Public sector organisations are accelerating AI adoption, but 73% of their infrastructure is currently unready to run complex AI workloads on-premises, according to new research from hybrid multicloud computing firm Nutanix.

The findings come from the public sector vertical edition of Nutanix‘s eighth annual Enterprise Cloud Index (ECI) survey, which gathered responses from 1,600 cloud, IT and engineering executives across 14 countries, including Singapore, in November 2025. The public sector cut covers federal, state and local governments, K-12 schools and higher education institutions.

Shadow AI a growing governance risk

The report found that 91% of government and education IT leaders surveyed believe unvetted AI usage — commonly referred to as Shadow AI — creates mission and security risks. At the same time, 87% of public sector technology leaders expect their reliance on application containerisation to increase over the next three years to better support AI deployments.

Nutanix said the findings point to a sector at a critical juncture: AI is already running across government and education environments, from containerised applications to hybrid infrastructure and, increasingly, direct service delivery, but the underlying foundation is uneven as agencies juggle legacy on-premises systems, private clouds and cloud migrations simultaneously, often without a unified control plane.

“The proliferation of unvetted AI, or ‘Shadow AI’, has emerged as the defining governance challenge for the public sector. Across Asia Pacific and Japan, the conversation has shifted from AI ambition to operational readiness. The organisations moving ahead are those that stop chasing every new capability in isolation and instead consolidate their foundations,” said Daryush Ashjari, Chief Technology Officer and Vice President of Solution Engineering, APJ, at Nutanix.

Relevance to Singapore’s AI strategy

Nutanix said the findings are particularly relevant to Singapore, which continues to expand AI use across the public sector under its National AI Strategy 2.0. As AI becomes more deeply embedded in public services, the company argued that secure, modern and well-governed hybrid infrastructure will be critical to scaling AI while safeguarding sensitive data and maintaining public trust.

For public sector IT leaders, Nutanix said this creates a clear mandate: AI workloads touching citizen services, public safety and educational outcomes require infrastructure that can deliver performance, maintain regulatory compliance and support mission governance — not just centrally, but at the point of delivery, where data sovereignty directly affects the communities government exists to serve.

The global ECI survey was conducted by Wakefield Research on behalf of Nutanix, gathering responses from organisations with 500 or more employees across Australia, Brazil, France, Germany, India, Italy, Japan, Mexico, the Netherlands, Saudi Arabia, Singapore, Spain, the UK and the US.