Government · UK · Office for National Statistics
UK Business AI Use Nearly Triples to 35% Since 2023
Artificial intelligence use among UK businesses has almost tripled since 2023, but new ONS figures show most firms are still only scratching the surface, running an average of 1.6 AI tools each.
Source published Summarised
Key findings at a glance
- 35% of UK businesses with 10 or more employees reported using at least one AI technology in June 2026
- 12% used AI in late 2023, meaning adoption has almost tripled in under three years
- 58% of information and communication businesses use AI, the highest of any UK industry
- 13% of construction businesses use AI, among the lowest adoption of any sector
- 18% of businesses use large language models, the single most common AI technology
- 16% use AI for visual content creation, the second most common technology
- 60% of larger businesses use AI mainly to improve existing operations, the top reported purpose
- 55% of employees report using AI for work or education, well ahead of the 35% of businesses that report it
- 53% of businesses using AI image processing report effects on creative or design roles
- 1.6 AI technologies used per adopting business on average, up only modestly from 1.4 in 2023
- 41% of UK businesses with 10 or more employees report no barriers at all to adopting AI
- 62% of firms citing an AI expertise gap respond by training or retraining their existing staff
What the report covers
The Office for National Statistics published Artificial intelligence in UK businesses: 2023 to 2026 on 20 July 2026, its most comprehensive look yet at how British firms are actually using AI. Drawing on the Business Insights and Conditions Survey (BICS), it tracks adoption from September 2023 through to June 2026, and adds new questions on how deeply AI is embedded, which roles it affects, and what is holding firms back.
The headline is growth with a caveat. The share of businesses with 10 or more employees using at least one AI technology has risen from around 12% to around 35% in under three years. Yet the average adopter runs just 1.6 AI technologies, barely up from 1.4 in 2023, so most firms are experimenting at the edges rather than rebuilding around AI. This picture builds on our earlier report on AI use among UK businesses, adding depth, sector detail and labour-market signals.
Adoption is deeply uneven. Almost three-fifths of information and communication firms use AI, against just 13% in construction, and larger companies are far more likely to have adopted it than the smallest.
What this means for UK marketers
For UK marketers, this is a map of where the real opportunity sits. Adoption is wide but shallow: a third of businesses use AI, yet most run only one or two tools and just 10% describe their use as extensive. That gap between having AI and using it well is exactly where agencies and in-house teams earn their keep, by turning a free chatbot licence into a repeatable workflow that moves a metric.
The sector split is a targeting signal. Information and communication firms are saturated at 58%, so the differentiator there is sophistication, not adoption. Construction, at 13%, and other low-adoption trades are greenfield: the brand that shows up with a genuinely useful AI-assisted service, from faster quoting to smarter follow-up, faces far less noise. If you sell to businesses, segment your messaging by how AI-mature the buyer's industry actually is rather than assuming everyone is at the same stage.
The labour-market findings matter for anyone producing content. Over half of firms using visual-content AI report effects on creative and design roles, and image-processing AI is reshaping admin and data work. This is not a signal to cut creative headcount, it is a signal to redeploy it: the value is shifting from production to judgement, brief-writing, brand consistency and editing the machine's output. That mirrors what the government's AI adoption research found, that efficiency is the main motive, not headcount reduction.
Most usefully, the barriers are addressable. A lack of expertise, not cost or regulation, is the leading blocker, and 62% of firms that hit it respond by training their own people. For a marketing partner, that is the offer: not another tool, but the skills and process to use the tools already on the desk. If you want help building that capability into your marketing, that is the heart of our digital marketing services. It also pays to watch public sentiment, which remains wary, as we covered in how the UK public views AI.
How they got the data
The figures come from the Business Insights and Conditions Survey (BICS), the ONS's fortnightly study of UK businesses and the most comprehensive timely source on business AI use. The estimates in this article are drawn from Wave 159, fielded in June 2026, to which 38,637 businesses responded, a response rate of 26.7%. BICS shares its sampling frame with the Monthly Business Survey but excludes some industries, including agriculture, energy generation, public administration, and finance and insurance, so the figures describe most but not all of the economy.
AI is measured not as a single concept but as a list of specific technologies, with the headline figure counting any firm that reports using at least one. New questions in Wave 159 on the depth of use and the share of staff using AI are early indications rather than settled trends. Some estimates exclude firms with fewer than 10 employees, whose large survey weights can distort aggregate results. For wider labour-market context, the ONS points to the Bank of England's Decision Maker Panel, which likewise found little AI impact on headcount to date.
Government · UK
Read the full report
From Office for National Statistics · published 20/07/2026