New research highlights rising costs and efficiency gaps in government cloud

New research released by Government Transformation Magazine, in partnership with IBM, reveals that central government’s cloud ambitions are being hampered by spiralling costs, entrenched legacy burdens and inconsistent use of optimisation tools - despite widespread cloud adoption across major departments.

The report, Unlocking Cloud Efficiency for Smarter Government, draws on findings from more than 100 senior DDaT leaders and operational teams across Whitehall. It provides a snapshot of how cloud spending is managed, where costs are escalating, and what departments need to do to unlock savings that can fund AI-driven innovation and improved citizen services.

A central theme in the research is the scale of legacy debt that continues to drain transformation budgets. All strategic decision-makers surveyed said that at least 10% of their technology budgets are absorbed by legacy systems, with nearly half reporting that between 26–50% of total spend goes toward maintaining outdated platforms. This burden directly diverts funding from integration (54%) and internal efficiency improvements (51%) that leaders say are essential for modernisation.

While cloud adoption has delivered performance benefits, with 51% citing faster delivery and 42% improved resilience, only 33% of respondents reported tangible cost savings. The findings suggest that although departments have embraced cloud for agility and service resilience, many have yet to translate those gains into measurable financial value.

Cost inefficiencies remain common at the operational level. Excessive data transfer and storage costs were reported by 48% of teams, while 43% highlighted overprovisioned compute as a recurring issue. Despite the prevalence of FinOps-style practices, adopted by 81% of organisations, many departments appear to struggle to move from visibility to optimisation. Tooling landscapes remain mixed, with third-party platforms (57%), AI-enabled tools (48%) and manual scripts (43%) all used in parallel, often without consistent standards or governance.

One of the most striking findings concerns the limited use of AI for cloud cost control. Although 67% of operational teams report using AI or automation in some form, just 19% have deployed these capabilities to optimise cloud workloads, despite emerging industry evidence that automated scaling, anomaly detection and rightsizing can significantly reduce expenditure.

Leaders nevertheless remain optimistic about the potential for reinvestment. A strong majority (77%) believe that cloud savings, if unlocked, could be redirected into service improvement and operational efficiency, a sentiment echoed by senior figures at HMRC, NAO and DSIT interviewed in the report. Many point to AI-enabled automation, legacy retirement and improved data quality as the areas most likely to deliver returns if funding can be freed.

The report concludes that government is not lacking tools or intent, but is held back by systemic blockers, including fragmented procurement, inconsistent tagging and accountability gaps between service and platform teams. It suggests that without clearer ownership, integrated standards and stronger FinOps maturity, savings will remain out of reach.

Download your copy of the research below.