AI, Automation and the Great Skills Transition: Why the Public Sector Must Lead a National Retraining Coalition

Artificial intelligence is no longer a distant horizon it is already embedded in the day‑to‑day machinery of public service delivery. Across councils, NHS trusts, housing associations, charities and combined authorities, AI is beginning to reshape how organisations operate, how services are delivered, and what work looks like. The UK public sector the country’s largest employer, with more than 5.9 million workers sits at the centre of this transformation. The question is no longer whether AI will reshape the workforce, but how we respond.

Recent analysis commissioned by Google Cloud found that AI could unlock billions of pounds in value for the UK public sector by 2030, improving productivity, reducing administrative burden and freeing up capacity across frontline services. The same study estimates that over 56% of public sector roles could be complemented by generative AI, particularly through automation of paperwork, data analysis and routine reporting tasks.

Yet the upside comes with profound risk. The Institute for Public Policy Research (IPPR) warns that up to 8 million UK jobs are at risk if AI adoption is unmanaged with administrative, clerical and routine roles among the most exposed. These roles are disproportionately concentrated in local government, the NHS and the wider public sector. Women and young people are particularly vulnerable because they are overrepresented in these job families.

The Alan Turing Institute’s 2025 analysis of public sector time‑use data reinforces this picture, identifying significant potential for AI to automate or augment routine administrative tasks across government, policing, education and health services. The National Audit Office has similarly projected that AI could deliver productivity gains “worth billions” across the public sector, but only if the right skills, data infrastructure and governance are in place.

The stakes could not be higher. We cannot afford a repeat of the 1980s and 1990s, when industrial decline devastated communities across the North East, South Wales, the Midlands and Scotland. In those decades, the collapse of coal, steel and manufacturing left deep scars: long‑term unemployment, intergenerational poverty, and the hollowing out of local economies. The North East where this podcast is proudly rooted still carries the legacy of those decisions.

This time, the disruption will not be confined to coalfields or shipyards. It will cut across administrative, clerical and routine roles in every part of the economy. Local government, the NHS, housing associations, charities and combined authorities all rely heavily on these functions.

The challenge is clear: how do we retrain and redeploy millions of workers at the pace and scale required to match technological change? And how do we do it collaboratively, across institutions that have historically operated in silos, with fragmented budgets, competing priorities and inconsistent national policy direction?

This is not simply a workforce challenge. It is a national economic challenge, a regional equality challenge, and a social justice challenge. It is also a test of whether the UK can finally deliver on the promise of devolution: shifting power, funding and capability away from Whitehall and into the hands of local leaders who understand their communities.

This is the conversation we’ll be opening in 2026 on The Truth About Public Sector podcast a conversation about the future of work, the future of place, and the future of public service. But here is the starting point.

The Scale of the Disruption

Artificial intelligence is not simply automating isolated tasks it is reshaping entire job families and redefining what “work” looks like across the UK economy. The shift is already measurable, and the evidence points to a labour market undergoing rapid structural transformation rather than a slow, generational evolution.

Recent modelling from the Institute for Public Policy Research (IPPR) paints a stark picture of what unmanaged AI adoption could mean for the UK workforce. Their analysis finds that up to 8 million jobs are at risk if the country fails to intervene strategically. This is not a hypothetical scenario: IPPR’s modelling covers 22,000 tasks across every job type, making it one of the most comprehensive assessments of AI exposure in the UK to date.

IPPR identifies two distinct waves of AI adoption:

• The first wave already underway exposes 11% of tasks across the economy. These are primarily “routine cognitive” tasks such as database management, scheduling, inventory management and basic information processing.

• The second wave deeper integration of AI into organisational systems exposes up to 59% of tasks. This includes more complex, non‑routine cognitive tasks and begins to affect higher‑earning professions as well.

The roles most exposed in the first wave are overwhelmingly concentrated in the public sector and its partner organisations. IPPR highlights that back‑office, entry‑level and part‑time jobs are at the highest risk, including:

• administrative roles

• secretarial and clerical work

• customer service

• routine data processing

These functions are heavily represented in local government, the NHS, housing associations and the wider public sector ecosystem.

The distributional impact is equally concerning. IPPR’s analysis shows that women are significantly more affected, because they are disproportionately represented in administrative and secretarial roles. Young people are also at heightened risk, as organisations reduce entry‑level hiring and automate early‑career tasks that traditionally served as stepping stones into the workforce.

The National Foundation for Educational Research (NFER) reinforces this picture with its long‑term labour market modelling. NFER forecasts that between one and three million low‑skilled jobs could disappear by 2035, largely due to AI and automation. The roles most vulnerable include:

• administrative and secretarial jobs

• customer service

• machine operation

• routine trades and basic service roles

NFER warns that these occupations are “declining at a much faster rate than previously predicted” and that the UK is entering a period of “profound labour market restructuring” that will not be felt evenly across regions or demographics.

