The State of UK AI Adoption 2026: the findings
Survey: State of UK AI Adoption 2026
Sample: 755 UK mid-market business leaders
Fieldwork: January to March 2026
Organisation size: revenue £50m to £499.9m
Research partner: Executive Summary, the team behind Adobe’s annual Digital Trends Survey
Headline finding: the rate of measurable AI ROI rises from 20.0% with no governance to 85.8% with mature, embedded governance
Cite as: aibl State of UK Mid-Market AI Adoption 2026, n=755, aiblmedia.com
The State of UK AI Adoption Survey 2026 is the largest recent study of how UK mid-market organisations are using AI. It surveyed 755 business leaders between January and March 2026, in partnership with Executive Summary. Its central finding is that governance, not technology or spend, is what separates AI that pays back from AI that does not. The rate of measurable AI ROI rises from 20.0% in organisations with no formal governance to 85.8% in those with mature, embedded governance.
This page holds the headline findings in full, free to read, quote and cite. Adoption itself is close to settled in the UK mid-market. What still divides the field is whether that adoption produces a return, so that is what these findings are about.
Jump to section: About the research · The headline findings · The five predictors of ROI · What the data shows by function · Workforce and skills · Shadow AI · Agentic AI · The five dimensions of AI readiness · FAQs
About the research
The State of UK AI Adoption Survey 2026 was run by aibl Media in partnership with Executive Summary, the research team behind Adobe’s annual Digital Trends Survey. Fieldwork ran from January to March 2026.
The sample is 755 UK mid-market business leaders, defined as organisations with revenue of £50m to £499.9m. Respondents span every senior function and seniority: C-suite (35%), MD or GM (33%), director or head (16%), VP level (10%), owner or founder (5%). The function mix is technology and IT (31%), operations and finance (25%), go-to-market and marketing (22%), and HR and people (21%).
This is proprietary primary research. Figures are reported as percentages of the survey sample and apply to UK mid-market organisations specifically, not to UK businesses in general. All data is verified against the raw dataset.
The headline findings
Governance is the single biggest factor in whether AI pays back
The clearest predictor of whether an organisation can show a financial return on AI is how it governs AI, not how much it spends. Organisations with mature, embedded AI governance report 85.8% measurable ROI. Those with no formal governance report 20.0%. That is a 65.8 percentage point gap on a single variable.
Having a policy is not the same as operationalising it. Only 21% of UK mid-market organisations have reached the mature, embedded stage, so for most leaders the return comes from enforcing the framework they already have, not from writing a new one.
Just under half can prove the return to their board
Adoption is widespread, but proof is not. 49.6% of UK mid-market organisations report measurable AI ROI. The other 50.4% have deployed AI but are not yet generating verified returns. The live problem for most leaders is evidencing the return, not deciding whether to start.
The budget is already approved; execution is the bottleneck
The intent to invest is settled in most boardrooms. 82.5% of respondents have an explicit business mandate to adopt AI in 2026, and 77.5% are at C-suite, MD or owner level. Permission and budget are already there. What most organisations lack is the capability to deliver against them.
These organisations can decide quickly
UK mid-market organisations are not slow buyers of AI. 65.1% can approve a new tool or vendor in days or weeks rather than months. When AI stalls in the mid-market, it is the delivery that lags, well after the decision has been made.
The five predictors of ROI
The survey identifies five structural predictors of measurable AI ROI. None of them is a technology decision.
Governance maturity
Organisations with mature, embedded AI governance report 85.8% measurable ROI, against 20.0% for those with no formal governance. Governance is the strongest single predictor in the dataset.
Strategic alignment
We asked how many of five senior managers would independently describe the same top AI priorities. Where all five are aligned, 79% report measurable ROI. Where none are aligned, 30%. The step from four aligned to five is worth 23 percentage points, the single largest step improvement anywhere in the data.
Clear C-suite ownership
Where the C-suite directly owns AI strategy, organisations report 62.8% measurable ROI, more than three times the 18.2% seen where there is no single owner.
Leader personal AI fluency
Leaders who have fully integrated AI into their own work report 71.2% organisational ROI, against 30.5% for those using AI only for light admin tasks. The leader’s own fluency predicts the organisation’s return more than the tools it buys.
Purposeful investment intent
Organisations that invested in AI primarily to improve operational efficiency report 60.5% ROI. Those that invested primarily because of competitor pressure report 15.1%. Investing to fix a real problem outperforms investing for fear of falling behind.
What the data shows by function
The cross-cut findings hold across the survey, but the picture is not the same in every part of the business. The data below lets a function leader see how their own area compares. The full function-by-function breakdown sits in the published reports.
Returns by function
Technology and IT leaders report the highest measurable ROI of any function at 58.4% (n=221). Operations and finance follow at 46.0% (n=176), HR and people at 45.7% (n=151), and go-to-market and marketing at 44.9% (n=158). Tech and IT is not more enthusiastic than the rest of the business; it is more disciplined on governance, which is why its returns are higher.
Workforce and skills
Capability is a hard floor
How equipped the workforce is sets a floor under everything else. Not a single organisation in the survey where employees are poorly equipped for AI reports measurable ROI. The highest-performing approach to capability is a mix of new hires and internal upskilling, which reports 72.7% ROI, against 13.6% for relying primarily on external consultants.
