Rishi Sunak says AI should not be viewed as a threat to workers’ jobs ahead of meeting with Elon Musk – as it happened

• Rishi Sunak has insisted that workers should not view AI as a threat to their jobs. He made the comment in a press conference at the end of his AI safety summit (see 4.21pm) where he also announced the creation of an AI Safety Institute to “carefully test new types of frontier AI before and after they are released to address the potentially harmful capabilities of AI models”. And he said governments and companies have agreed a new strategy on AI safety testing. In a news release summarising it, No 10 says:

In a statement on testing, governments and AI companies have recognised that both parties have a crucial role to play in testing the next generation of AI models, to ensure AI safety – both before and after models are deployed.

This includes collaborating on testing the next generation of AI models against a range of potentially harmful capabilities, including critical national security, safety and societal harms.

They have agreed governments have a role in seeing that external safety testing of frontier AI models occurs, marking a move away from responsibility for determining the safety of frontier AI models sitting solely with the companies.

• Matt Hancock told officials that he – rather than the medical profession – “should ultimately decide who should live or die”, if the NHS was overwhelmed during the pandemic, the Covid inquiry has been told by the former head of NHS England.

• Chris Wormald, the permanent secretary at the Department of Health and Social Care, has told the Covid inquiry that a full lockdown may have been avoided if the government had issued its non-statutory stay at home request earlier. (See 3.18pm.) The inquiry was also shown a message from Mark Sedwill, the then cabinet secretary, showing that on 12 March 2020 – less than a week before Boris Johnson ordered the nation to stay at home – Sedwill was pushing a “herd immunity” strategy and saying “we want people to get [Covid]” (just not all at once). (See 2.59pm.)

• Andy McDonald, the suspended Labour MP, has said he’s suing the Conservative MP Chris Clarkson, who accused him of “seeking to justify the murderous actions of Hamas” in the Israel-Hamas conflict.

Bruno Le Maire, the French finance minister, has described the AI safety summit as “a great success for the UK”, Christina Gallardo from Sifted reports.

Was it a mistake not to invite more AI startups to the #AISafetySummit?
“I strongly believe this summit has been a great success for the UK and for the prime minister Rishi Sunak,” @BrunoLeMaire replies.

Dominic Cummings, Boris Johnson’s former chief adviser, says he has been vindicated by the publication of a message at the Covid inquiry today showing that on 12 March 2020 – less than a week before Johnson ordered the nation to stay at home – Mark Sedwill, the cabinet secretary, was pushing a “herd immunity” strategy and saying “we want people to get [Covid]” – just not all at once. (See 2.59pm.) Cummings was one of the No 10 figures most influential in pushing for an alternative, lockdown strategy. He said on X:

When I first revealed the ‘chickenpox parties’ horror in 2021 remember No10 lied and said never happened, Cummings liar etc.

Today we found out the Perm Sec in charge of fighting covid was pushing chickenpox parties ‘in every meeting’ & to the country’s most senior official, who then tried to get the PM to announce the parties to the nation.

This explains a massive amount of why wave 1 was a total shitshow but because it makes clear the deep state was responsible for this, not Boris, it will probably get buried.

Even if you support herd immunity plan, the whole point was to DELAY - not ACCELERATE WITH PARTIES!!!

This was on any objective basis an insane plan & is a catastrophe for the civil service

Cummings also claims this shows why the nation should be grateful to him and his Vote Leave allies in No 10.

all of you fuckers who whine about Vote Leave face this fact -

on 12 March the DHSC Permanent Secretary was pushing chickenpox parties on the Cabinet Secretary who tried to get the PM to say this to the country that day - AND VOTE LEAVE SPADS STOPPED THE INSANE FUCKING PLAN

you can all say ‘thanks Vote Leave, thanks Ben Warner, thanks weirdos & misfits’ in replies

let’s hear it

Cummings may be anxious to see Vote Leave get some approval because polling shows that voters now are far, far more likely (by a margin of six to one) to view Brexit as “more of a failure” than “more of a success”.

But his argument does not work because, without Brexit, Johnson may well not have been PM in 2020, and the handling of Covid would probably have been in more competent hands.

Rishi Sunak has arranged to hold a Q&A with Elon Musk later today. It will be shown on X (formerly Twitter), which is owned by Musk, at some point after 9pm.

Given Musk’s reputation as a maverick, there may have been some nervousness in No 10 about setting up the encounter and this tweet from Musk won’t help. Maybe that “sigh” signals disapproval, but he is publicising a cartoon gently mocking the AI safety summit, which Sunak has just described as something that will “tip the balance in favour of humanity”.

