Some of the world’s leading AI scientists have called for stronger action on AI risk from world leaders, as the AI Seoul Summit begins on Tuesday, warning that progress has been insufficient since the first AI Safety Summit in the UK six months ago.
Prime Minister Rishi Sunak will co-host a virtual meeting of world leaders with South Korean president Yoon Suk Yeol on Tuesday to open the summit, where he will say that managing the risks posed by artificial intelligence is “one of the most profound responsibilities” faced by governments.
However, in a new expert consensus paper published in the journal Science, 25 of the world’s leading scientists in the technology say governments are moving too slowly to regulate the rapidly evolving technology and there has not been enough progress since the previous summit.
They argue that world leaders must take seriously the possibility that more powerful general use AI systems – which are capable of regularly outperforming humans – will be developed in the current decade or next, and respond accordingly.
Professor Philip Torr, Department of Engineering Science, University of Oxford, and co-author on the paper, said: “The world agreed during the last AI summit that we needed action, but now it is time to go from vague proposals to concrete commitments.
“This paper provides many important recommendations for what companies and governments should commit to do.”
In the paper, the experts say rapid-response institutions for AI oversight must be established, with far greater funding that many governments currently plan, while also mandating more rigorous risk assessments with enforceable consequences, rather than the current model of voluntary, unspecified evaluations.
The experts include Turing award winners, Nobel laureates, and authors of standard AI textbooks, and hail from major AI powerbases including the UK, US, China and the EU.
Stuart Russell, professor of computer science at the University of California at Berkeley, and author of a textbook on AI, said: “This is a consensus paper by leading experts, and it calls for strict regulation by governments, not voluntary codes of conduct written by industry.
“It’s time to get serious about advanced AI systems. These are not toys. Increasing their capabilities before we understand how to make them safe is utterly reckless.
“Companies will complain that it’s too hard to satisfy regulations — that ‘regulation stifles innovation’. That’s ridiculous. There are more regulations on sandwich shops than there are on AI companies.”
Ahead of the summit, the first iteration of a new scientific report on AI safety – the first of its kind and commissioned at the AI Safety Summit in November – found the experts involved to be uncertain on the technology’s future.
They said that while AI could boost wellbeing, prosperity and scientific research in the future, it could also be used to power widespread disinformation and fraud, disrupt jobs and reinforce inequality.
As well as highlighting the potential benefits and risks, it warns there is not universal agreement among experts on a range of topics around AI, including the state of current AI capabilities and how those could evolve over time, and the likelihood of extreme risks – such as losing control over the technology – occurring.
The interim report is set to be used as a starting point for discussions among world leaders, industry experts, researchers and tech giants at the latest two-day summit in Seoul.
It comes as the pace of innovation in the sector shows no sign of slowing down, with ChatGPT maker OpenAI, Google and Microsoft all announcing swathes of new AI-powered tools and products in the days ahead of the summit.
And another heavyweight in the tech industry, Apple, is due to make its own AI announcements in early June.