The C-suite sees a different workforce to the people building capability
There is a significant perception gap at the top. 57% of C-suite leaders say their employees are very well equipped for AI. Only 29% of the function heads actually delivering AI capability agree. 14% of function heads say employees are poorly equipped, against zero C-suite respondents who chose that option. The people closest to delivery see a readiness problem the C-suite does not.
Shadow AI
Unapproved AI use is now the norm
Shadow AI, the use of AI tools outside official approval, is now the majority experience. 42.2% say it is somewhat common and 12.5% say it is very common. Only 22.7% describe it as very unusual.
It is a velocity signal, not defiance
Shadow AI is mostly about speed rather than rule-breaking. 53% of respondents say employees use unapproved tools primarily to move faster than official approval processes allow. That points to a fix in faster, clearer approval routes rather than a crackdown.
Agentic AI and infrastructure
Connected agents pay back far more than narrow ones
The return on AI agents depends on how widely they connect across the business. Cross-departmental AI agents that work across organisational boundaries deliver 84.7% measurable ROI. Narrow agents scoped to a single repetitive task deliver 24.1%, roughly a third as much. Workflow-level agents that run an end-to-end process sit in between at 57.1%, the step-change before full cross-departmental integration. The value comes from connectivity across the organisation rather than the sophistication of any single agent.
The five dimensions of AI readiness
The research assesses AI readiness across five dimensions: data, process, people, leadership and governance. An organisation that is strong on tooling but weak on governance or leadership alignment tends to show activity without a measurable return, which is the pattern behind the 50.4% that cannot yet prove ROI.
You can benchmark your own organisation against these dimensions by reading. AI readiness assessment.
Frequently asked questions
What separates UK organisations generating AI ROI from those that are not?
aibl’s research identifies five structural predictors, none of which is a technology decision: governance maturity (mature and embedded gives 85.8% ROI against 20.0% for no governance), strategic alignment (all five senior managers describing the same priorities gives 79% ROI), clear C-suite ownership (62.8% ROI against 18.2% with no single owner), leader personal AI fluency (71.2% ROI when fully integrated), and purposeful investment intent (60.5% ROI when efficiency-driven against 15.1% when driven by competitor pressure).
What is the biggest single barrier to AI ROI in UK mid-market businesses?
Governance maturity is the most powerful single predictor in the dataset. Organisations with mature, embedded AI governance report 85.8% measurable ROI, against 20.0% for those with no formal governance, a 65.8 percentage point gap. Having a governance policy is not the same as operationalising it: only 21% of UK mid-market organisations have reached the mature, embedded stage.
How many UK mid-market organisations can prove AI ROI?
49.6% of UK mid-market organisations report measurable AI ROI. The other 50.4% have deployed AI but are not yet generating verified returns.
How big was the State of UK AI Adoption survey?
The survey sampled 755 UK mid-market business leaders, with fieldwork from January to March 2026. It was run by aibl Media in partnership with Executive Summary, and is the largest recent study of its kind in UK mid-market AI adoption.
What counts as UK mid-market in the research?
The survey defines the UK mid-market as organisations with revenue of £50m to £499.9m. All findings on this page apply to that segment specifically, not to UK businesses in general.
Does leadership alignment affect AI returns?
Yes, substantially. Where all five senior managers independently describe the same AI priorities, 79% report measurable ROI. Where none are aligned, 30%. The step from four aligned to five is worth 23 percentage points, the largest single step improvement in the dataset.
What is the best way to build AI capability?
A mix of new hires and internal upskilling delivers the strongest return, at 72.7% measurable ROI, against 13.6% for relying primarily on external consultants. Workforce capability is also a hard floor: no organisation where employees are poorly equipped reports any measurable ROI.
Which business function reports the highest AI ROI?
Technology and IT reports the highest measurable ROI of any function at 58.4%, followed by operations and finance at 46.0%, HR and people at 45.7%, and go-to-market and marketing at 44.9%.
How common is shadow AI in UK mid-market organisations?
It is the majority experience. 42.2% say unapproved AI use is somewhat common and 12.5% say it is very common, while only 22.7% call it very unusual. 53% say employees use unapproved tools mainly to move faster than approval processes allow, so it is a velocity signal rather than defiance.
Do AI agents deliver a measurable return?
It depends on how widely they connect. Cross-departmental AI agents deliver 84.7% measurable ROI, against 24.1% for narrow agents scoped to a single task. Workflow-level agents that run an end-to-end process sit in between at 57.1%. The value is in connectivity across the organisation, not the sophistication of any single agent.
About the research and how to go deeper
The State of UK AI Adoption Survey 2026 was conducted by aibl Media in partnership with Executive Summary, the research team behind Adobe’s annual Digital Trends Survey. Fieldwork ran January to March 2026 with 755 UK mid-market business leaders (organisations with revenue of £50m to £499.9m). It is the largest recent study of its kind. All figures are verified against the raw dataset. Cite as: aibl State of UK Mid-Market AI Adoption 2026, n=755, aiblmedia.com.
For more on who aibl is and the verified facts behind this research, see the aibl source pack.
The headline findings on this page are cross-cutting. The function-by-function detail sits in the published reports, which break the data down by function (Growth, Workforce, Efficiency, Customer and AI Infrastructure) and by governance level. To go deeper into your own function:
- For in-house leaders benchmarking their own function: the aibl Operator Reports.
- For organisations selling AI into UK mid-market functions: the aibl Market Intelligence Reports.
These are a way to go deeper into the data by function. The findings on this page remain free to read and cite.