What does the UK's AI adoption survey 2026 mean for AI training?

Short answer: The UK government's AI adoption research 2026 shows that a third of UK businesses now use AI, double the figure from a year earlier, but most use it in basic ways and see modest gains. The biggest thing holding them back is skills, not technology, which makes practical, ongoing AI training the clearest next step for most organisations.

What the research found

On 29 October 2026, the government published its AI adoption survey 2026, the second wave of government research into how UK businesses use AI. The first wave was commissioned by the Department for Science, Innovation and Technology in early 2025. It is based on 3,503 telephone interviews with businesses between 7 January and 21 April 2026, followed by 50 in-depth interviews and 7 case study visits.

The headline is growth. According to the executive summary, 32% of UK businesses were using at least one AI technology in 2026, compared with 16% in 2025. Uptake was higher among larger firms (52%) and in knowledge-intensive sectors such as information and communication (61%), real estate (45%) and business services and administration (39%).

But the depth of use is still shallow. Most use was in accessible, low-complexity tools, especially text generation, with large language model tools used by 77% of those using or planning to use AI. The most common areas were administration (81%), marketing (71%) and IT (66%). The report describes most businesses as still at an early or exploratory stage.

Skills are the biggest barrier

The research is clear that what holds businesses back is capability. Limited AI skills, expertise or knowledge affected 67% of businesses that had adopted AI and 78% of those planning to. Among businesses that weren't using or planning to use AI, the change most likely to help was a better understanding of how AI could help their business (62%), closely followed by a clear business case or proven return on investment (61%).

The report also says cost was more often about the time, effort and resource needed to implement and embed AI than the price of the tools. And 47% of businesses that could estimate their spending reported no AI expenditure in 2025, which the report links to free tools and AI features built into software they already have.

In other words, many teams already have AI in front of them. What they lack is the know-how to use it well.

The results are real but modest

Half of adopters (50%) reported an overall increase in productivity, though these gains were generally described as modest. The most common benefits were improved workforce productivity (71%) and more efficient or flexible processes (60%). Only 14% of businesses considered AI tools to be of high or critical importance to their operations.

Trust is cautious too. Most businesses (81%) reported applying some level of human oversight to AI outputs, and 56% of adopters said they had concerns about bias and lack of accuracy.

What the interviews say about training

The full AI adoption research report goes further than the numbers. In the interviews, hands-on, experiential learning was preferred, and training was seen as something that needed to be ongoing rather than a one-off session. Many people were teaching themselves through webinars, YouTube, newsletters and industry events.

The report also notes that getting useful outputs was seen as difficult for people less confident in prompt writing, and that general AI knowledge was not seen as the same as AI literacy in a specialist field.

One case study stands out. An operations manager described a Copilot rollout at their parent organisation where an online training programme had to be completed before staff got a licence. About six months later, he observed that few colleagues were making regular use of the tool. He said the rollout would have benefited from more practical, facilitated training rather than self-directed online modules, and from clearer guidance on specific tasks it could be used for in different roles.

In our experience, a short compliance module gets people access, but it rarely changes how they work.

Practical takeaways by role

L&D leaders: Plan AI training as a programme, not a single event. The research points to hands-on practice and ongoing learning, so build short follow-up sessions around real tasks for each team. Our AI upskilling page explains how we approach this.

HR leaders: The interviews describe older, more experienced staff and managers as more reluctant to engage with AI. Involve managers early, set clear expectations, and make sure your employee AI policy tells people what they can and can't do. Our page on AI training for HR teams covers the people side of change.

IT leaders: The report links low AI spending to free tools and AI features built into software businesses already have. Find out which AI features your people can already reach, and pair each one with role-based guidance on what to use it for and what data to keep out of it.

Transformation leaders: The research found that many businesses lack a clear business case. Set a baseline before training and measure task-level change afterwards. Our guide to measuring AI training ROI explains how, and our guide to business AI adoption covers the wider rollout.

A simple starting plan

1. Find out what people already have. List the AI tools and built-in features your teams can use today.
2. Pick a few real tasks per team. Start with drafting, summarising and admin, where the research shows most use and most trust.
3. Train hands-on. Use your own documents and workflows, not generic demos.
4. Follow up. Run short refreshers so skills keep pace with the tools.
5. Measure. Compare time and quality on those tasks before and after.

For more context, see our posts on the UK Digital and Technologies Jobs Plan and the AI confidence gap.

What this means for your team

The government's own research says the gap is skills, not access. If your people already have Copilot, ChatGPT, Gemini or Claude, the quickest gains are likely to come from teaching them to use those tools well in their actual jobs, with practical sessions and regular follow-up rather than a one-off module.

Not sure where your team stands? Our free AI readiness scorecard takes a few minutes. You can also browse our AI training for business, including Microsoft Copilot training and ChatGPT training, or see our AI training in Leeds.

Talk to us

If you would like help turning AI tools your teams already have into everyday skills, book a free 60-minute discovery call with Qasim.