A suspended Labour MP is suing the Conservative MP Chris Clarkson, who accused him of “seeking to justify the murderous actions of Hamas” in the Israel-Hamas conflict. Andy McDonald, who represents the Middlesbrough constituency, said much of what he has said over the last few days about recent events in Israel and Palestine had been “deliberately distorted and misinterpreted”. Aletha Adu has the full story here.

Q: China was invited today. But it was not invited to your session today. So why should people trust them to comply?

Sunak says that is a glass half-empty take. He says China came, and signed up to the declaration. Inviting China was not an easy decision, he says. But he says he was taking a decision for the long term, and any serious conversation has to involve the leading AI nations.

And that’s the end of the press conference.

Q: You have said that people should not lose sleep about the lack of AI regulation. At what point should people start losing sleep?

Sunak says this summit has seen the world coming together. More is being done internationally than ever before. That is a significant achievement, he says.

He says AI can do incredible things. Every pupil should be able to have a personalised AI tutor.

But he says he wants that to happen with safeguards in place.

Q: Shouldn’t the government be legislating to control AI?

Sunak says government needs to act quickly. That is why he held this summit this year.

But legislating takes time, he says.

And, before you regulate, you need to do the research first.

He says the safety institute will do this work. It will provide the evidence base for legislation.

Other countries are also researching this, he says.

Ultimately, binding requirements will probably be necessary, he says. But those should be based on “empirical evidence” that will come from the testing.

Sunak does not mention Labour, but this answer is a reply to the challenge from the opposition released this morning. (See 9.18am.)

UPDATE: Sunak said:

The lesson is that we need to move quickly and that’s what we’re doing.

The technology is developing at such a pace that governments have to make sure that we can keep up.

Now, before you start mandating things and legislating for things I think A) by the way that takes time and we need to move faster, and we are, but secondly you need to know exactly what you’re legislating for and that’s why our safety institute is so important.

So far we’ve got the cooperation we need, but, of course, I think everyone would acknowledge, ultimately, binding requirements will likely be necessary, but it’s important that we do those in the right way and that needs to be based on empirical evidence that we’ll get from our testing.

View image in fullscreen

Q: Do leaders need to be more candid about the risks AI poses to jobs?

Sunak says he knows this is an anxiety people have. He says we should look at AI as “a co-pilot” more than something that will replace workers. It will help people do their jobs more effectively.

He says that is how he thinks of it.

Technology always has the ability to change labour markets. It is hard to predict.

But AI has already created 50,000 jobs.

He thinks technology like AI that enhances productivity will be good for the economy.

He says he also wants to focus on education, so that people get the skills they need.

UPDATE: Sunak said:

I know this is an anxiety that people have. We should look at AI much more as a co-pilot than something that necessarily is going to replace someone’s job. AI is a tool that can help almost everybody do their jobs better, faster, quicker, and that’s how we’re already seeing it being deployed.

I’m of the view that technology like AI which enhances productivity over time is beneficial for an economy. It makes things cheaper, it makes the economy more productive. But that doesn’t mean jobs can change.

View image in fullscreen

Q: What will you discuss with Elon Musk tonight? And is it not going out live because you are worried what he will say?

Sunak says Musk is a leading actor in AI. He has been concerned about this for a long time; he started talking about the risks 10 years ago.

But he does not just want to focus on Musk, he says. He says more than 100 leading AI nations were represented here.

View image in fullscreen

Sunak says the UK is ahead of any other country in creating the tools to keep people safe.

Under the agreement today, AI models will be tested before they are released.

He says people should be reassured by this.

He concludes:

The late Stephen Hawking once said ‘AI is likely to be the best or worst thing to happen to humanity’. If we can sustain the collaboration we’ve fostered over these last two days, I profoundly believe that we can make it the best.

Because safely harnessing this technology could eclipse anything we’ve ever known.

If, in time, history proves that today we began to seize that prize, then we will have written a new chapter worthy of its place in the story of Bletchley Park and, more importantly, bequeathed an extraordinary legacy of hope and opportunity for our children and generations to come.

Rishi Sunak is speaking now.

He says this summit has considered the risk from AI. And he thinks it has “tipped the balance in favour of humanity”.

People said China should not be invited, and they would not reach an agreement. But they were wrong, he says. He says every country at the summit has signed the Bletchley Declaration.

He says Yoshua Bengio has agreed to produce the inaugural report from the new international panel on AI safety.

And he says a landmark agreement on testing the safety of new AI models has been agreed.

Rishi Sunak is about to hold a press conference at the end of the AI safety summit at Bletchley Park.

View image in fullscreen