The UK has a £400bn AI skills gap. Skills England just published what actually closes it.

On 10 June, Skills England published a report on what actually works in AI training, written by Dr Nisreen Ameen of Royal Holloway, University of London, and drawn from over 150 employers. The headline finding is the one every UK leader should sit with: around 44% of UK workplaces now use AI every day, but adoption is “uneven and often limited in impact.” The tools are everywhere. The value mostly isn’t. That gap, the government estimates, is worth £400 billion.

This is the same story we covered when the government put £200m behind AI adoption two days earlier: the bottleneck is not the technology, it is the skill to use it well. What is new here is that Skills England has now set out, in plain terms, what effective AI training looks like. And the moment you read its framework, the conclusion writes itself.

The report is the centrepiece of the SKAI programme (Skills for AI: What Works for AI Upskilling in the UK). It moves, in Skills England’s words, “from diagnosis to delivery”: not another warning about the skills gap, but evidence-based guidance on closing it, with an employer guide and case studies from organisations including LinkedIn, Airbus, Roche, KPMG and the NHS.

At its core is the PRIMES framework, six principles for AI training that works:

We are not claiming a coincidence. Skills England’s own report names the routes it complements: a Level 4 AI & Automation Practitioner apprenticeship “supporting practical workplace application”, new flexible AI leadership apprenticeship units “helping employers respond quickly to emerging needs”, and the free AI Skills Boost short courses. Those are the three things TESS delivers. But put PRIMES side by side with how a good apprenticeship actually runs, and the fit is almost line for line.

We did not design our programmes to match a framework that did not exist yet. It is the other way round: Skills England studied 150 employers and described what good AI training does, and it turns out good AI training is an apprenticeship. Practical, on the job, modular, governed. That is the model, and it is the one the levy already pays for. Lisa O'Reilly, Director, TESS Group

The report also ships three free tools every employer can use today: an AI Skills Framework (which skills which roles need), an AI Skills Adoption Pathway Model (the stages from awareness to strategic scaling) and an Employer AI Adoption Checklist (a structured readiness assessment). They are genuinely useful, and they pair neatly with our own free resources: the 2-minute AI readiness scorecard tells you which stage you are at, and the AI Opportunity Map template turns that into a prioritised plan.

The honest read of all of this: the government has now removed every excuse. There is a free baseline (AI Skills Boost), a free diagnostic (the new tools), a published account of what works (PRIMES), and a funded delivery route (the apprenticeships). What remains is the decision to start.

The most quietly alarming line in the report is that adoption is “often limited in impact.” A workforce that uses AI every day but badly is not a safe middle ground; it is a liability. It means staff pasting sensitive data into free tools, inconsistent outputs reaching clients, and no audit trail when something goes wrong. We hear it on calls constantly: teams “winging it” with whatever tool they found. PRIMES’s “Sustainable” principle, and the governance built into the apprenticeship, is the difference between AI that compounds value and AI that quietly accumulates risk.

Forty-four per cent using AI daily sounds like progress. But using it daily and using it well are different things, and the gap between them is where the £400bn is hiding. You close it with people who can design the workflow and govern the result, not with another licence. Rod Doyle, Director, TESS Group

The wider context

Three government moves in one fortnight tell a single story. The £200m AI Adoption Summit put money behind getting Britain to use AI. The leadership-apprenticeship defunding redirected the levy away from generic management toward AI-era skills. And now this report tells employers, with evidence, exactly what effective AI training looks like. All three point the same way: fund the people, on the job, with governance, and do it through the routes the levy already covers.

Sources & further reading

GOV.UK / Skills England: New tools will help employers maximise AI productivity gains (10 June 2026). Skills England: Skills for AI: What Works for AI Upskilling in the UK. TESS: the £200m AI Adoption Summit, what employers get.