Small firms and AI in 2026: what the evidence says

The main gap in the market is diagnosis, not software.

Adoption has roughly tripled since 2023

Between a quarter and a half of UK small firms now use some form of AI, up from about one in eight in 2023. The estimates vary because the surveys ask different questions. Government surveys ask whether the business uses AI. Business-group surveys often count any use by staff, including free chatbots. The trend is the same in all of them.

Size matters. 28% of firms with 0–9 staff use AI, against 49% of firms with 250 or more (ONS). Business-to-business service firms adopt fastest: 46% use AI, against 26% of consumer-facing firms and manufacturers (BCC 2025).

Use is wide but shallow

Only about 10% of ONS adopters, and 11% of SMEs in the BCC survey, use AI extensively. Firms use 1.6 AI technologies on average, and a third of AI-using small firms only use it for routine admin (ONS, Simply Business). 85% of adopters use text-generation tools and only 7% use anything agentic (DSIT).

The barrier is "where would it fit?"

71% of UK firms say they have no identified need for AI, ahead of limited skills (60%) and a lack of suitable tools (48%) (DSIT). Cost comes far down the list, cited by 7–14% depending on firm size (ONS).

Staff are ahead of their employers

55% of workers use AI for work, but only 35% of businesses report using it (ONS). 80% of AI users at small and mid-sized companies bring their own tools, only 39% have had any training from their employer, and 60% of leaders worry their firm has no plan (Microsoft Work Trend Index). In practice that means company information going into personal accounts, with nobody checking the output.

Value follows process

Among accountancy practices, 87% of those using AI every day have documented processes, against 18% of those not using it (Xero). The tool works when the work around it is clear.

The gains are real, and bounded

Controlled studies find gains of 14–40% on tasks that suit AI. A support assistant raised issues resolved per hour by 14% on average and by 34% for newer staff (Brynjolfsson, Li and Raymond). ChatGPT cut professionals' writing time by 40% (Noy and Zhang). But consultants using GPT-4 on a task outside its abilities were 19 percentage points less likely to get it right (Dell'Acqua et al.). Danish data show almost no effect on earnings or hours two years on, because firms hadn't reorganised work around the tools (Humlum and Vestergaard).

And it is mostly support, not replacement: 95% of SMEs saw no change in headcount over the past year (BCC 2026).

What this means for a small firm

1. Start with one named task, not a strategy. The firms getting value picked a specific job and measured it.
2. Use what you already pay for. Most adopters use off-the-shelf tools, and 27% spent nothing on AI at all.
3. Write the process down first. Documented work is what separates daily users from dabblers.
4. Set simple rules. Which tools staff may use, what data stays out, and who checks the output.
5. Measure hours, not transformation. Count the time on that one task before and four weeks after.

Sources

1. Office for National Statistics (2026). Artificial intelligence in UK businesses: 2023 to 2026. BICS Wave 159, June 2026.
2. Department for Science, Innovation and Technology (2026). AI Adoption Research. 3,500 UK businesses with 5+ staff.
3. British Chambers of Commerce and Intuit (2025). The Turning Point for SMEs.
4. British Chambers of Commerce, Atos and University of Essex (2026). Half of SMEs using AI, with limited headcount impact so far.
5. Simply Business (2026). AI use among UK small businesses more than doubles in a year.
6. Microsoft and LinkedIn (2024). Work Trend Index: AI at Work Is Here. Now Comes the Hard Part.
7. Xero (2026). Modern Practice Playbook, reported by ResultSense. 520 UK practices.
8. Brynjolfsson, Li and Raymond (2025). Generative AI at Work. Quarterly Journal of Economics 140(2).
9. Noy and Zhang (2023). Experimental evidence on the productivity effects of generative AI. Science 381.
10. Dell'Acqua et al. (2026). Navigating the Jagged Technological Frontier. Organization Science.
11. Humlum and Vestergaard (2025). Still Waters, Rapid Currents. NBER Working Paper 33777.