Britain’s AI Moment Is Real — But It’s Nowhere Near Big Enough

Britain’s AI Moment Is Real — But It’s Nowhere Near Big Enough

Last week, UK Chancellor Rachel Reeves set out her vision for long-term investment in Britain’s future in the Spending Review. Reeves set a tone of “national renewal” and some of the initiatives were promising:

  • £86 billion for science and technology across the next three years
  • £2 billion earmarked specifically for AI, including a new £750 million supercomputer in Edinburgh
  • major backing for NHS digitisation (up to £10 billion)

But, in AI, ambition isn’t judged by intentions. It’s judged by scale and speed. And on those metrics, Britain is still playing in the lower leagues.  While the UK is investing billions, others are investing hundreds of billions — and already breaking ground.

The UK is trying. But it’s not trying hard enough.

Reeves is taking a bold political gamble: avoid near-term tax increases, front-load spending, and hope that economic growth bails us out later. The logic makes sense — up to a point. But the AI doesn’t wait for careful, incremental policy cycles. It rewards whoever moves fastest with the most capital.

A key component of the Spending Review's technology impetus is the £2 billion AI investment package — including £750 million for the new Edinburgh compute facility. This marks a much-needed reversal from earlier dithering (in August 2024, this project had actually been cancelled!), and the stated aim is important: a 20x increase in compute capacity by 2030.

But to understand why this still falls short, consider what else is happening around the world:

  • Germany just secured a $500 million investment from Nvidia to build the “world’s first industrial AI cloud” — with 10,000 GPUs on day one.
  • The UAE and US are backing Stargate, a series of AI compute infrastructure projects totalling $100 billion to $500 billion.
  • US venture investment into AI this year: $80.8 billion. UK AI VC funding in the same period? Just $4.3 billion.

This is the scale of the arms race. And right now, the UK is showing up with a lemon to a knife fight (to quote the Wombats).

A reality check from Jensen Huang

When Nvidia’s CEO, Jensen Huang, took the stage alongside Keir Starmer at London Tech Week, he didn’t mince his words:

“The UK is the largest AI ecosystem in the world without its own infrastructure.”

Huang described the UK’s position as a “Goldilocks moment”: the right mix of talent, research excellence, and supportive government — but missing the most important piece: infrastructure to actually do something with it.  It was a backhanded compliment.

Jensen is right, Britain has all the raw ingredients for global AI leadership but is falling short on investment:

  • Oxford, Cambridge, Imperial, UCL producing world-class technical talent
  • DeepMind and Isomorphic Labs as global leading labs
  • Synthesia and ElevenLabs showing global product reach
  • A vibrant (albeit underfunded) early-stage VC ecosystem
  • Real commitment to AI safety and governance

But we are stuck because the infrastructure that modern AI requires — cheap energy, dense compute, data infrastructure — just isn't there in sufficient quantity and it's getting hit by arcane planning laws and NIMBYism.

Look at Havering in East London: a single proposed mega-campus there — at 600 MW, over one-third of the UK’s entire data-centre capacity — has been stuck in planning limbo for years, tangled in local objections and power-connection delays. Meanwhile, the UK, the sixth largest economy in the world, deploys less than 1% of global GPU capacity, while the US and China break ground on hyperscale clusters at breakneck speed. The UK can’t lead in AI if it keeps blocking the concrete and cables that power it. Local resistance must be balanced with national urgency — or Britain’s AI dreams will keep crossing the Atlantic to where the servers actually get built.

The bigger crisis no one wants to talk about

Underneath the AI story lies a deeper economic tension: Britain is quietly becoming a healthcare state.

According to the FT, by 2029 nearly half of all UK government spending will go to health and social care. This is a structural trend — driven by demographics, not politics. And it puts an enormous squeeze on capital available for long-term investment in productivity-enhancing sectors like AI.

Spending £86 billion over four years sounds bold. But set against the tidal wave of healthcare obligations, it’s a drop in the ocean.

What this means is simple: if the UK wants to remain a serious player in AI, it needs to do one of two things — or ideally both:

  1. Massively scale up public investment in critical infrastructure like compute, energy access, and AI-ready cloud.
  2. Attract global private capital to build major AI projects on British soil.

And the UK just got outdone by Germany with the announcement of the Nvidia deal, a signal that Germany is open for business and can offer the scale, regulatory clarity, and industrial planning that these projects need.  The UK needs to do the same and quickly to attract private capital.

UK talent is ready — but leaving

A Tech Nation study recently found that 43% of UK AI founders are actively considering relocation outside of the UK (with most heading to the US). The reasons? Limited access to compute, regulatory uncertainty, and the sheer gravitational pull of the US venture and talent market.

There should be red alert in Westminster.  This isn’t a marginal industry (like steel or fishing) this is THE entire future of economies.  

Some reasons for optimism — but only if followed through

To give credit where it’s due, the current government is showing signs of urgency:

  • The ARIA fund is getting real capital and political backing
  • The NHS is getting a £10 billion digitisation boost — one of the largest AI-adjacent projects in Europe (albeit details are vague).
  • The British Business Bank is being empowered to deploy more strategically

These are good initiatives. But they need to be followed by even bolder ones:

  • Offer tax incentives for building UK-based AI infrastructure
  • Court big tech and AI companies with land, labour, and long-term certainty
  • Get serious about building energy capacity to support data centres and bring down the aggressively high cost of electricity for industry in the UK.

Final thought: it’s not too late, but it’s close

AI isn’t just another tech wave. It’s the foundation of economic competitiveness for the next 30 years. It touches every sector: energy, health, defence, education, manufacturing.

Missing this moment — or underplaying it — would be a catastrophic mistake.

The UK is blessed with world-class minds, deep technical talent, and a credible commitment to ethics and safety. That’s not nothing. But without the infrastructure to match, the UK will keep exporting talent, losing companies, and falling further behind.  

And what to do as an early stage investor?  Invest with a patriotic hat on in founders who seem at times to be swimming against a fast tide in the UK or back them to move to the US where the tides, at least in respect of AI and critical technologies, is moving in the right direction....and fast.

To view or add a comment, sign in

More articles by Lomax Ward

Others also viewed

Explore content categories