Firstly let me say an enormous thanks to Onward.

I stated once I first launched this new and thrilling Department that individuals might count on to see a continuing drumbeat of motion from my officers, from my ministers and my whole staff…

What I didn’t count on was for Onward to take that problem on too!

From your good report on generative AI earlier this yr, to Allan Nixon’s Wired for Success Report which gave us insights into DSIT that illumined Whitehall.

And I hear via the grapevine that Onward are publishing one other AI-focused report within the coming weeks, so I look ahead to studying that.

It is protected to say that the world is now unsleeping to the alternatives and the challenges posed by synthetic intelligence.

In the final 3 years alone, MPs have talked about Artificial Intelligence extra occasions within the House of Commons than they did in all the 30 years earlier than that.

And once I stood as much as give my first speech because the Secretary of State for the Department for Science, Innovation and Technology, I made it clear that we had been going to be totally different and we had been going to do issues otherwise.

So, with regards to AI, I feel it’s particularly vital as a result of we can’t afford for DSIT to be a traditional ‘business as usual’ Whitehall division.

Nor can we ship extraordinary issues with out extra extraordinary folks in our Departmental staff.

We have to be as agile, progressive and as accessible identical to the entrepreneurs and companies and researchers that we symbolize.

And I’m happy to report that we have now carried out precisely that.

Over a matter of months we have now added world-renowned AI consultants to our Department and Taskforce…

From famend AI professors like David Krueger and Yarin Gal…

To one of many Godfathers of AI, Yoshua Bengio….

With our abilities, Frontier AI Taskforce and our international AI Safety Summit, it’s clear that the UK is completely positioned to guide the worldwide cost on seizing the alternatives of AI while gripping the dangers.

However, right now I wish to get past the statistics and go deeper into the philosophy that’s driving our method to AI.

Many of you’ll have heard me discuss AI security being the UK’s precedence, and the way we are able to solely really utilise the extraordinary advantages AI has to supply as soon as we have now tackled a few of the security challenges related to the frontier.

To some this will likely sound overcautious, or that we’re being pushed by worry of the dangers moderately than optimism concerning the alternatives. But really I feel it’s moderately concerning the alternatives that we’re specializing in.

Today, I wish to smash these myths head on.

Here at Onward, I wish to set out how the UK’s method to security and safety in AI will make it the most effective place on the planet for brand spanking new AI corporations to not solely develop however find right here.

Safety and safety are key to unlocking innovation.

The nation which tackles key AI security dangers  first would be the first to totally reap the benefits of the massive potential that AI has to offer.

That is why the UK is placing extra funding into these questions proper now than another nation on the planet.

Questions like “how do we prevent misuse of Frontier AI by malicious actors?”

And “how do we ensure we don’t lose human control and oversight of this new technology?”

And how can we shield our society together with our democratic system.

Think of how air journey – as soon as thought of a harmful new expertise by many – is now one of many most secure and most useful applied sciences in human historical past.

We received there by working with international locations internationally to verify we have now the correct security mechanisms in place –  and now all of us reap the advantages of flying safely. 

Safety is totally important to unlocking adoption throughout our financial system.

Boosting shopper confidence is what is going to make the distinction between a rustic taking a couple of years to undertake new expertise into their lives, or a couple of many years.

I wish to ensure that the UK is on the forefront of reaping the advantages of this transformative expertise.

Our method to AI has been counseled for being agile, open and progressive.  But we want extra analysis to information our method.

In many circumstances, we merely don’t perceive the dangers in sufficient element or certainty proper now as a result of this an rising expertise that’s growing faster than another expertise in human historical past.

It took mankind simply over a lifetime to go from the horse and cart to the house race.

Yet within the final 4 years massive language fashions have gone from barely with the ability to write a coherent sentence to now with the ability to move the bar examination and medical exams and who is aware of what massive language fashions have are set to carry.

