AI in UK business: adoption has more than doubled, but most firms have barely started
More businesses use AI every quarter, yet in most of them a few people use it for a few tasks. The gap between trying AI and benefiting from it is where the advantage now lies, and small firms can close it faster than they think.
Ask whether British businesses are using AI and the answer depends on who is asking and how. Two official surveys measure it. The Office for National Statistics, which asks firms with ten or more employees, found that around 35% used at least one AI technology by June 2026, up from around 12% at the end of 2023. The government’s own survey, which includes the smallest firms, puts it at about one business in five using AI or planning to.
Adoption has roughly tripled in under three years.
Source: Office for National Statistics, Business Insights and Conditions Survey. Approximate shares.
Show the data
| Late 2023 | 12% |
|---|---|
| June 2026 | 35% |
Either way the direction is clear, and so is the pattern underneath. Size matters a great deal. In the government’s survey, 36% of businesses with 250 or more employees used AI, against 23% of mid-sized firms and 14% of the smallest firms, those with five to nine people.
- 250 or more36%
- 50 to 24923%
- 5 to 914%
Source: DSIT AI adoption survey, reported in Assessment of AI capabilities and the impact on the UK labour market (January 2026).
Show the data
| 250 or more | 36% |
|---|---|
| 50 to 249 | 23% |
| 5 to 9 | 14% |
Wider, but not yet deeper
The more telling figure is how far AI reaches inside the firms that have it. According to the government’s assessment, fewer than a third of employees in adopting firms use it. The ONS finds the average adopter uses fewer than two kinds of AI technology, and only around one in ten describes its use as extensive. In most businesses AI is a tool a few people have tried, not a change in how the work gets done.
Source: DSIT AI adoption survey, reported in Assessment of AI capabilities and the impact on the UK labour market (January 2026).
The gains are the businesses’ own estimates, not independent measurements.
Show the data
| of businesses using AI report productivity gains, most of them of up to 20%. | 56% |
|---|
The gains are real but uneven. The government is careful to say these are self-assessed, and that there is not yet strong statistical evidence linking AI to higher productivity across the economy. In our experience the firms that see results have something in common: they chose one process, changed how it works and measured the difference.
Why small firms lag, and why that can change quickly
Small businesses have less time to experiment, fewer specialists and tighter budgets. But the tools they need are now mostly built into software they already pay for: accounting packages that read receipts and suggest codes, email and document tools that draft and summarise. The barrier is less the technology than the time to set it up well, and the confidence that it is safe.
Help is growing. In October 2026 the Chancellor announced a package of AI investment that includes £30 million for local AI partnerships led by mayors, bringing councils, businesses including small ones, universities and colleges together to adopt AI and give workers AI skills.
Where to begin
- Pick one process that costs time or causes errors, such as coding receipts or chasing late payers.
- Measure how long it takes today, so you can tell whether AI helps.
- Decide what data the tool may see, and keep personal and client data out of tools you have not checked.
- Name the person who checks the output and answers for it.
- After 90 days, keep it, change it or stop it, and choose the next process.
Sources
This insight is general information, not advice for your circumstances. Talk to us before acting on it.