Britain’s AI race: leading Europe but lagging on ambition
AWS’s annual ‘Unlocking the UK’s AI Potential‘ report finds that nearly two-thirds of UK organisations now use AI – but a £35 billion opportunity is being left on the table as advanced adoption stalls.
The UK is now the fastest AI-adopting nation in Europe – and according to Amazon Web Services‘ latest research, a company is choosing to deploy artificial intelligence every 40 seconds. That is a sharp acceleration from the 60-second rate recorded just a year ago. Yet beneath that headline momentum, a more uncomfortable story is emerging: the country risks squandering a historic opportunity by thinking too small.
AWS’s fourth annual ‘Unlocking the UK’s AI Potential’ report – produced by independent consultancy Strand Partners – finds that 64% of UK organisations now use AI, up from 52% last year and well ahead of the European average of 54%. Productivity gains are widespread: 68% of adopters report measurable improvements, 72% expect AI to boost growth in the coming year, and 79% say their innovation timelines have accelerated.
But Phil Le Brun, Executive in Residence at AWS, is candid about where the optimism ends. “We’re approaching that S-curve where more and more companies are actually adopting AI,” he told Startups Magazine. “The downside is we’ve seen a shift from 23% of companies using advanced AI last year to just 24%.”
The smartphone that only makes calls
The report draws a pointed analogy for the current state of British AI: most organisations are like someone who owns a smartphone but uses it only to make phone calls. Basic AI – off-the-shelf chatbots, document summarisation, marginal efficiency gains – is what the majority of businesses have settled for. Advanced AI, which involves redesigning workflows, autonomous multi-agent processes, and building entirely new products and services, remains the territory of a minority.
The economic stakes of that gap are vast. Organisations using advanced AI report average efficiency gains of 68%, compared with 40% among basic users. Closing the divide could unlock an estimated additional £35 billion in economic growth for the UK by 2030 – roughly equivalent to the entire economy of the City of Manchester.
Le Brun frames the challenge in blunter terms. “This isn’t a new version of a steam engine – this is electricity. This isn’t a new version of machine learning – this is agentic. It’s very different and it’s much more powerful.”
The pace of change underscores his point: it took a decade to go from dial-up Internet to 3G, whereas the leap from Generative AI to Agentic AI – systems capable of autonomously planning and carrying out complex tasks – happened in months.
The imagination gap
When I asked Le Brun why so many organisations remain at the basic level, he resists the obvious answer. Skills shortages matter – 49% of organisations cite them as their main barrier, up from 46% last year – but he argues there is a deeper problem that goes all the way to the boardroom.
“One of the barriers is really imagination,” he says. “If you can do accounts payable in 60 days and now you can do it in 50 days, that’s efficiency. It’s a different proposition entirely to say: how do I fundamentally rethink how I do business and what the customer wants?”
The report identifies this as an “imagination gap” with leaders. Rather than asking how AI can help them do the same things faster, Le Brun argues that the most successful organisations are those that start with the customer problem and work backwards.
Skills: a crisis at every level
The skills challenge is real, but Le Brun is quick to complicate the picture. “Often we think of skills at the engineering level, the grassroots,” he says. “But this is skills from the C-suite down.” A recent Harvard report on Agentic AI reached the same conclusion: leadership capability, not just technical talent, is the binding constraint.
The numbers are alarming. Only 17% of organisations say they have a strong AI skillset today, yet 84% expect AI skills to be critical within five years. It now takes eight months on average to fill a digital role – up from five and a half months in just a single year. Organisations are willing to pay a 41% salary premium for AI talent, worth an additional £16,000 per year on the UK median full-time salary.
Industry-led upskilling programmes are part of the solution, but Le Brun argues they are not sufficient on their own. “This starts in schools and universities, but it also starts in the workplace itself. Are employees being given the opportunity to use the technology in their day job? Are they being allowed to experiment, to improve their own role? That’s how learning really works.”
He draws a telling generational comparison. “When I grew up, the first time I used a laptop was in the office. The first time I used email was in the office. We’re all using this technology at home now.” His point: organisations that fail to make AI available to employees within the workplace are creating an artificial barrier where none need exist.
A nation of regional revivals
One of the more surprising findings of this year’s report is where AI growth is actually happening. While London leads in overall adoption at 72%, it is the regions that are accelerating fastest. The North West grew AI adoption by 28% year-on-year; the North East by 26%; Wales by 25%. All outpace London’s 16% growth.
Manufacturing stands out as a particular bright spot. Some 58% of manufacturing organisations are using AI for process optimisation, with 74% reporting productivity gains. As Physical AI and Edge AI – systems embedded directly into machinery and production lines – continue to mature, the industrial heartlands of the UK could become early and deep beneficiaries.
Healthcare and pharmaceuticals represent perhaps the largest untapped opportunity. Adoption stands at 69%, and 83% of those organisations say AI has accelerated innovation – yet advanced adoption is growing at only 2% a year. The UK’s combination of world-class health data, NHS delivery infrastructure, and research-led institutions gives the sector a competitive advantage that greater AI adoption could translate into global leadership.
Government as a role model
The report makes a striking case for the public sector’s importance as a catalyst for private adoption. More than three-quarters (78%) of organisations say they are more likely to adopt AI if the public sector integrates it visibly into its own systems. Over a third of startups say public sector procurement practices are critical to their ability to scale.
On this measure, the data is more encouraging. Some 58% of public sector organisations have now adopted AI – and, notably, the government’s advanced use of the technology is proportionally ahead of industry in general. Le Brun expressed some genuine surprise at that finding. “Is the government serious about this? They’re actually starting to role-model how this technology can be used to improve outcomes, not just drive efficiency.”
AI Minister Kanishka Narayan welcomed the report’s findings, noting that full potential “will only be realised when businesses and workers have the skills they need to use it effectively,” and pointing to the government’s target to equip 10 million workers with AI skills by 2030.
What comes next
The report identifies three priorities for UK organisations and policymakers in the year ahead: close the digital skills gap at pace through training, public-private partnerships, and AI literacy at every level of education; help organisations move from adoption to transformation through AI strategies that make the most of Cloud capabilities; and scale AI across public services so government leads by example and unlocks what independent analysis suggests could be over £45 billion in annual savings from full digitisation.
Le Brun is watching for a tipping point. He expects the S-curve of adoption to steepen, but is realistic about the pace. “I would hope for a big leap. My gut is that we’re going to see us start to go up that curve – but it’s not going to be halfway up. With large enterprises, there’s a fundamental rewiring going on: how do I make decisions faster? How do I become more customer-centric? That makes the transition slower. But it doesn’t mean they can’t use advanced AI – and in pretty much every customer we work with, there are already pockets where incredible things are being done.”
His hope for next year’s report is simple: “More advanced use of AI. AI doesn’t replace people. People use AI to drive exceptional outcomes – but we need those people with the right skill sets.”
