UK Councils Risk Poor ROI on AI Without Service Design

What happened

New research published by a local government technology body has found that UK councils are at significant risk of failing to realise the returns they expect from investments in digital transformation and artificial intelligence. The study warns that without adequate change management, staff capability-building and a clear focus on resident outcomes, technology deployments are likely to underdeliver — or be abandoned altogether.

The research identifies a persistent gap between the ambition councils express in their digital strategies and the operational reality on the ground. Procurement of new platforms and AI tools is accelerating, but the organisational conditions required to embed those tools — process redesign, workforce readiness and leadership alignment — are frequently absent at the point of deployment.

Why it matters

For anyone designing or managing public-facing services, this finding is a pointed reminder that technology is never the transformation itself — it is an enabler of transformation that has already been designed. Local government is one of the highest-volume, lowest-discretion service environments that exists: residents cannot simply switch to a competitor when a council's digital channel fails them. That asymmetry makes the cost of a poorly implemented AI rollout far greater than a wasted budget line; it erodes trust in public institutions and pushes vulnerable residents back toward expensive, high-friction contact channels.

From a behavioural economics perspective, the research points to what practitioners would recognise as an optimism bias at the institutional level — decision-makers systematically overestimate the benefits of new technology while underestimating the friction involved in changing how people actually work. The ROI gap is not primarily a technology problem; it is a human-behaviour problem dressed in a procurement process.

The Renascence take

The conversation around public-sector AI tends to get captured by the technology vendors and the procurement teams, leaving the service-design and behaviour-change disciplines as afterthoughts. That sequencing is precisely backwards, and this research confirms it.

Most councils will read this finding as a call to invest more in training programmes or change-management consultants — and that instinct, while not wrong, still misses the deeper point. The ROI from digital transformation is unlocked at the moment of service encounter, not at the moment of system go-live. A customer-obsessed operator would start by mapping the resident journey in forensic detail, identifying where friction, anxiety or distrust already exist, and only then ask which technology — if any — removes those barriers. AI deployed into a broken process produces a faster broken process. The discipline required here is service design first, technology selection second, and change management woven throughout — not bolted on at the end.