A clear majority of voters are not confident in the Government’s ability to protect the public from the dangers of artificial intelligence (AI), a poll has suggested.
The More In Common survey appears to demonstrate how great a task ministers have given that it was carried out as Rishi Sunak hosted the first ever global AI safety summit in Bletchley.
The gathering has been praised by some as a success after tech giants agreed they will not release any more powerful AI models until their safety has been tested by governments, although the Prime Minister in conversation event with X boss Elon Musk has been criticised elsewhere.
But the poll, carried out during the summit, showed that nearly six in ten (59 per cent) of voters are not confident the Government can protect them from threats posed by AI, with just 29 per cent confident, giving an overall net score of -30 per cent.
Voters were even less confident about the Government’s ability to protect them from low level crime (-45 per cent), violent crime (-34 per cent), climate change (-36 per cent) and online fraud (-38 per cent).
The public also lacks confidence in the Government’s ability to protect the UK from another pandemic like Covid (-30 per cent).
But voters were more confident that Whitehall can protect people from terrorism (-7 per cent) and nuclear war (-12 per cent).
Luke Tryl, UK Director of More in Common, said: “More than anything else the public think the core job of Government is to keep people safe, but with the Covid Inquiry dominating the news and scenes of terror in the Middle East the public are sceptical about their Government’s ability to manage major threats.
“No 10 may well be particularly disappointed that far from having reassured the public that we control Artificial Intelligence (AI), six in ten people instead say they’re not confident that the Government will keep them safe from the risks posed by this emerging new technology.
“We know that the public is worried about artificial intelligence and what it means for their jobs, safety and security.
“If this week’s AI conference was about reassuring the public that the Government has got a handle on the risks of AI, our research shows they’ve still got a long way to go.”
Responding, a Government spokesman said: “The UK is leading the way on AI safety, securing a landmark commitment to safe and responsible AI development from 28 countries, the Bletchley Declaration, at the first-ever AI Safety Summit which we hosted.
“The summit also secured agreement on the roles to be played by government and AI companies in AI safety testing, while the launch of the world’s first AI Safety Institute here in the UK means we continue to set the pace globally, on this critically important issue.”
:: More In Common, a member of the British Polling Council, surveyed 2,042 British adults from 31 October to 2 November.