Leverloft Free audit
Figures

Guides

How many UK businesses use AI? The 2026 figures, in plain English

About a third of UK businesses with 10 or more staff now use some form of AI. Most use it lightly, few have trained their people, and very few have cut jobs because of it. Here's what the Office for National Statistics found.

By Leverloft · Published 8 October 2026 · 5 min read

In short
  • About 35% of UK businesses with 10 or more employees used at least one AI technology in June 2026, up from about 12% in late 2023.
  • Use is shallow: the average adopter uses 1.6 kinds of AI, and only 10% of adopters say they use it extensively.
  • 58% of information and communication businesses use AI. In construction it's 13%.
  • Lack of expertise and cost are the most cited barriers. Earlier ONS work also found businesses struggle to see where to use it.
  • Only 11% of businesses have trained more than half their staff. 63% of businesses using AI to improve operations report no change in headcount.

Where these figures come from

The Office for National Statistics asks UK businesses about AI in its Business Insights and Conditions Survey. The figures below come from its article published on 20 July 2026, which covers the survey run from 5 to 28 June 2026 and compares it with late 2023.

Unless we say otherwise, they cover businesses with 10 or more employees. The ONS describes these as official statistics in development, so treat them as a good guide rather than a precise count.

How many businesses use AI

About 35% of businesses with 10 or more employees reported using at least one AI technology in June 2026. In late 2023 it was about 12%, so use has almost tripled. An earlier ONS bulletin, for late March 2026, put it at 26% of businesses.

By size

Bigger businesses are more likely to use it: 49% of businesses with 250 or more employees, against 28% of those with 0 to 9.

By industry

The gap between industries is wider than the gap between sizes. 58% of businesses in information and communication use AI, against 13% in construction.

How deeply they use it

This is the more telling number. The average business that uses AI uses about 1.6 kinds of it, up only slightly from 1.4 in late 2023. Of the businesses using AI, only 10% say they use it extensively.

Staff use tells the same story. Only 15% of businesses say more than half their employees use AI in their daily work.

What kind of AI, and what for

Type of AIShare of businesses using it
Large language models (ChatGPT, Claude, Copilot and the like)18%
Making images and other visual content16%
Processing data with machine learning12%
Image processing6%
Robotics2%

Improving business operations is the most common use in every size band: close to 60% of businesses using AI say so. Providing or personalising products and services comes second.

What holds businesses back

41% of businesses said nothing had prevented or delayed them adopting AI in the previous three months. Among those that did name a barrier, lack of expertise was one of the most common; around 18% of businesses with 100 to 249 employees said so. Cost was named by around 7% to 14% of businesses, depending on size.

Earlier ONS analysis, which the article refers back to, found the most common reasons were difficulty identifying where to use AI, cost and a lack of expertise.

What it has done to jobs and training

So far, very little to headcount. Of the businesses using AI to improve operations, 63% reported no change in the number of staff, 6% a fall and 1% a rise. The ONS reads this as AI changing tasks within roles rather than replacing them.

Training lags well behind. Only 11% of businesses said more than half their workforce had received any AI-related training. Firms that named lack of expertise as a barrier were far more likely to be training or retraining their own staff, at about 62%.

Where firms do see an effect on roles, it's mostly admin and clerical work, creative and design work, and data analysis.

What this means for a small business

  • You're not late. About two thirds of businesses with 10 or more staff aren't using AI at all, and most of the rest use it lightly.
  • The gap is skills, not software. The tools are cheap and widely available. What's missing is people who know which jobs to give them and how to set them up.
  • Start with a named job. "Difficulty identifying use cases" is a polite way of saying nobody had time to work out where to start. Pick the most repetitive weekly task and begin there.
  • Train the people who do the work. The businesses getting somewhere train their own staff rather than hiring specialists.

Sources

  1. Office for National Statistics, Artificial intelligence in UK businesses: 2023 to 2026 (20 July 2026)
  2. Office for National Statistics, Business insights and impact on the UK economy: 2 April 2026

We checked every figure and rule above against these sources by hand on 8 October 2026. Plans and prices change often. If something here is out of date, tell us at hello@leverloft.com.

Free audit

Start with a free audit.

Tell us what your team does by hand. Within three working days you get a one-page note: the first job, the hours it should save, a six-week plan and the price.

A sentence or two is plenty. It makes our estimate honest. Not sure which job?

We use your details to reply, and for nothing else. No mailing list.

  1. 1

    We read what you've told us, then look at your site, your ads and how enquiries reach you. If we need more, we email. A call is optional.

  2. 2

    You get the note by email, written by the person who'd run your sessions. It's yours to keep.

  3. 3

    If you go ahead, we agree a start date and rehearse the week before.

See a sample note for a made-up business