AI and technology tasked with cutting civil service costs by 15%

Experts raise concerns over feasibility and impact

The UK government’s plans to cut civil service running costs by 15% by the end of the decade will rely heavily on advances in technology and AI, chancellor Rachel Reeves has confirmed.

Reeves stated that AI is already being used in departments such as HM Revenue and Customs and the Department for Work and Pensions to tackle fraud and error and that further efficiencies could be achieved through increased automation. “Anyone running a business will think that the 15% savings target is more than possible because of advances in technology and AI,” she told BBC News. “Using technology alongside people, but using more technology to do some of those jobs.”

This follows comments from science and technology secretary Peter Kyle, who recently suggested that the civil service headcount would “almost certainly” decrease as AI is integrated into government operations, though he insisted that the government would not impose an “arbitrary overall figure” for job cuts.

Around the same time prime minister Keir Starmer said technology will be essential to reshape the “flabby” British state, with AI deployed to drive efficiency and streamline public sector operations.

In January, the government unveiled its AI Opportunities Action Plan, with 50 key recommendations to drive AI innovation in government, followed shortly afterwards by a blueprint for digital government, which included a new package of AI tools – nicknamed “Humphrey” – to be made available to civil servants to boost productivity and deliver better public services.

Impact of AI adoption

However, critics have warned about the feasibility of relying so much on new technology to drive change, given the current state of the digital infrastructure.

In an email to Computing Rob Anderson, chief analyst and service director, public sector at analytics and consulting company GlobalData, warned that the UK government faces “a huge challenge in introducing AI-enabled change due to its somewhat archaic structure and siloed culture.” He noted that AI implementation requires “a robust data infrastructure which is currently lacking and very segmented by department.”

Anderson also highlighted structural barriers to effective AI integration, stating that “HM Treasury funding allocation mechanisms need a radical overhaul,” as departments are currently incentivised to focus on their own needs rather than cross-departmental collaboration.

Laura Petrone, principal analyst at GlobalData, echoed these concerns, describing the government’s approach as “mainly aspirational” and lacking in detail. “Beyond the rhetoric, there is little clarity around how AI could operate within the UK civil service, what type of work it would do, and on which timeframe,” she said.

Petrone also warned against the perception that AI could replace entire roles, stating that AI is currently best suited to performing specific tasks rather than taking over jobs completely. “The extent to which AI can take on white-collar jobs, with what outcomes, and when is still unclear. We don’t expect this to happen within the next few years.”

She emphasised the importance of AI literacy among civil servants, noting that many will need to be upskilled to work effectively with AI. “AI shouldn’t be imposed on civil servants, but it should be viewed as a tool that can help them do their jobs more effectively,” she said.

While the government insists that AI-driven efficiencies will lead to “better public services,” experts caution that without a clear strategy, appropriate oversight and investment in data infrastructure, the promised benefits of AI integration in the civil service may remain elusive.