Hampshire needs £179.7 million by 2028. Windsor and Maidenhead residents each owe £419.71 toward their own council’s shortfall. AI transcription is already saving social workers hours. What it can’t do, experts warn, is fill a hole this size.
English councils face a funding gap topping £4 billion over the next two years, according to the Local Government Association, and a growing number are turning to AI to help close it. A BBC Shared Data Unit study found 146 of 148 English councils have already signed off savings worth £2.86 billion, with 218 upper-tier authorities cutting £3.2 billion between them. Almost every upper-tier council in the UK will offer fewer services this financial year as a direct result. Jonathan Carr-West, chief executive of the Local Government Information Unit, didn’t soften the warning: “In just three years, councils will need around 25% more money simply to stand still.”
What’s Happening & Why It Matters
Council by Council
The numbers behind the headline figure vary dramatically by region. Hampshire County Council needs to find £179.7 million by the end of 2027/28. The Royal Borough of Windsor and Maidenhead carries England’s second-largest per-head budget gap, at £66.9 million — working out to £419.71 for every resident. West Northamptonshire Council identified a £24 million gap for the same period, while neighbouring North Northamptonshire needs £13.6 million. Both Northamptonshire authorities, controlled by Reform UK, have already introduced AI in adult social care — an area that’s seen consistent spending increases in recent years.
Barnet Council, one of the largest budget gaps in the UK relative to its size, will need close to £80 million in government support this year. A spokesperson described the council as a “tightly run ship” facing “the perfect storm” of a decade of reduced government funding colliding with year-on-year rises in service costs. Six of the ten councils making the deepest cuts per head in 2026-27 are London boroughs, adding to concerns that the capital is being hit hardest.
Where AI Is Helping
Oxfordshire County Council considers itself an early adopter, deploying AI in children’s and adults’ social care, including home-to-school transport planning. Councillor Tim Bearder, the council’s leader, described a concrete use case: “Routes can be complicated, perhaps a child with autism who can’t be in a taxi for more than 20 minutes so we need to plot journeys which AI can do.” That’s a genuinely narrow, solvable logistics problem — the kind AI handles well.
Transcription is where the technology has spread fastest. Suffolk County Council is rolling out AI transcription to every adult social care practitioner following a successful pilot, saving time on the detailed case notes social workers previously wrote by hand after every visit. Angus Council in Scotland signed a £240,000 contract for a similar AI note-taker. Not every deployment has gone smoothly, though — the Scottish authority awarded that contract without a tender, citing an absence of competition, after discussing the deal in a private session.
The Unsolved Governance Gaps
Speed of adoption is outpacing oversight in some cases, and two recent incidents illustrate exactly how. North Lanarkshire Council has admitted passing a resident’s details to a data centre developer following a planning enquiry—a data-handling failure, not a model performance problem. Neither incident reflects AI working badly. Both reflect the accountability and safeguard gaps that come with deploying new systems faster than governance structures can keep pace.
An eighteen-month study led by Heriot-Watt University and GoLLM, analysing 208 councils across England, Scotland, Wales, and Northern Ireland, found a genuinely mixed national picture. Some councils are moving quickly to build data foundations and develop internal AI capability. Many others remain at the earliest stages of basic digital infrastructure, with legacy systems and unclear guidance making it considerably harder to move beyond initial experimentation. Separately, new research found 84% of the UK public believes government prioritises tech companies over citizens on AI regulation. The trust gap widens every time a story like North Lanarkshire’s data handling failure makes headlines.
AI Won’t Close The Gap Alone
Every expert TF’s research surfaced on this story agrees on one point. AI will not solve the council cash crisis. Carr-West was direct: “Without action, the services people rely on every day, from social care to safe streets, will be eroded.” He noted Labour’s recent local government reforms — multi-year funding settlements that tell councils how much they’ll receive several years in advance, plus a reworked allocation formula directing more money toward areas with higher need and less local revenue-raising capacity — had raised some hope, but haven’t closed the underlying gap.
The practical reality is that AI is genuinely useful at the margins — faster transcription, smarter route planning, fewer hours spent on paperwork — while structurally incapable of replacing the billions in central government funding that’s disappeared over the past decade and a half.
TF Summary: What’s Next
Councils across England continue rolling out AI transcription and case-management tools through 2026, with adoption speed varying widely by authority. The Local Government Association‘s multi-year funding settlements, introduced under Labour’s local government reforms, continue phasing in. No confirmed government response has been announced to the £4 billion gap the LGA has flagged, beyond the funding formula changes already underway.
MY FORECAST: Expect the AI adoption pattern already visible — genuine efficiency gains in narrow, well-defined tasks like transcription and route planning, paired with real governance failures like North Lanarkshire’s data mishandling — to continue at roughly the same ratio through 2027. The £4 billion gap won’t close through technology; it will close, if it closes at all, through the funding formula reform Carr-West referenced, or it won’t, and services will keep eroding, as councils warn. Watch whether the Angus Council no-tender controversy triggers tighter procurement rules for AI contracts specifically, given how directly it exposes the accountability gap researchers have already flagged as the sector’s real weakness — not the AI itself, but the oversight structures meant to govern it.
Related Stories
- A Reverse Lookup Service Exposed Millions of People’s Faces
- Flock Admits It Should Have Stopped Police From Abusing Its Cameras
- GOLD EAGLE: White House Launches an AI Cybersecurity Clearinghouse