Copy article

AI Adoption In UK Firms Is A Mile Wide And An Inch Deep

ended 22. July 2026

New official statistics show that most businesses now dabbling with artificial intelligence are barely using it, and the smallest firms are furthest behind. The Office for National Statistics reports that self-reported AI use among UK businesses with 10 or more employees has risen from around 12 per cent to around 35 per cent since late 2023, yet only about 10 per cent of those adopters say they use it extensively. That is the catch the headline growth hides: adoption is broad but shallow, a tool switched on rather than a tool paying its way. It also splits sharply by size. The largest firms, those with 250 or more staff, are at 49 per cent, against 28 per cent for the smallest firms with under 10 people, a gap of roughly 21 percentage points. The blockers owners cite are rarely the technology itself: they are the difficulty of pinning down a real use case, the cost, and a lack of expertise, which around 18 per cent of medium-sized firms with 100 to 249 staff say has delayed their adoption. Most businesses report AI has had no impact on their headcount so far. The person caught in the middle is the sole trader or small-firm owner weighing whether a subscription actually earns back its fee, without an IT team to tell them which one does. 

  1. AI use has nearly tripled among larger firms in under three years, but only about 1 in 10 users call their use extensive. Is this real adoption, or a mile wide and an inch deep?
  2. The biggest firms are at 49 per cent and the smallest at 28 per cent. Who is being left behind by that gap, and is it a problem of cost and know-how rather than reluctance?
  3. What should a small-firm owner actually do to find a tool that pays back, rather than buying AI because everyone else is? Do you have a client whose plans this would change? If so, please give as much colour and detail as possible.

4 responses from the Newspage community

Copy all

Star Quote
Copy

Adopting a tool and getting value from it are two different things. Official figures show AI use in firms with 10 or more staff has jumped from around 12 per cent to around 35 per cent since late 2023, yet only about 10 per cent of those users say they use it extensively. That is a mile wide and an inch deep. It also splits by size: the smallest firms, under 10 staff, sit at 28 per cent, against 49 per cent of the biggest. That gap is not because small owners are luddites. A big firm has a team to choose the tools and a budget to pay for them; a sole trader has an evening. The blocker is rarely the technology. It is cost, and knowing which job you are actually trying to fix. So if you run a small firm, do not buy AI because everyone else is. Pick one slow, repeated task, a slice of your bookkeeping, quoting or invoice chasing, put a tool on that alone, and measure the hours it saves. Adopt one thing well before you adopt ten things badly.
Copy

Am I using AI? Every day. Are most of my small business clients? Not really. And that's the problem.A mile wide and an inch deep is bang on. I see it every week. A few firms are flying. They trained their staff, built it into the daily job, and it earns its keep. Most just switched something on, had a quick go, got confused, and left it sitting there like a gym membership they never use.The best ones point AI at the boring stuff. The heavy lifting. The admin. That is where it shines.

Where it goes wrong is firms using it to cut jobs. Wrong tool, wrong job. AI is only as good as the prompt you give it, and bias creeps in every single time. Aim it at people decisions and you don't fix the problem, you speed it up. Use it for the paperwork that eats your week. Not for deciding who stays and who goes.Point AI at the process, not the people. Then train your team to use it properly.
Copy

Britain’s AI sector is genuinely world class, but too many employers are still using it like a shiny spell-checker. That is not a failure of the technology; it is a failure to turn it into productivity.

The adoption figures look impressive, but they are a mile wide and an inch deep. Switching on an AI tool is not the same as redesigning how a business works. Large firms can afford teams to test, train and integrate it. Small firms are left staring at subscriptions, jargon and vague promises while trying to run payroll and answer customers.

The risk is not that British businesses hate AI. It is that they admire it from a safe distance while competitors make it pay. Small firms should start with one dull, expensive bottleneck — admin, quoting, customer emails or stock control — and prove the saving before buying the next shiny toy.
Copy

There's a real difference between an employee using ChatGPT for the odd email and a business embedding AI into how it works. With only one in ten adopters using it extensively, we're still experimenting, not transforming.

The size gap is understandable: a large firm can afford an AI team and absorb failed experiments. In a small business, the owner is often simultaneously the CEO, FD, and IT dept.

So start with the problem, not the product: find the task eating up the most time, test one tool against one measurable outcome, and only keep paying for it if it earns its place. AI is useful for research, analysis and first drafts, but we treat it like a bright graduate: fast and often impressive, yet equally capable of both valuable insight and significant error. It may reduce the need for incremental hiring rather than replacing existing jobs.

For many firms, the biggest risk is doing nothing. Those that learn to use these tools well will gain a real edge over those that merely dabble.