Tag: 2025

The speech made by Peter Kyle, the Secretary of State for Science, Innovation and Technology at the Mansion House in London on 3 September 2025.

I want to talk about a society and economy where AI benefits every person and community in the country.

Where there is opportunity for anyone, no matter their background. Where our huge potential for wealth creation isn’t centred in the capital but is distributed where talent lies – which is everywhere people are.

And where we finally become a country that celebrates entrepreneurial zeal, and those can who move from innovation to commercialisation, and thrive in a modern Britain.

To get there, the question I’m seeking to answer in government isn’t ‘do we want to be a country that adopts AI or not’…because AI is going to happen. We know it is.

That’s true for every country from Britain to North Korea.

The better question is: do we want to use all the power and agency we have, as a government to shape how it unfolds?

This will be my focus tonight.

And what better setting than here, in Mansion House.

The place where the future of this country has so often been debated, and yes, defined.

Almost 70 years ago, in April 1956, this room was the setting of a major Cold War summit.

Where Nikita Khrushchev came, after Stalin’s death, with a delegation from the USSR.

On the way here he stopped by Claridge’s. Clearly nothing is too good for the workers.

Before coming for dinner here, with Clement Attlee and Anthony Eden.

Mansion House was done up, looking its finest.

The Lord Mayor gave a speech.

And the team put on a delicious spread, as always.

It was a huge effort.

But we know now, looking back at history…the charm offensive didn’t really work.

Khrushchev returned to the USSR as resolved as he ever had been to lead, through technology.

Racing to be first to the space rocket, the microchip, the bomb.

Thankfully, those Cold War days are behind us.

But on the world stage, the tech competition remains as fierce as ever.

Only this time, the defining competitive advantage of this century – I believe…is going to be AI.

Artificial intelligence will shape our economies, our security, and our place in the world.

Those who wield it in their national interest – who invest in the right skills and hardware, while they have the chance…will be the economic superpowers of the future.

Look at the US, and China.

Or the Gulf states, vying to compete.

Time and again, we see that a nation’s sovereign interest rests on its technological edge.

It’s the lesson Khrushchev almost learned – but not quite.

You see, in the years after his Mansion House visit, Khrushchev set about increasing the USSR’s tech capacity.

As part of the plan, he wanted to have his own Soviet Silicon Valley.

He ordered the creation of a new city, on the outskirts of Moscow: Zelenograd.

The city had its own cinema, billboards, homes and offices. And a massive statue of Lenin.

And the whole place was designed for engineers, racing to design microchips.

The idea was simple: copy what worked in America, but do it faster.

The Soviets got their hands on a prototype for a US microchip, the SN-51.

Alexander Shokin, the official in charge, summoned the engineers of Zelenograd into his office, and he ordered them:

“Copy it, one-for-one” – without a single deviation.

This, ultimately, was their big mistake.

The Soviets chose to imitate rather than innovate.

At a time when the pace of change in chips was impossible to keep up with, without their own domestic research capacity.

American speed proved too difficult for the Soviets to match.

Individual US entrepreneurialism outpaced Soviet central control.

And Silicon Valley won the chip race.

Today, history is repeating itself in the development of AI and the new technological revolution.

And the UK must think like the US, not act like the USSR.

The computing power needed to train leading AI systems has doubled roughly every 6 months for the last decade.

If we don’t keep up – with a domestic AI ecosystem of our own, on British shores, we’ll always be beholden to others, following where they point us.

Buying off-the-shelf, from overseas.

That’s a precedent I am not willing to set when it comes to our military tech, the integrity of our NHS, and data protection.

Or when the prize is a huge competitive edge for our economy.

Fortunately, we’re starting from a good place.

We currently rank third for AI, after the US and China. We have 4 of the world’s top ten universities. The lowest corporation tax in the G7. And more venture capital investment than anywhere else in Europe.

Only this year, the chief executive of NVIDIA, Jensen Huang, said the UK is in a ‘Goldilocks’ moment for AI.

Because we are country not burdened by over-regulation, or a lack of ambition.

Britain is striking the right balance.

I remember, before this government came to office, I spent ages asking businesses what they needed from us.

You certainly weren’t shy in telling us.

Take AI seriously. Regulatory reform. Make sure we don’t fall short on talent.

And ever since, we’ve been getting on with it.

It started in January, with the AI Opportunities Action Plan.

It outlined the 50 steps we are taking to grow the economy, and create scores of new jobs as part of the government’s modern industrial strategy.

An early priority for me was skills.

In June, we launched our TechFirst programme – backed by £187 million in funding which will bring digital and AI learning directly into the classroom and reach every secondary school pupil in the country.

