The AI adoption gap: Britain is buying AI faster than it is training people to use it
UK businesses are adopting AI at record pace. Yet 97% of organisations report an AI skills gap, and two in three employees have never had a single hour of formal AI training. This is the gap that decides who actually gets value from AI, and it is a training problem, not a technology one.
A TESS Group analysis of official UK data (ONS, DSIT, British Chambers of Commerce), June 2026.
Across every official measure, AI use in UK workplaces is rising fast. The disagreement is only about how fast. What every source agrees on is the second half of the story: the skills to use AI well are not keeping up. We call the distance between the two the AI adoption gap, and closing it is the single most useful thing most employers can do this year.
The numbers that matter
Six figures that define the UK AI adoption gap in 2026. Every one is sourced; full references are at the foot of the page.
Hold two of these side by side. More than half of SMEs now use AI, but a third of all organisations say their skills gap is already hurting their ability to hit business goals.3 The tools have arrived. The capability has not.
The AI Adoption Ladder
To make the gap concrete, here is the framework we use with employers. Most UK organisations are stuck on the same rung, and it is not the one they think.
Watching from the sidelines
No AI in use. Aware of it, waiting to see. A shrinking group as adoption spreads.
Shadow AI
Individuals quietly use ChatGPT and similar tools off their own back. No strategy, no governance, no shared learning.
Tools rolled out, people self-taught Most UK firms are here
AI tools are officially available, but staff are left to teach themselves. This is where adoption statistics look healthy and real value stays low.
Trained teams building
People are properly trained to design, build and govern AI in real workflows. Value compounds. This is the rung the skills gap blocks.
AI-native
Building capability is continuous and organisation-wide. AI is part of how work is designed, not a tool bolted on.
Why the gap is not closing on its own
Most organisations are trying to cross it with the wrong vehicle. 88% still rely on informal, on-the-job learning rather than structured training.3 That is why 83% of people can claim daily AI use while fewer than 5% use it in advanced ways.5 Casual exposure produces casual users. Building real capability, the kind that reaches Rung 3, takes structured, applied training against real work.
The stakes are not small. AI could add up to £400 billion to the UK economy by 2030,5 and the number of jobs directly involving AI is projected to rise from 158,000 in 2024 toward 3.9 million by 2035.5 The organisations that build capability now compound an advantage; the ones that stay on Rung 2 watch the value land elsewhere.
What actually closes it
The data points to one answer that is both structured and affordable: apprenticeships. They have moved from the margins to the mainstream of AI hiring, from 3% of AI hires in 2020 to 19% in 2025,5 because they do the one thing casual training cannot, they build capability against real workplace projects over time.
The funding now backs it too. Under the 2026 Growth and Skills Levy, training is levy-funded for larger employers, and from August 2026 the government fully funds apprentices under 25 for smaller employers. A new Level 4 AI apprenticeship is open to all sectors, with the first cohorts starting in 2026.6 The barrier was never really cost. It was knowing this is the lever.
Methodology and a note on the next edition
This first edition of the TESS AI Adoption Index synthesises the most recent official and published UK data on AI adoption and skills, from the Office for National Statistics, the Department for Science, Innovation and Technology, and the British Chambers of Commerce, alongside TESS Group's own published delivery data. Adoption figures vary by source because each uses a different definition of "using AI"; we present the range rather than a single number, and cite every figure below.
The AI Adoption Ladder is a TESS Group framework. The next edition will add primary research: a survey of UK employers and a survey of our own learner base, to put first-party numbers against where organisations sit on the ladder and what moves them up it. If you would like to take part, get in touch.
Frequently asked questions
What is the AI adoption gap?
The AI adoption gap is the distance between how widely organisations have adopted AI tools and how well their people can actually use them. In the UK in 2026, AI adoption is rising fast (the ONS puts business use at 21%, the British Chambers of Commerce reports 54% of SMEs), yet 97% of organisations report an AI skills gap and two in three employees have had no formal AI training. The tools are in place; the capability is not.
How many UK businesses use AI?
It depends on the definition. The Office for National Statistics found 21% of UK businesses using AI by mid 2025 (36% of firms with 250 or more staff). The government's DSIT research, using a tighter definition, reports about 16%. The British Chambers of Commerce, surveying SMEs in 2026, reports 54%. The figures differ because each measures "using AI" differently, but all show a steep upward trend.
What is the AI Adoption Ladder?
It is a five-rung framework from TESS Group describing how organisations mature with AI: Rung 0 Curious, Rung 1 Experimenting (shadow AI), Rung 2 Adopting (tools rolled out, people self-taught), Rung 3 Embedding (trained teams building), and Rung 4 Transforming (AI-native). Most UK organisations sit on Rung 2. The valuable, and hardest, move is from Rung 2 to Rung 3, which is exactly where the skills gap bites.
How do apprenticeships close the AI skills gap?
Apprenticeships build capability against real workplace projects over time, rather than through one-off courses, which is why they have grown from 3% of AI hires in 2020 to 19% in 2025. They are levy-funded for larger employers and, from August 2026, fully government funded for under-25s in smaller employers, so cost is rarely the barrier. They move people and organisations from owning AI tools to building with them.
Sources
- ONS, Business Insights and Conditions Survey (AI adoption, 2025): ons.gov.uk
- DSIT, AI adoption research (businesses with 5+ employees): gov.uk
- British Chambers of Commerce, "Powering Productivity: AI and the Future of UK Work" (March 2026): britishchambers.org.uk
- British Chambers of Commerce, "Half of SMEs Using AI" (March 2026): britishchambers.org.uk
- Sector and economic projections (AI economic value and AI-related jobs to 2035): DSIT and published 2026 research.
- UK apprenticeship and skills policy (Growth and Skills Levy, new Level 4 AI apprenticeship, Skills England priorities): gov.uk
Figures are drawn from the sources above and reflect the most recent data available at publication. Definitions of "AI use" differ between surveys; where they conflict we present the range. TESS Group delivery figures (Ofsted Good, 72% distinction rate) are from our own published performance data.
Find out which rung your organisation is on
A 30-minute call to map your team against the AI Adoption Ladder and the levy-funded route from Rung 2 to Rung 3. No pitch, just a clear read on where you stand.
Published 19 June 2026 by The TESS Group. The AI Adoption Index is updated annually.