So the tempo of improvement is quick and unpredictable which implies our focus must be on understanding the dangers.

And in life I do assume its vital to know the issue earlier than speeding to options. And with AI this ought to be no totally different.

That is why we established the Frontier AI Taskforce and have appointed a world-leading analysis staff to turbo-charge our understanding of frontier AI with skilled insights.

The Taskforce is already making speedy progress, forging partnerships with trade and growing progressive approaches to addressing the dangers of AI and harnessing its advantages.

The Global AI Safety Summit can also be a possibility to construct that understanding, share learnings and set up a community globally to work collectively to make sure our analysis can sustain with this transformative rising expertise.

Indeed, one of many key targets of the Summit is to kind an settlement on what the important thing dangers really are.

By bringing international locations, main tech organisations, academia and civil society collectively, the UK will lead the worldwide dialog on frontier AI.

The international nature of AI implies that worldwide concerted motion is totally important. AI doesn’t cease at geographical boundaries and nor ought to our method.

But in fact, we do additionally want to verify we have now the correct home method in place to drive protected, accountable AI innovation.

Earlier this yr we set out a principles-based method via the AI Regulation White Paper.

Our method is agile, focused and sector-specific.

We right here within the UK perceive that AI use circumstances are drastically totally different throughout totally different sectors.

A one measurement matches all system that treats agri-tech the identical means as army drones as a result of they each use AI is unreasonable and undeliverable.

By empowering present regulators to control AI in their very own sectors the place they’ve their very own experience, we have now created probably the most tailor-made and responsive regulatory regime anyplace on the planet.

We have additionally supported totally different sectors with a Central danger perform – horizon scanning.

Later this yr, we’ll publish a full response to our White Paper – exhibiting how our method is preserving tempo with this transformative expertise.

So, what does this all imply for small companies?

The regulatory method set out within the White Paper is particularly designed to be versatile, assist innovation and be certain that small, new and challenger AI corporations can develop and succeed right here within the UK.

And certainly, we’re already taking steps to ship on these goals by working with the Digital Regulation Cooperation Forum to pilot a brand new advisory service for AI and digital innovators so corporations can convey their merchandise extra rapidly and safely to market.

We need the UK to stay some of the nimble and progressive locations on the planet for AI corporations to develop.

That is why it’s proper to reaffirm our dedication to our principles-based method to regulation while additionally taking daring steps to deal with dangers on the frontier,

investing in world-class analysis capabilities and dealing intently with trade and civil society to verify our AI governance method retains tempo.

Our proportionate and focused method will allow us to foster innovation and encourage corporations to develop to meet up with the frontier – as a result of we’re not pulling up the drawbridge –the truth is what we wish to do is give shoppers and the general public confidence to spice up adoption and it’ll guarantee we are able to seize the alternatives safely.

Far from a race to the underside, the important thing AI builders internationally and right here within the UK are telling me they’re on the lookout for international locations the place they’ll have certainty, readability and assist with regards to constructing and deploying AI safely.

They desire a mature, thought of and agile method to AI that maximises the potential for innovation by mitigating the dangers.

They wish to open up in international locations the place shoppers are open-minded and enthusiastic about utilizing their AI instruments to enhance their lives, which is why with the worldwide AI Safety Summit, we’re not solely speaking about dangers, but additionally speaking about alternatives for the advantage of mankind.

And the proof is within the pudding Open AI, Anthropic have opened their worldwide places of work right here.

I need the UK folks to make use of AI with the identical confidence and lack of worry as we do after we e book an airplane ticket.

And I need AI corporations to know that the UK is probably the most up-to-date, agile and economically rewarding place on the planet to construct their enterprise in.

So, to all of the smaller AI corporations on the market, let me ship this message out to you right now: the UK is and can stay probably the most agile and progressive place so that you can develop your small business.

Safety on the frontier means prosperity throughout the sector.

We will grip the dangers in order that we are able to seize the alternatives.

Thank you.