Rishi Sunak has vowed to tackle fears around artificial intelligence (AI) “head-on” as the government prepares to host a world-first global safety summit.
The prime minister has made taking advantage of the fast-improving technology, while also regulating against its potential dangers, a focus since taking office last year.
In a speech on Thursday, he announced he would establish the world’s first AI safety institute in the UK to explore “all the risks – from social harms like bias and misinformation, through to the most extreme risks”.
And he said their work would be shared around the world, for both “moral” and “economic” benefits.
Politics latest: Potential by-election headache for Sunak
Mr Sunak’s pledges came as the government published a first-of-its-kind paper into its capabilities and risks.
The prime minister acknowledged the “dangers and fears” AI brings – with concerns around misinformation and deepfakes, job losses, and a threat to human life itself all cited by experts in recent months.
In his speech, he said that while the technology offered opportunities for economic growth and advances in human capability, it also brought new challenges.
“The responsible thing for me to do is to address those fears head-on,” he said.
“Giving you the peace of mind that we will keep you safe, while making sure you and your children have all the opportunities for a better future that AI can bring.”
Highlighting the positives of the technology, the prime minister pledged a further £100m of investment to “accelerate the use of AI on the most transformational breakthroughs in treatments for previously incurable diseases”, such as aggressive cancers and dementia.
He added: “I believe nothing in our foreseeable future will be more transformative for our economy, our society and all our lives than this technology. But in this moment, it is also one of the greatest tests of leadership we face.
“It would be easy to bury our heads in the sand and hope that it’ll turn out all right in the end, to decide it’s all too difficult or that the risks of political failure are too great to put short term demands ahead of the long term interests of the country.
“But I won’t do that. I’ll do the right thing, not the easy thing. I’ll always be honest with you about the risks. And you can trust me to make the right long term decisions.”
AI may be transformative, but it can’t generate an election win for Sunak
In a speech in central London this morning the prime minister was in his element.
Discussing the risks and opportunities posed by artificial intelligence and how the UK can be a global tech leader, Rishi Sunak was keen to explain how transformative AI could be to British society.
When it comes to technology, this is an area in which Rishi Sunak thinks he can shine; as one of the youngest prime ministers in British history, Number 10 believes he will be able to show off his fluency in the sector and therefore offer something Sir Keir Starmer – and even other world leaders – cannot.
But that political calculation must be seen in the context of the Conservatives’ dire poll ratings, bruising by-election losses in Tamworth and Mid-Bedfordshire (and another by-election most likely on the way in Wellingborough), and NHS waiting lists remaining high.
Moreover, the ongoing war in the Middle East continues to overshadow government business: aptly demonstrated when, after his speech, a member of the media asked about the conflict between Israel and Hamas and the UK’s response to it.
That is not to say that AI will not have a significant impact on British politics – we only have to look to the recent deepfake audio recordings of the Labour leader that circulated online at the beginning of his party conference to see how elections could be impacted.
In his speech, Mr Sunak claimed he was “doing the right thing, not the easy thing” on dealing with AI – but with the cost of living crisis continuing to plague British households and war in the Middle East, how much will the public sit up and take note?
Next week’s AI Safety Summit at Bletchley Park, home of Britain’s Second World War codebreakers, will see other world leaders and tech bosses unite to discuss how the technology should be regulated.
It will focus on the misuse of it, such as to carry out cyberattacks or develop bioweapons, and the threats posed by the loss of control of AI.
Talks will also cover its impact on wider society, such as in elections.
Mr Sunak defended inviting China to the summit, despite criticism from some on his own backbenches, saying it was “not an easy decision… but I do believe it is the right thing”.
The prime minister added: “China is unquestionably the world’s second AI power behind the US, that’s just a fact.
“And AI doesn’t respect borders, so if you are seriously trying to address this issue, it is right to try and engage with AI’s leading powers.”
Sunak must make concrete decisions on AI approach
Rishi Sunak’s government is determined to be seen as a leading force in artificial intelligence and next week’s summit at Bletchley Park is the prime opportunity for them to do it.
But it is not enough to tell us that the benefits of AI need to be balanced with the risks.
To make clear where the UK government stands on AI, we need officials to explain how they are going to prioritise these benefits and the value system they will use to evaluate the safety risks.
Mr Sunak has announced an AI safety institute that will evaluate new types of AI so we know what the models are capable of, but we still don’t know how much power this institute will have to influence government decision-making based on its findings.
The EU is leading the way in terms of regulation, with its draft AI act that is currently going through a legislative process – meanwhile the UK’s white paper, published in March, suggests the government hopes to adopt a more light-touch approach.
But if the country is to be a tech superpower by 2023 as Sunak has pledged, concrete decisions need to be made.
Mr Sunak has already met several leading figures working in AI development, including the bosses of ChatGPT creator OpenAI and UK-based Google DeepMind.
Thursday’s paper, which includes assessments from UK intelligence agencies, will form part of the discussions.
Technology Secretary Michelle Donelan will open the summit next Wednesday, 1 November.