This is not a slow, generational shift. It is a structural transformation already underway one that will reshape the labour market within a decade, not a lifetime. The question is no longer whether disruption will occur, but whether the UK will respond with the urgency, coordination and scale required to protect communities and create new pathways into meaningful work.

The Scale of the Opportunity

The same IPPR analysis shows that the future is not predetermined. With the right choices, AI could:

• Boost GDP by 13% (£306bn per year)

• Increase productivity across the public sector

• Create new roles in digital, green, care and technical sectors

• Avoid mass displacement entirely, if we choose augmentation over replacement

Skills England identifies huge growth in:

• digital technologies

• adult social care

• construction and housing

• engineering

• green industries

The UK will need:

• 725,000 clean energy workers by 2030

• 1.5 million new homes

• a significantly expanded NHS and care workforce

AI frees up labour. The question is whether we redeploy it or allow communities to be hollowed out again.

Why We Cannot Repeat the Mistakes of the Past

The UK has lived through one major economic transition already and we mishandled it. When the mines closed, when steelworks shut down, and when manufacturing moved overseas, the country failed to put in place the foundations that would have protected workers and communities. The consequences were not short‑term shocks; they were generational scars.

Deindustrialisation reshaped Britain’s economic geography. Entire regions including the North East, South Wales, the Midlands and parts of Scotland saw the collapse of industries that had sustained communities for more than a century. According to the Economics Observatory, the disappearance of coal, steel and shipbuilding contributed to higher rates of long‑term sickness, declining life expectancy and surges in regional economic inactivity, with effects still visible today. These impacts were not limited to those who lost their jobs; evidence shows they extended to their children and grandchildren, creating intergenerational costs that continue to shape life chances across former industrial areas.

Research from Sheffield Hallam University describes this as the “long shadow of job loss” a pattern of unemployment, economic inactivity and welfare dependency that persisted long after the factories and pits closed. In many towns, job growth never recovered to pre‑closure levels, and the labour market remains characterised by lower earnings, lower qualifications and fewer opportunities than in nearby cities Frontiers.

The reasons for this long‑term decline are well documented:

• No early warning closures were often sudden, with little time for workers or local authorities to prepare.

• No coordinated retraining large‑scale, accessible reskilling programmes were never put in place.

• No local economic diversification many towns were left with a single dominant employer gone and nothing to replace it.

• No long‑term investment in skills the UK’s skills system remained fragmented, underfunded and poorly aligned to local labour markets.

• No community‑led transition planning decisions were made in Whitehall, far removed from the people and places most affected.

The result was generational unemployment, poverty and regional decline. Former industrial areas today still experience higher rates of ill health, lower employment and persistent economic inactivity compared to the national average. And yet, the scale of disruption from AI could be even larger.

Where deindustrialisation hit specific geographies coalfields, steel towns, shipbuilding communities AI will affect every geography, because administrative, clerical and routine roles exist everywhere: in councils, hospitals, housing associations, charities, universities, and private sector employers across the country.

This time, the risk is not the collapse of one industry it is the transformation of entire job families.

But unlike the 1980s, we are not walking into this blind. We have foresight. We have modelling. We have labour market data. We know which roles are exposed, which communities are vulnerable, and which sectors will grow. The question is whether we act on that knowledge or repeat the mistakes of the past.

A National Skills Coalition: What Collaboration Must Look Like

No single institution can deliver a transition of this scale. AI is reshaping labour markets faster than any previous technological shift, and the UK’s current skills system fragmented, under‑funded, and overly centralised is not designed for rapid, economy‑wide retraining. What’s needed is a National Skills Coalition, bringing together local government, central government, the NHS, housing associations, charities, combined authorities, employers and communities.

This coalition must be built on five pillars.

1. Local Government as the Convenor

Local government and combined authorities are closest to labour markets. They understand the real‑world dynamics of local economies, employer needs, and community vulnerabilities. They are also the only institutions with the democratic legitimacy to convene employers, colleges, universities, training providers and anchor institutions. Skills England’s labour market assessments show that local variation in skills demand is significant, and that effective planning requires granular, place‑based data on job vacancies, training outcomes and future workforce needs.

Local government should:

• Map local skills exposure using real‑time labour market data, including dashboards now available through Skills England and the Department for Education.

• Coordinate Local Skills Partnerships, ensuring employers, colleges and training providers work to a shared plan.

• Align retraining with local economic strategies, particularly in growth sectors such as green jobs, health innovation, construction and digital technologies.

• Use devolved powers to shape training provision, ensuring it reflects local needs rather than national averages.