Next, we looked at the workforce. Forming a skills partnership with firms like Barclays, Amazon, BT, and Google.

Together, committing to train 7.5 million people in AI – a fifth of the country’s workers.

It’s fantastic to see how so many of you have risen to that challenge.

After that, we looked at hardware.

If we wanted to compete, we knew we had to improve our physical machinery.

The raw processing power we have on offer, here in the UK, to churn through the mountains of data that will be required.

The Compute Roadmap I set out in July charts that course.

And I recently launched Isambard – our new supercomputer, the most powerful in the country.

A machine that will be able to process an unthinkable amount of information, in seconds.

We have another one, Dawn, in a lab in Cambridge.

And I’ve announced the creation of a national supercomputer that will be based in Edinburgh.

All told we’re on track to increase our compute capacity 20-fold between now and the end of the decade.

We have our plan for the National Data Library.

AI is pretty straightforward in its basic form. It is chips. It is data. It is software.

We talk about chips a lot. We talk about software a lot. But we need to talk much, much more about the data that fuels it.

AI is only as good as the data it uses, and Britain has the best data in the world.

We will be safely harnessing it to power scientific and medical discovery, to drive our understanding of the human condition, and as potentially the biggest engine for the commercialisation of innovation in our country’s history.

And we’re not slowing down any time soon.

Our next big priority is our AI Growth Zones.

These will be dedicated hubs of AI development.

The first will be just 60 miles away, in Culham.

And we’re getting spades in the ground for sites in Wales and in Scotland.

Each has the potential for a full campus – bringing together companies, researchers, and investors.

These efforts have been met by a wave of commitment from the private sector.

With £14 billion in investment announced by firms like Vantage Data Centres, NScale, and Kyndryl.

A brilliant British company, Synthesia, has announced they are expanding their London office just a few weeks ago.

And global firms like Cohere, Open AI, and Anthropic have followed suit – choosing our capital as their home from home.

That is a vote of confidence not just in our tech sector, but in the UK’s future.

So this evening, I’m proud to publish the next 2 parts of our plan.

The first is a roadmap for a new British AI assurance industry.

Backed by a fund worth £11 million.

In the next few years, AI assurance will bloom into a unique profession, worth up to £18.8 billion to our economy, based on a growing pool of independent experts with the skills to verify that new AI innovations are secure, and trustworthy.

We hope it will give firms the tools they need to build trust with both customers and markets, especially smaller teams, who lack the in-house expertise to do this work themselves.

Applications for that fund will be opening in the Spring – please do keep an eye out.

Lastly, we’re looking very closely at regulation.

I know this is a crucial issue for many of you. So I want to make it plain:

British companies shouldn’t have to wait months for approvals, whilst competitors overseas race ahead. If AI can speed things up, even a little, then we will do everything we can to make that a reality.

As part of this effort, today we announced our new AI regulator capability fund.

Designed to support 5 UK regulators – from Ofgem, to the Civil Aviation Authority with up to £2.7 million in funding, to help them both use and regulate AI better.

Whether it’s a new AI assistant. Analysing huge datasets. Or streamlining approvals.

This is our challenge to regulators:

Use every tool at your disposal to get new products to market quickly, without sacrificing safety.

In aviation, for example, this might mean getting faster at clearing the skies for new drone technologies.

Or, for the Office for Nuclear Regulation, we’re investing more than a quarter of a million pounds in a project that will enable the nuclear industry to test new AI tools in nuclear plants.

Including things like making us more efficient at handling high risk nuclear waste.

I want to personally thank the team at our Regulatory Innovation Office, set up last year, for being so forward-looking on this. I’m immensely proud of the work that it’s already doing.

So that’s where we are today.

A little over a year after I first set foot in the Department for Science, Innovation, and Technology, I don’t think a single person in this room could say, hand on heart, that we haven’t got stuck in.

And it’s starting to pay off.

In our first 12 months of in government we’ve attracted more than £44 billion worth of investment into the British AI sector. The average deal last year was worth £5.9 million.

And we have doubled the number of AI firms in Yorkshire, Wales, the Midlands, and the North West compared to just 3 years ago.

We have learned the lesson of history: countries can only prosper if they get the big calls right; if they decide to go beyond the expected and embrace the future; to innovate not imitate; refusing to be constrained by the problems of today by taking on the challenges of tomorrow.

In these uncertain times, I am certain that’s what it takes to get a global competitive edge.

So, if there is anyone here who still doubts our commitment.

My message to you is simple:

Britain is preparing for the challenge of the new technological revolution.

We want you to keep investing here, keep building here. List here. Scale here.

And if you invest in Britain, you will share in that competitive edge.

I look forward to working with you all as, together, we create the security and opportunity society people are counting on us to deliver.

Thank you.