AWS published its annual UK AI maturity report at the AWS Summit London this week, and the headline number is worth sitting with: 64% of UK organisations now use AI in some capacity. That sounds impressive. But only 24% have reached what AWS classifies as advanced-use maturity.

That gap — four in five businesses using AI but most of them stuck in the shallow end — tells you almost everything about where the real problem lies in 2026.

Adoption Is Not the Problem Anymore

The adoption rate has climbed fast. In 2024 it was 52%, then 64% today. The UK is now 10 percentage points ahead of the European average, with a business adopting AI every 40 seconds compared to 60 seconds across the continent. The early scepticism is over.

The new problem is the ceiling. Most businesses that adopted AI are using it for drafting emails, summarising documents, or handling basic customer queries. They ran a pilot, it worked, and they declared victory. Then they stopped.

AWS estimates that closing the gap between basic and advanced AI use could add £35 billion to UK GDP by 2030. That is roughly the entire annual economic output of Manchester.

The productivity data backs this up. Advanced AI adopters report 68% efficiency gains. Businesses stuck on basic use cases report 40%. The difference between those two numbers is not incremental — it is what separates a business that is genuinely competitive and one that has ticked an AI box.

The Real Barrier Is Skills, Not Technology

When asked what is holding them back from advanced adoption, 49% of organisations cited AI and digital skills shortages as their main challenge. That is up from 46% the year before. The technology barrier has largely fallen. The human capability barrier has not.

This shows up in hiring. Businesses are waiting an average of eight months to fill digital roles, and many are willing to pay a 41% salary premium for people with genuine AI literacy. The market for people who can actually do things with AI — not just prompt a chatbot, but build workflows, interpret outputs, and make sound decisions with AI-generated data — has become extremely competitive.

The AI skills gap is not about whether your team knows what ChatGPT is. It is about whether they can evaluate model outputs critically, build reliable automations, and connect AI tools to actual business processes.

What the 24% Are Doing Differently

The organisations reaching advanced maturity are not all large enterprises with dedicated AI teams. Some of them are mid-sized companies that made a deliberate decision to invest in the capability of their people rather than just buying software.

The pattern is consistent: they trained their teams on data fundamentals before layering AI on top. They built internal champions who could translate between business problems and AI solutions. They moved from “use AI on this task” to “redesign this process around AI.”

AWS also notes significant sectoral variation. Manufacturing is performing well — 58% of manufacturing organisations use AI for process optimisation and 74% report productivity gains. Healthcare adoption sits at 69% with 83% reporting accelerated innovation, but advanced adoption is growing at only 2% annually in the sector. Professional services are advancing faster.

What This Means for Business

If your organisation is in the 64% that has adopted AI but the 76% that has not yet reached advanced use, you are not in a safe position. The businesses pulling ahead are doing so quickly, and the gap compounds.

The skills shortage is the solvable problem here. Waiting eight months and paying a 41% premium for external talent is one option. Building that capability internally — through structured programmes that develop data literacy and AI fluency across existing teams — tends to be faster, cheaper, and stickier.

The fundamental shift AWS is pointing to is not that more businesses need to try AI. It is that the ones who have tried it need to stop treating it as an add-on and start treating it as a capability that every person in their organisation needs to develop.

The £35 billion opportunity is real. It goes to businesses whose people can actually use the tools in front of them.

---

Enterprise DNA put together a free field guide on exactly this: the full Claude ecosystem, Claude Code, and how to roll agents out without breaking things. Get the guide.