Some combined authorities such as Greater Manchester and the West Midlands are already doing this well. But national consistency is missing. Without a coordinated framework, the UK risks a postcode lottery in AI readiness.

2. Central Government as the Enabler

Whitehall’s role is not to micromanage skills delivery, but to create the conditions for local success. The CIPD’s analysis of UK skills policy highlights that centralisation has historically slowed innovation and reduced responsiveness to local labour market needs.

Central government must:

• Devolve funding flexibly, allowing places to design training that fits their economies.

• Provide long‑term certainty, replacing short‑term competitive pots with multi‑year settlements.

• Set national standards for AI literacy and digital skills, ensuring consistency across regions.

• Reform the apprenticeship levy into a true Growth and Skills Levy, enabling employers to use funds for modular, short‑course training not just traditional apprenticeships.

3. NHS, Housing Associations and Charities as Anchor Employers

Anchor institutions large, place‑based employers that are not going anywhere are essential to a just transition. They collectively employ millions and have the scale to absorb displaced workers.

Skills England’s 2025 assessment shows:

• The NHS employs around 1.5 million people in England alone.

• Adult social care employs 1.6 million people, with 76% in direct care roles.

• The sector faces chronic shortages, with vacancy rates in some areas exceeding 40% in key clinical and support roles.

These organisations can:

• Offer guaranteed interviews for retrained workers.

• Co‑design training pathways with colleges and providers.

• Provide real‑world placements to accelerate learning.

• Absorb displaced workers into care, support, housing and community roles.

Housing associations and charities often overlooked in national workforce planning are equally important. They employ hundreds of thousands in roles that require empathy, communication, digital capability and community engagement: skills that many displaced workers already possess.

4. A New Skills Infrastructure

The UK’s current skills system is too slow, too rigid and too qualification‑heavy to meet the demands of AI‑driven change. The Department for Education’s labour market projections to 2035 show that skills needs are shifting faster than traditional training routes can adapt.

We need a system built for speed, flexibility and accessibility. That means:

• 16‑week accelerated bootcamps for digital, green and technical roles.

• 8‑month fast‑track apprenticeships for critical sectors such as health, construction and engineering.

• AI “bolt‑on” training for existing workers, enabling rapid upskilling without career breaks.

• Foundation apprenticeships for those furthest from the labour market, providing stepping stones into growth sectors.

• 
  

Skills must be modular, stackable and portable, allowing workers to build capability over time rather than committing to multi‑year programmes.

5. Community‑Led Transition Planning

The people most affected by AI disruption must shape the solutions. Deindustrialisation taught us that top‑down transition planning fails when it ignores lived experience. Community‑led planning should include:

• Local transition boards, bringing together residents, employers, unions and civic leaders.

• Community skills guarantees, ensuring every displaced worker has access to retraining.

• Retraining hubs in libraries, community centres and housing estates, making learning accessible.

• Wraparound support childcare, transport, mental health recognising that barriers to training are often practical, not motivational.

Skills policy fails when it ignores the realities of people’s lives. A just transition requires dignity, agency and support.

What Success Looks Like

Success in this transition cannot be measured simply by productivity gains, GDP growth or the number of new digital jobs created. Those metrics matter but they are not enough. A successful transition must be judged by whether it strengthens communities, expands opportunity and protects the dignity of work.

A successful transition would mean:

• No community left behind not the former industrial towns that still carry the scars of the 1980s, not coastal communities, not rural areas, not the estates where opportunity has been rationed for decades. Every place must have a plan, and every plan must be backed by long‑term investment.

• No worker displaced without a pathway because displacement without direction is abandonment. Every person whose role is automated should have a guaranteed route into training, support and a new job. Not a leaflet. Not a website. A pathway.

• A workforce ready for AI‑augmented roles where people are equipped not just to survive technological change, but to thrive within it. Where AI becomes a tool for empowerment, not a threat to livelihoods. Where digital confidence is a universal skill, not a privilege.

• Local economies strengthened, not hollowed out with new industries, new skills pipelines and new opportunities rooted in place. Where the benefits of AI are captured locally, not extracted elsewhere. Where growth is shared, not concentrated.

• A public sector that leads by example modelling ethical AI adoption, investing in its people, and demonstrating what a just transition looks like. The public sector is the country’s largest employer. If it gets this right, the rest of the economy will follow.

It is a commitment to fairness, dignity and shared prosperity. It is a promise that technological progress will not come at the expense of the people and places that hold this country together. It is a recognition that the future of work is not predetermined it is shaped by the choices we make now.

And that is why this conversation matters. Why it belongs at the centre of public sector leadership. Why it will be one of the defining themes of The Truth About Public Sector in 2026.

The transition is coming. The question is whether we manage it or whether it manages us.