The UK government has forecast up to £45 billion in annual productivity savings from the full digitisation and AI adoption across public sector services, framing the figure as an unrealised opportunity rather than a scheduled reduction in spending. Publicly described as a “£45bn jackpot,” the forecast is positioned as a major lever for fiscal consolidation, with potential gains compared to the annual funding requirements of all UK primary schools to illustrate the scale involved.
The government’s breakdown attributes roughly £36 billion of the total to simplifying and automating service delivery across public sector organisations, from the NHS to local councils and central government departments. A further £4 billion is linked to shifting services toward cheaper online channels and away from more costly offline alternatives. Fraud and error reduction through digital compliance systems accounts for the remaining £5 to £6 billion, depending on the source consulted.
Internal analysis suggests productivity-related savings could range between £36 billion and £72 billion annually, with the lower figure selected as a conservative baseline.
The methodology behind the headline number has drawn scrutiny. Core analysis was initially conducted by the Central Digital and Data Office, examining personnel spending of approximately £79 billion across central government, the NHS, and police services. Those findings were then extrapolated to the entire UK public sector to generate the broader forecast.
The underlying model assumes automation productivity gains of between 15 and 30 percent of baseline spend, alongside assumptions of 100 percent automation of routine tasks and 10 percent automation of non-routine tasks. The National Audit Office has noted the indicative nature of the analysis and flagged the absence of a full feasibility and cost assessment, raising questions about how reliably the projections reflect what is practically achievable.
On the implementation side, the Department for Science, Innovation and Technology has previewed an AI package called Humphrey, designed to assist civil servants as part of a broader productivity drive. Government trials of AI-generated code reported savings of approximately one hour per programmer per day, equivalent to around 28 working days per year. Technology secretary Peter Kyle described the adoption of AI tools as central to unlocking the public sector’s unrealised digital potential.
Separate DSIT research claims AI tools could save nearly two working weeks per person annually and assist with roughly half of office-based tasks. These deployments are directly tied in government communications to ambitions around cutting backlogs and modernising processes.
External analysis by the Tony Blair Institute has referenced government claims of up to £40 billion annually in productivity gains from AI and digital transformation, broadly aligning with official figures. However, with the savings characterised as unrealised rather than banked, and with the National Audit Office highlighting gaps in feasibility assessment, the gap between forecast and delivery remains a significant open question for policymakers and public sector workers managing the shift.
