Rishi Sunak warns AI safety should be world priority alongside pandemics and war

The Prime Minister said he wanted to be “honest” with the public about the risks of AI, as he made a speech on the emerging technology.

As the Government published new assessments on artificial intelligence, Mr Sunak said they offered a “stark warning”.

“Get this wrong and it could make it easier to build chemical or biological weapons.

“Terrorist groups could use AI to spread fear and disruption on an even greater scale,” he said.

“Criminals could exploit AI for cyber attacks, disinformation, fraud or even child sexual abuse.

“And in the most unlikely but extreme cases, there is even the risk that humanity could lose control of AI completely through the kind of AI sometimes referred to as ‘super intelligence’.

“Indeed, to quote the statement made earlier this year by hundreds of the world’s leading AI experts, mitigating the risk of extinction from AI should be a global priority, alongside other societal scale risks such as pandemics and nuclear war.”

But he said that it was “not a risk that people need to be losing sleep over right now” and he did not want to be “alarmist”.

The Government will host an AI Safety Summit at Bletchley Park next week, bringing together world leaders, tech firms and civil society to discuss the emerging technology.

US vice president Kamala Harris is set to be among the senior figures attending the summit, and will deliver a speech on the US approach to AI on November 1 before attending the event on November 2.

UK officials do not see the planned speech by Ms Harris as in any way overshadowing the summit.

The Prime Minister’s official spokesman said: “We will work obviously extremely closely with the US…. The US and the UK have the leading companies in this space, so it is logical that both the US and the UK would be at the forefront of it.”

Ahead of the summit, Mr Sunak announced the Government would establish the “world’s first” AI safety institute, which the Prime Minister said would “carefully examine, evaluate and test new types of AI so that understand what each new model is capable of” and “exploring all the risks”.

He said tech firms had already trusted the UK with privileged access to their models, making Britain “so well placed” to create the world’s first AI safety institute.

The Prime Minister said the Government would use next week’s summit to push for a first international statement about the nature of AI risks, and said leaders should follow the example of global collaboration around climate change and establish a global expert panel on the issue.

But Mr Sunak said the Government would not “rush to regulate” AI, although he added that countries should not rely on private firms “marking their own homework”.

“Only governments can properly assess the risks of national security,” he said.

He also defended the decision to invite China to the AI Safety Summit, arguing there can be “no serious strategy for AI without at least trying to engage all of the world’s leading AI powers”.

“That might not have been the easy thing to do but it was the right thing to do,” he said.

Ahead of the Prime Minister’s speech, the Government published several discussion papers showing its evaluation of the risk of AI, which suggested that there were new opportunities for growth and advances, but also a range of “new dangers”.

The papers said there is insufficient evidence to rule out a threat to humanity from AI and that it is hard to predict many of the risks because of the broad range of potential uses in the future.

It adds that the current lack of safety standards is a key issue, and warns that AI could be used to carry out more advanced cyber attacks and develop bioweapons.

It also warns that human workers could be displaced by AI and both misinformation and disinformation could be spread more easily, and potentially influence future elections.