STATISTICS / UPDATED JUNE 2026

UK small business AI statistics

As of June 2026, 54% of UK SMEs are actively using AI (British Chambers of Commerce), 23% of all UK businesses use some form of AI technology (ONS), and the median adopter spent £2,000 on it in a year against a mean of £19,000 (UK government research). Every number on this page links to the organisation that measured it. No vendor surveys, no extrapolation, no rounding up.

Every figure re-verified against its primary source quarterly. Last verified June 2026.

SMES ACTIVELY USING AI, 2024 TO 2026
25%2024
35%2025
54%2026

Source: British Chambers of Commerce / Atos, Future of Work, March 2026. The first row of the table below links to it.

01 / 06

How many UK small businesses use AI?

The two headline numbers disagree because they measure different things. The BCC surveys SMEs and asks about any active use, including generic chatbots. The ONS asks all businesses about AI technologies in use. Both series are climbing fast.

StatisticFigureDetailSource and date
UK SMEs actively using AI54%Up from 35% in 2025 and 25% in 2024, the fastest jump the BCC has recordedBritish Chambers of Commerce / Atos, Future of WorkMarch 2026
SMEs running bespoke AI systems~1 in 10Only around one in ten firms report AI built into workflow automation, operations or customer service rather than generic toolsBCC / Atos, Powering Productivity reportMarch 2026
All UK businesses using AI technology23%Up from 9% when the ONS first asked the question in September 2023ONS Business Insights and Conditions Surveylate September 2025
Large businesses vs micro businesses using at least one AI technology36% vs 14%The firms with the least spare admin capacity adopt the tools that remove admin at the slowest rateDSIT AI Adoption Research (3,500 business interviews)February to May 2025

02 / 06

Which sectors use AI most?

Adoption splits hard by industry. Sectors whose work is documents and data adopted early. Sectors whose work is physical have barely started, and most of them are not planning to.

StatisticFigureDetailSource and date
Information and communication businesses using AI43%The most AI-heavy sector in the UK government research, against 16% of businesses overallDSIT AI Adoption ResearchFebruary to May 2025
Finance and real estate businesses using AI21%The next tier of adoption sits in finance and professional work: the more of the job that is text, the earlier the adoptionDSIT AI Adoption ResearchFebruary to May 2025
Construction firms neither using nor planning to use AI88%Transport and storage sits at 90%, hotels and catering at 88%, retail and distribution at 86%DSIT AI Adoption ResearchFebruary to May 2025
B2B service firms vs B2C firms using AI46% vs 26%Accountants, lawyers and agencies adopted first because their work looks like documentsBritish Chambers of Commerce SME surveySeptember 2025

03 / 06

What do UK businesses spend on AI?

The spend distribution is the single most useful budgeting fact in this table: most adopters spend a little, a small minority spend a lot, and the mean is pulled nine times higher than the median by that minority.

StatisticFigureDetailSource and date
Median annual AI spend by adopters£2,000Against a mean of £19,000 in the same year, 2024DSIT AI Adoption Research2024 spend, surveyed 2025
Mean annual AI spend by adopters£19,000A minority of businesses making a deliberate, larger bet pulls the average far above the typical spendDSIT AI Adoption Research2024 spend, surveyed 2025
Businesses that spent nothing on AI in 202431%A large slice of adoption starts on free tiers and bundled tools before any budget line appearsDSIT AI Adoption Research2024 spend, surveyed 2025
Businesses reporting AI spend grew over the last five years85%With 73% saying it had increased by more than 20%DSIT AI Adoption Research2025
Planned adopters with no specific AI budget47%Among businesses planning to adopt within twelve months. Buyers with no anchor get quoted whatever the seller thinks the market will bearDSIT AI Adoption Research2025

04 / 06

What stops UK small businesses adopting AI?

Cost is real but it is not the whole story. Not knowing what to point AI at beats price as the most common barrier, and the most heavily weighted concern is ethics, not money.

StatisticFigureDetailSource and date
Businesses rating ethical concerns a significant barrier80%The most widely rated significant barrier among businesses considering adoptionDSIT AI Adoption Research2025
Businesses rating high costs a significant barrier76%Second only to ethical concerns, with unclear or uncertain regulation close behind at 72%DSIT AI Adoption Research2025
Non-adopters who see no identified need for AI71%With 60% also citing limited skills. The gap is use-case knowledge, not appetiteDSIT AI Adoption Research2025
Current AI users hindered by limited skills and expertise54%The blocker does not disappear after adoption: over half of active users say skills limit how far they can push itDSIT AI Adoption Research2025
Firms citing difficulty identifying use cases as a barrier39%Ahead of cost at 21% in the ONS management practices analysisONS, Management practices and AI adoption in UK firmsMarch 2025

05 / 06

What do UK small businesses actually use AI for?

Text first, operations later. The overwhelming majority of adoption starts with writing and research, which is why depth, not adoption, is now the differentiator.

StatisticFigureDetailSource and date
AI-using businesses applying it to natural language processing and text generation85%The chat assistants are the form of AI most UK businesses adopt firstDSIT AI Adoption Research2025
SME adopters using AI for automating tasks54%Followed by marketing and advertising at 45%, customer service at 31% and logistics at 28%YouGov poll of 1,000 UK SME decision-makers2024
AI users applying at least some human checking to outputs84%And 67% apply significant checking. Human review is the norm, not timidityDSIT AI Adoption Research2025

06 / 06

What return do UK businesses report on AI?

Productivity arrives first and shows up in most firms quickly. Revenue takes longer, and it mostly arrives for the businesses that push adoption deepest. The pattern in the data matches what we see in practice: AI removes cost and time before it adds sales.

StatisticFigureDetailSource and date
Adopters reporting workforce productivity improvements75%And 57% had developed new or improved processes or operationsDSIT AI Adoption Research2025
AI-using businesses reporting a rise in overall employee productivity56%A measured rise, not a feeling: the survey asked businesses to put a number on the changeDSIT AI Adoption Research2025
Adopters who had not yet seen any change in revenue77%Just 12% reported an increase. The return shows up in time and cost long before it shows up in salesDSIT AI Adoption Research2025
Businesses with over half their staff using AI that reported increased revenue19%Against 12% of adopters overall. Depth of adoption, not the fact of it, is what eventually moves revenueDSIT AI Adoption Research2025
SMEs using generic AI reporting no impact on headcount95%Generic tools help existing staff move faster rather than restructure anything. Bespoke adopters are the group reporting structural changeBCC / Atos, Future of WorkMarch 2026

HOW THIS PAGE IS MAINTAINED

Primary sources only, or it does not go in

The web is filling with AI statistics that trace back to nothing: a model invented a percentage, a content farm published it, fifty blogs repeated it. This page exists to be the opposite. Four sources supply every row: the British Chambers of Commerce Future of Work research with Atos, the ONS Business Insights and Conditions Survey, the UK government's AI Adoption Research run by DSIT with 3,500 business interviews, and YouGov's SME panel.

Once a quarter we re-open every source linked on this page and check each figure against it. When a series updates, the table updates and the date at the top changes. The last full pass was June 2026.

Quote any figure here freely, with a link back to the original source or to this page. If you spot a fresher official number than the one we show, email dc@operosus.com and we will update it ahead of schedule.