Asia Pacific governments boost sovereign AI priorities

Governments across Asia Pacific are placing greater emphasis on sovereign AI, according to research commissioned by Dell and conducted by IDC. The study found it rose from seventh to second in public sector investment priorities over the past year.

The survey covered 360 government IT decision-makers in Australia, India, Indonesia, Japan, Malaysia, the Philippines, Singapore and South Korea, alongside interviews with senior officials. It suggests many administrations have moved beyond early debate over AI adoption and are now assessing how to deploy systems while keeping tighter control of data, infrastructure and policy.

Almost half of respondents, 46.1%, said they were evaluating sovereign AI technologies, while 36.1% were running pilots or proofs of concept. Only 3.1% reported significant investment, highlighting a gap between strategic interest and large-scale rollout.

That caution appears to reflect practical concerns. More than three-quarters of respondents, 76.9%, said sovereign AI investment would improve resilience against geopolitical risk and supply chain disruption. Even so, most governments are still weighing how to build those systems in ways that meet security and regulatory requirements.

The findings indicate that many public sector organisations are pursuing what the report calls selective sovereignty. In practice, that means keeping tighter control over sensitive data, critical systems and regulated workloads while continuing to use international technology suppliers where they see value. Hybrid models combining on-premises systems with sovereign cloud arrangements are emerging as the preferred approach.

Agentic AI

The research also points to strong interest in agentic AI across the region. Almost all respondents, 99%, said they saw agentic AI as an accelerator for government adoption. Of those, 36.9% said it would play a major role, while 62.1% expressed strong confidence in its use when backed by governance and oversight.

Just 1.1% said they remained uncertain. That confidence reflects a view among officials that AI systems could help administrations cope with staff shortages and rising operational pressure, particularly in back-office and analytical work.

Skills shortages emerged as one of the clearest constraints on progress. Nearly nine in 10 government organisations reported digital skills gaps, and more than half said those shortages were having a major effect on their digital programmes.

The hardest roles to fill were not limited to technical engineering posts. Respondents identified AI safety and alignment researchers as the most difficult to recruit, at 42.5%, followed by data architecture, management and analytics specialists at 35%, sovereign data governance specialists at 30%, sovereign cloud architecture and operations specialists at 25.3%, and AI policy and governance specialists at 25%.

Those figures suggest the main challenge for many governments is not simply access to software or computing infrastructure, but building teams that can govern AI systems safely and in line with national requirements.

Public services

The study found governments expect the biggest gains from sovereign AI in areas where public accountability and operational risk are high. National security and cyber resilience ranked first at 45.6%, followed by justice and public safety at 37.5%, financial and taxation functions at 37.5%, public healthcare at 34.4%, social services and welfare at 32.2%, education at 31.7%, and workforce development at 31.1%.

National policy priorities are also shaping investment decisions. More than half of respondents, 53.3%, said alignment with national security and sovereign priorities was the leading factor in technology investment decisions. Security and provider reliability followed closely at 52.5%.

The report recommends a four-layer model for governments seeking to scale sovereign AI. Under this approach, administrations would retain direct control over policy, governance and data stewardship, while relying on external partners for more specialised AI work and broader deployment.

Nicole Jefferson, vice president of global government affairs at Dell, said the findings reflected what public sector leaders were discussing across the region. "This research confirms what we're hearing from government leaders across Asia Pacific: the question is no longer whether Sovereign AI matters, but how to operationalise it at national scale," Jefferson said. "What stands out is the region's confidence in agentic AI as an accelerator and the understanding that strong governance is an enabler of progression, not a hinderance. The findings show a region that is serious, structured, and pragmatic about building AI capabilities it can trust. Governments want partners who understand that sovereign-ready infrastructure, skills transfer, and governance maturity are inseparable from the technology itself. At Dell Technologies, we're committed to helping public sector organisations build AI on their own terms - with the security, resilience, and openness that mission-critical national services demand."

IDC said momentum behind autonomous systems would depend on whether governments believed the necessary safeguards were in place. "Agentic AI is moving quickly from concept to practical consideration for government and executive decision-makers," said Ravikant Sharma, research director at IDC. "The study shows strong momentum, with public sector leaders looking to autonomous systems to help close skills gaps, ease workforce pressure and accelerate AI adoption. However, that momentum is conditional. Governments will only move at scale if they have confidence in the security, privacy, sovereignty and infrastructure foundations underpinning these systems."