AI Adoption Doubles to 32% of UK Firms as Skills Gap Bites
A third of UK businesses with five or more staff now use AI, double the 16% of a year earlier, according to the government's AI adoption survey 2026. Marketing is one of the most common places they use it, yet most adopters say they lack the skills to go further, and only 14% call their AI tools highly important.
01Telemetry
Key findings at a glance
- 32%of UK businesses with five or more staff used at least one AI technology in 2026, double the 16% in 2025
- 52%of large businesses (250 or more employees) use AI, the highest of any size band
- 27%of micro businesses (5 to 9 employees) use AI, nearly double the 2025 level
- 61%of information and communication businesses use AI, the highest of any sector, up from 43% in 2025
- 69%of retail and distribution businesses neither use nor plan to use AI, down from 86% in 2025
- 71%of businesses using or planning to use AI apply it in marketing, second only to administration at 81%
- 75%of businesses using or planning to use AI use it, or plan to, for creative and content creation
- 50%of AI adopters report an overall increase in productivity, though most describe the gains as modest
- 14%of businesses using AI rate their AI tools as highly important to their operations
- 47%of businesses using or planning to use AI that could estimate their spend spent nothing on AI in 2025
- 67%of AI adopters cite limited AI skills, expertise or knowledge as a barrier, rising to 78% of those planning to adopt
- 81%of businesses using AI apply at least some human oversight to its outputs, down from 84% in 2025
02Scope
What the report covers
The AI adoption survey 2026 was published on 29 October 2026 by the Department for Business, Innovation, Science and Trade (BIST). It is the second wave of a government series that tracks how UK businesses use AI, the barriers they face and the support they need, and the Department for Science, Innovation and Technology (DSIT) commissioned BMG Research to run it. The first wave's results are in the AI adoption survey reports collection on gov.uk. Across 3,503 UK private-sector businesses with five or more employees, 32% were using at least one AI technology (including pilots), up from 16% in 2025. A further 6% planned to adopt, and the share saying AI is not relevant to their organisation fell from 51% to 38%. It is the demand-side companion to our report on the government's AI sector study, which counts the 10,488 UK companies that supply AI.
Adoption rises with size: 52% of large businesses use AI, against 47% of medium, 34% of small and 27% of micro businesses (5 to 9 employees). Information and communication leads the sectors at 61% (43% in 2025), with business services and administration at 39% (23%). At the other end, 69% of retail and distribution businesses neither use nor plan to use AI (86% in 2025), alongside hotel and catering (73%), transport and storage (71%) and construction (69%). Use is concentrated in accessible tools. 77% of businesses using or planning to use AI rely on large language model based approaches, and OpenAI models such as ChatGPT are used by 92% of adopters, with Google's Gemini at 38%.
Among businesses using or planning to use AI, the most common business areas are administration (81%), marketing (71%, against 72% in 2025) and IT (66%), and 75% use it or plan to for creative and content creation. Efficiency is the main driver (71%), while gaining or keeping a competitive advantage rose from 12% to 25%. The returns so far are incremental: half of adopters (50%) report an overall productivity increase, generally described as modest, 64% have seen no immediate effect on operating costs, and only 14% rate their AI tools 8 or more out of 10 for importance. Spending is light too. Of those able to estimate it, 47% spent nothing on AI in 2025, which the report puts down to free tools and AI features built into existing software.
03Briefing
What this means for UK marketers
There are three readings here for UK marketers.
First, AI-assisted marketing is now normal among adopters, not an edge. 71% of businesses using or planning to use AI apply it in marketing, 75% to creative and content work, and most use the same handful of chat assistants. When competitors draft with the same models, copy converges, and AI writing is already visible across a large share of new web pages. The small online fashion retailer profiled in the report saved real time on meta descriptions across hundreds of products, but staff worked to protect their brand voice after one email was criticised internally for sounding generically AI-written. Use AI for volume and first drafts, and keep the distinctive parts (product knowledge, tone, proof) human. Oversight is slipping: 81% of AI users apply some human checking, down from 84%, and 5% now apply none. Anything that makes a claim about price, product or performance still needs a named person to sign it off.
Second, retail is behind. 69% of retail and distribution businesses neither use nor plan to use AI. For ecommerce teams, the low-risk starting points are the ones that retailer chose: product copy, meta descriptions and email drafts, checked before they go live. The same business abandoned its platform's built-in AI assistant because it kept returning inaccurate analytics, so check any AI-reported number against the source data. That product-page work at scale is the core of our ecommerce SEO service.
Third, if you sell to businesses, the barriers tell you what to write. Adopters most often cite limited skills (67%) and limited knowledge of how AI could help (64%), and 65% of firms neither using nor planning to use AI see no clear business case. The price of tools is not the main obstacle, since 47% of those who could estimate spent nothing. Content that explains use cases, implementation effort and measured outcomes will do more than a list of capabilities, especially in regulated sectors such as finance, legal and healthcare, where the report found the most caution. ONS figures point the same way: about three in ten UK businesses use AI, and only 6% of adopters say it has cut headcount. Demand is real but early, and buyers still want proof.
04Method
How they got the data
Check the method before you quote the number.
BMG Research ran the survey by computer-assisted telephone interviewing (CATI) between 7 January and 21 April 2026, speaking to the person responsible for technology in each business. Interviews averaged 15 minutes. The 3,503 businesses were drawn from Market Location, a commercial database of UK businesses, and cover the private sector only: businesses with fewer than five employees were screened out, and public administration, education, and health and social work were excluded. Results are weighted by business size and sector; the regional split matched the population, so no regional weight was applied. Of the 69,370 businesses called at least once, 5% completed an interview. The 2025 comparisons come from the first wave, which interviewed 3,500 businesses between February and May 2025.
The survey was followed by 50 in-depth interviews between February and June 2026 with businesses that had taken part, deliberately over-representing AI users, and seven ethnographic site visits to AI adopters, written up as case studies. Many of the percentages are based on subgroups (businesses using AI, or using or planning to use it) rather than all businesses, and the report states the base on every chart. Because the figures describe firms with five or more staff, they are not directly comparable with the ONS estimates above, which come from a different survey. The technical annex and question wording are in the full AI adoption research report.
