Rs. 40 billion, a figure that dwarfs the entire AI allocation for Sri Lanka in 2026.
The absence of a publicly released feasibility study has become a central point of contention. Analysts warn that without a detailed assessment of site suitability, power availability, cooling requirements, network connectivity, and disaster risk, and long-term operational costs, government projections risk becoming aspirational slogans rather than realistic strategies.
One top official involved in planning acknowledged that natural disasters, such as Cyclone Ditwah and subsequent landslides, have made site evaluation far more critical than previously understood. Choosing unstable land for data centres could expose the entire project to prolonged outages, higher insurance costs, and long-term reputational damage.
Supporters of the project point to Sri Lanka’s non-aligned geopolitical stance and strategic location as potential competitive advantages. Officials have floated the concept of hosting “data embassies”secure AI and data services for foreign governments positioning Colombo as a regional digital hub. Yet such high-level aspirations require substantial backing, robust infrastructure, and investor confidence grounded in feasibility analysis rather than budgetary tokenism.
Another concern is energy capacity. AI-centric data centres require immense continuous power, sophisticated cooling, and redundancy systems. Even in neighbouring markets with far stronger technology ecosystems, building an AI-ready facility can run into hundreds of millions of dollars, excluding the ongoing electricity and maintenance costs.
Critics argue that the government’s approach risks repeating past patterns—announcing high-profile technology initiatives before establishing the technical, fiscal, and risk frameworks to support them. For Sri Lanka to genuinely compete in the global AI economy, analysts say, a realistic budget aligned with feasibility planning and staged investment—not token allocations—must be prioritised
