The full figures

The numbers, checked

Everything behind the short answer on the front page — the real British adoption figures, read from the original government release rather than a summary of one, and updated every couple of months.

Last checked 30 July 2026 · next review 30 September 2026 · every figure carries its source

The short answer

If you’ve opened an AI tool a couple of times, you’re not behind. You’re exactly average — and so is almost everybody else.

The figure you’ll have seen quoted is that 35% of UK businesses now use AI. It’s true, it comes from the Office for National Statistics, and it is the only part anyone repeats.

Three paragraphs further down the same release is the number that actually describes the country.

What the same ONS release found · June 2026
35%of businesses use at least one AI technology — the figure everyone quotes
10%use it “extensively” across their operations
15%have more than half their workforce using it daily
11%have trained more than half their staff to use it
1.6AI technologies used by the average business — up from 1.4 in late 2023

ONS, Business Insights and Conditions Survey, Wave 159 · fieldwork 5–28 June 2026 · published 20 July 2026

Adoption looks like it tripled. Depth went from 1.4 technologies to 1.6 in nearly three years.

That is the whole picture, and it is the opposite of the one being sold to you. It isn’t that British business raced ahead while you were busy. Almost everybody dipped a toe and stopped. The country is wide and shallow.

So — am I behind?

Almost certainly not, and the numbers say so plainly. If AI has been opened in your business at all, you’re inside the 35%. If it isn’t running anything much, you’re inside the 90% who aren’t using it extensively either. That is not falling behind. That is the national position.

The thing hardly anyone has done is the next bit — getting one real piece of work handed over and keeping it there. That’s where the gap is, and it’s a much shorter distance than “catch up with everyone else” sounds like.

Where I was, for fifteen months

I’d have been a yes on that survey. I’d used it. I’d have ticked the box and gone in the 35%. Nothing whatsoever was happening in my business — four courses paid for and one attended, two dozen conference papers saved and never opened, a phone full of screenshots.

That’s not a criticism of the survey. It’s just what a headline figure hides — and I’m the evidence for it.

For the record, the numbers

12% → 35%

Adoption has roughly tripled in under three years.

In the ONS’s own words, self-reported use among businesses with ten or more employees “has increased from around 12% to around 35% since late 2023.” If you think British small business is ignoring this, that stopped being true a while ago.

ONS, BICS Wave 159, June 2026

28% v 49%

The smallest firms are twenty-one points behind the largest.

Twenty-eight per cent of businesses with nought to nine employees report using at least one AI technology. For those with 250 or more, it’s forty-nine per cent. The gap is real, current, and British.

ONS, BICS Wave 159, June 2026

UK AI use by business size · June 2026
28%0 to 9 employees
35%10 or more employees — the headline figure
49%250 or more employees

What businesses are actually asked

Worth knowing, because it explains why one number gets repeated and the rest don’t. The survey asks six separate questions about AI. These three are the ones that never make the headlines:

“To what extent does your business use artificial intelligence technologies in its business operations?”

Answer options: Used extensively · Used on a limited basis · Used only for testing or pilots · Not currently used · Not sure

“Approximately, what proportion of your workforce, if any, currently use artificial intelligence technologies as part of their daily work?”

ONS, BICS questionnaire, June 2026

“What proportion of your workforce, if any, have received training in the use of artificial intelligence technologies?”

ONS, BICS questionnaire, June 2026

The data exists. It is published. It is simply not the bit anyone repeats — and the bit everyone repeats is the one that makes people feel behind.

What the numbers still can’t tell you

International, for context only

Across the OECD, 2024 — not British figures

Firms with 10–49 employees11.9%
Firms with 50–249 employees20.4%
Firms with 250+ employees40%

The OECD’s own words: large firms are “more than three times” as likely to use AI as small ones. These are OECD-wide figures from 2024 and they are not a description of Britain — the UK’s own numbers are higher and more recent, and they’re the ones above.

OECD, “AI adoption by small and medium-sized enterprises”, discussion paper for the G7, 2025

A correction to this site

Until 30 July 2026 this site said British small firms used AI at 11.9%. That was the OECD figure above, from 2024, across all member countries — not a British number. The real UK figure for the smallest businesses is more than twice it.

It’s corrected here, and everywhere else it appeared. Saying so plainly seemed better than quietly changing it — the whole point of this site is that the figures can be checked, which has to include the ones I got wrong.

So what actually moves it

The same thing that moves anything: doing the first bit, badly, this week. Not a strategy, not a budget, not a decision about which platform. One small piece of real work handed over, seen through to the point where it either helps or it doesn’t.

That’s the whole of it, and it’s why this site is a record of doing rather than a set of predictions. Here’s the order I’d do it in — written down as I found it, including the parts that didn’t work.

How this page is kept honest

Non-Tech CEODoing it in the open

Privacy · How this is made · Home
Last checked 30 Jul 2026 · next review 30 Sep 2026