Government needs central coordination to turn AI ambition into action
UK government has published strategy documents and committed £2 billion to AI adoption across the public sector. The real work starts now: turning that commitment into working systems that serve citizens and save money.
The risk is clear. Without coordinated support from the centre, individual departments and councils will solve the same problems separately, waste resources, and progress will stall. Government has done this before - and succeeded.
The digital precedent
In 2010, the Martha Lane Fox report fired the starting pistol on public sector digitisation. Government Digital Services emerged as a central function that set standards, shared tools, and built communities of practice. Good ideas stopped staying siloed.
AI needs the same structure. The AI Opportunities Action Plan (January 2025) and AI Playbook (February) are steps in the right direction. But guidance and strategy documents alone won't get the job done.
What organisations actually face
Government bodies making AI decisions confront a long list of hard questions: how fast to move; how to align AI strategy with organisational goals; which use cases matter most; whether their data and systems are ready; how to avoid vendor lock-in; how to keep control of their own systems.
The hardest part isn't technical. People and culture account for at least 70 percent of transformation work. Technology and data make up the rest.
Each organisation is starting from a different place. Some departments are years ahead. Many councils are just beginning. Without central tools and shared knowledge, the slower ones will take longer and spend more.
Where the work is happening now
The National Audit Office recently developed data and AI strategies aligned to its five-year plan, positioning itself to audit government AI spending in the years ahead.
Large departments are working through the ethical and technical challenges of putting AI into citizen-facing services. HMRC's trade team is exploring how AI can help small and medium-sized businesses navigate tariffs and trade guidance - work that removes real barriers to commerce.
Local councils are deploying AI across adult social care, housing, planning, and customer services. The Local Government Association research shows the potential is real. So is the risk of fragmentation if each council builds its own solution.
What comes next
Government has the technology. It has the ambition. What's missing is the connective tissue: shared standards, central tools, and active knowledge exchange.
The GDS model proved that common standards and shared assessment frameworks let good ideas spread quickly. A council in the North East should be able to adopt a solution pioneered by a Whitehall department without starting from scratch.
Central government should coordinate communities of practice, publish guidance, develop standards, and build tools for everyone to use. That's how individual experiments become system-wide change.
This work has been done before. It can be done again.
Learn more: AI for Government and explore the AI Learning Path for Policy Makers.
Your membership also unlocks:
