Where does Britain stand on AI?

The headlines say AI adoption is booming. The evidence is more nuanced — and more interesting — than that.

In the first article in this series, I asked who is shaping Britain's AI future and whether the choices being made are deliberate or default. In this piece, I want to look at the evidence. Where do UK organisations actually stand with AI right now? Not where the headlines say we are, or where the strategy documents aspire to be, but where the data suggests the reality lies.

The honest answer is: it's a mixed picture. That's not necessarily bad news. But it does demand a different kind of leadership conversation than the one most organisations are having.

The adoption story is real — but shallow.

Let's start with the good news. Over half of UK firms are now actively using AI, according to the British Chambers of Commerce, up from around a quarter two years ago. That's genuine, rapid progress. AI startups raised over £6 billion in UK venture capital last year. The government's AI Opportunities Action Plan estimates that widespread adoption could contribute £47 billion annually to the UK economy. The money, the attention, and the ambition are all there.

But look a layer deeper and the picture shifts. A recent study of 500 UK executives found that fewer than one in four can point to clear, measurable productivity gains from their AI investments. Most enterprise AI usage remains basic (simple questions in and answers out) rather than something woven into how organisations actually operate. And 60 per cent of businesses cite limited AI skills as a key barrier to doing more.

This isn't a story of failure. It's a story of early-stage adoption where the technology is running ahead of organisations' ability to absorb it. But it does mean that the gap between what leaders expect from AI and what their organisations are currently equipped to deliver is significant — and growing.

The pilot problem

Perhaps the most telling pattern is what's sometimes called "pilot purgatory." Organisations launch AI experiments, often with impressive results in controlled conditions, but then struggle to scale them into production. This is not a uniquely British problem. Estimates suggest that globally, the vast majority of AI pilots never make it beyond the experimental stage. But it matters enormously, because an organisation stuck in perpetual piloting is spending real money and real leadership attention without building lasting capability.

The reasons are remarkably consistent across sectors. It's rarely the technology that stalls. It's the things around the technology: data that isn't ready for production use, teams that lack the skills to operate AI systems at scale, governance frameworks that haven't been designed yet, and (perhaps most importantly) a gap between senior leadership's expectations and the reality on the ground. Recent UK research found that 65 per cent of senior leaders rated their organisation's AI maturity highly, while only 44 per cent of mid-level managers agreed. That perception gap is where AI projects go to die.

A leadership challenge, not a technology challenge

What strikes me most, having spent two years researching this, is that the UK’s AI challenge is fundamentally about leadership and organisational readiness, not about access to technology. The tools are available. The platforms are powerful. The investment is flowing. What's often missing is the connective tissue: the leadership literacy to make good decisions about AI, the internal capability to move from experiment to operation, and the strategic clarity to know which AI investments will create lasting value and which are just keeping up appearances.

This is where the conversation gets uncomfortable for many senior leaders. It's tempting to treat AI as a technology project that can be delegated to the CTO or the innovation team. But the evidence is increasingly clear that the organisations making real progress are those where leadership at the highest level is engaged. Not necessarily in the technical detail, but in the strategic choices about where AI fits, how the organisation needs to change to absorb it, and what "success" actually looks like beyond the pilot stage.

The UK has navigated transitions like this before. The shift to digital government, the move to cloud, and the transformation of financial services. In each case, the organisations that succeeded were those that treated the change as a leadership and organisational challenge, not just a technology one. AI is no different. If anything, the stakes are higher, because the decisions being made now will be harder to reverse.

So, if this is where we are with real momentum, but shallow adoption, persistent skills gaps, and a leadership perception problem, what should we actually do about it? That's where I'll pick up in the final piece in this series.

Next in this series: In Part 3, I'll set out what I think an adaptive path for Britain's AI future could look like — practical steps for leaders who want to move from experimentation to real impact.

Alan Brown is Professor of Digital Economy at the University of Exeter Business School and Research Director at the Digital Policy Alliance. His new book Making AI Work for Britain is published by London Publishing Partnership (April 2026). Follow Alan on LinkedIn and find out more at alanbrown.net.