Why government must move beyond bolt-on AI

The UK public sector is accelerating its adoption of artificial intelligence, with organisations across central government, local authorities and the NHS exploring how the technology can improve services, automate work and increase efficiency.

But as investment and experimentation grow, an important question is emerging: how can public bodies ensure that AI adoption translates into meaningful improvements for citizens rather than simply more technology?

I believe the reality is that the public sector does not have an AI problem. It has an orchestration problem.

Activity is not the same as progress

On the surface, AI adoption across government appears healthy. From central government departments to local authorities and the NHS, organisations are piloting new tools, developing strategies and exploring opportunities to improve services through automation and intelligence.

Yet delivery continues to lag behind ambition.

There is a growing disconnect between AI activity and AI outcomes. While leadership confidence remains high, many organisations are still struggling to translate experimentation into meaningful transformation. Recent research found that only 29 percent of public sector workers believe their organisation is delivering on most of its AI commitments, despite widespread investment and enthusiasm for the technology.

This is not a failure of AI itself, but a failure of implementation.

Too often, organisations are mistaking experimentation for transformation. Deploying AI may be relatively straightforward, but transforming services with AI requires a far more fundamental rethink of how work gets done.

The limits of bolt-on AI

A significant part of the challenge lies in how AI is being introduced into public sector environments.

Nearly half of public sector AI initiatives are still being deployed as standalone tools or bolt-on technologies, rather than being embedded within the processes that underpin service delivery.

The appeal is understandable. Bolt-on solutions are quick to deploy, carry relatively low risk and provide visible evidence that organisations are embracing innovation. Whether it is a chatbot, a co-pilot or a standalone analytics platform, these technologies allow teams to demonstrate progress without fundamentally changing existing systems.

But this approach rarely delivers the transformational outcomes organisations are seeking. When AI sits outside the process, it cannot transform the process.

Instead, AI often inherits the inefficiencies, fragmentation and duplication that already exist. While individual productivity may improve, the wider organisation sees little change. The result is frequently a collection of disconnected tools with limited impact on citizen-facing services.

This may also help explain why many members of the public remain unaware of how government is using AI today. AI is being deployed across the sector, but the benefits are not yet being felt or communicated in ways that citizens can easily recognise.

Why process matters

The missing ingredient in many AI strategies is process.

AI delivers the greatest value when it is integrated directly into the flow of work, connected to the right data and given a clear role within a wider operational system. In that environment, decisions lead directly to actions, outcomes can be measured and improvements can be made continuously.

In an integrated process, AI is not an add-on. It is part of how work gets done.

This is particularly important in the public sector, where transparency, accountability and trust are essential. Citizens are not simply asking whether AI can make government more efficient; they also want assurances that decisions are fair, secure and properly governed.

Those expectations cannot be met through isolated tools alone. They require end-to-end processes that embed oversight, auditability and accountability from the outset.

Fix the process before scaling the intelligence

There is growing recognition across the public sector that successful AI adoption depends on strong foundations.

Many organisations are beginning to understand that AI is not a shortcut around broken processes. Instead, it amplifies whatever it is given. If the underlying workflow is inefficient or fragmented, AI will scale those problems rather than solve them.

Conversely, when processes are structured, connected and measurable, AI can unlock significant improvements in speed, accuracy and service quality.

That is why organisations should focus on process transformation and AI adoption together, rather than treating them as separate initiatives. Redesigning processes is often more challenging than deploying a new technology because it requires organisations to rethink how work flows across systems, teams and departments.

It requires coordination, not just capability. In short, it requires orchestration.

Moving from pilots to public impact

As government looks to accelerate AI adoption, I believe the next phase of progress will not be defined by new models or more pilot programmes alone.

Recent policy discussions have rightly focused on the need to accelerate adoption, but speed without direction risks delivering more activity rather than more value. The real challenge is ensuring AI is applied in ways that improve public services and deliver outcomes that citizens can see and experience.

That means moving beyond isolated use cases and towards integrated systems. It means designing AI into processes from the outset rather than adding it later. And it means measuring success not simply through internal efficiency gains, but through tangible improvements in public outcomes.

The public sector has already demonstrated its willingness to embrace AI. The more important question now is whether organisations are taking a sufficiently strategic approach to implementation.

Until AI is designed into processes rather than added onto them, it will continue to deliver pockets of value while falling short of the broader transformation that government and citizens expect.