- Social Market Foundation critiques £3.25 bn AI Transformation Fund
- Highlights inconsistent adoption and unclear strategy across departments
What happened: Capita‑backed think tank scrutinises UK Civil Service AI fund
The Social Market Foundation (SMF), supported by Capita, has criticised the UK government’s AI Transformation Fund. This fund allocates £3.25 bn to embed artificial intelligence across the civil service. The criticism emerged through Freedom of Information requests. These revealed wide variations in AI progress across departments. Some bodies have started trials and AI tool deployments. Others are still in initial planning stages. The SMF noted that the fund’s allocation through existing Treasury bidding may not suit AI’s iterative development approach. It described this model as poorly aligned with agile innovation methods.
What’s more, the SMF pointed out that around £150 m from the fund was directed towards voluntary redundancy schemes rather than AI projects. Additionally, £8 m supported administrative technology in probation services with uncertain AI relevance. The Foundation argued that this spending pattern reveals a lack of strategic fall‑through from fund headlines to substantial AI deployment. The Department for Business and Trade reported 34 AI use cases as of April 2025, compared to just a few in other major departments. Many initiatives are still at pilot stage, with unclear plans for scaling or sunset. The SMF concluded that inconsistent adoption and the lack of a clear, cohesive rollout plan may risk undervaluing the fund’s effectiveness.
Also Read: Ori partners with Kao Data for UK AI cloud
Also Read: SoftBank acquires Graphcore, boosting UK AI chipmaker’s future
Why it’s important
The SMF’s critique raises concern about the UK’s strategy for embedding AI across government departments. It highlights how the £3.25 bn AI Transformation Fund lacks strategic coherence and end‑to‑end clarity. Spending on redundancy and admin tech suggests that AI adoption may not be the primary objective. The wide disparity in readiness between departments reveals it. Without clear leadership and design, pilot projects risk remaining isolated. That could limit potential benefits such as efficiency, improved public services, and cost savings.
Additionally, traditional budgeting frameworks may delay or overcomplicate funding release. These systems are misaligned with AI’s agile and iterative development cycles. This mismatch may impede innovation and scalability. With the government targeting up to £45 bn in productivity gains and planning to automate significant civil‑service tasks, the lack of direction may threaten intended outcomes. Policy coherence and rapid iteration are vital to government AI deployment.
The SMF’s findings may prompt reviews of fund strategy and governance. They also reinforce calls for clearer and faster AI funding models aligned to technological cycles. The outcome will likely shape AI’s role in public services delivery, efficiency, and trust. As AI expands across the civil service, a robust roadmap and clear milestones are essential to avoid value slippage and deliver impact.
