Rishi Sunak vows security services will check 'frontier' AI systems
Rishi Sunak vows government and security services will check ‘frontier’ AI systems before they are deployed as he warns companies cannot be allowed to ‘mark their own homework’
Rishi Sunak today vowed that government and security services will check ‘frontier’ AI systems before they are deployed.
The PM committed to providing the public with ‘independent assurance’ about the safety of new platforms as he wrapped up the first global summit at Bletchley Park.
Mr Sunak held a press conference at the wartime codebreaker base after a session with allies on how to handle the national security implications.
He said that ‘until now the only people testing the safety of new AI models have been the very companies developing it’.
The premier hailed a deal between ‘like-minded governments and AI companies’ to work together on testing before tech is released.
The UK and US are establishing AI Safety Institutes which he said will have ‘the public sector capability to test the most advanced frontier models’.
Rishi Sunak today vowed that government and security services will check ‘frontier’ AI systems before they are deployed
Rishi Sunak today took centre stage as he gathered world leaders and tech bosses to discuss the dangers of Artificial Intelligence on the second day of his Bletchley Park summit
Mr Sunak said: ‘Drawing on the expertise of some of the most respected and knowledgeable AI experts in the world our Safety Institute will work to build our evaluations process in time to assess the next generation of models before they are deployed next year.’
The PM added that that intelligence agencies would form part of the safeguarding operation.
‘Fundamentally governments… do need to provide their citizens with the independent assurance that the models are safe,’ he said.
Mr Sunak has hosted representatives from 27 countries alongside tech firms and civic society groups at the two-day summit.
He will be interviewed billionaire Elon Musk this evening in an event that will be put on the X social media site afterwards.
But while he tried to play down ‘alarmist’ fears, Mr Sunak also risked fueling concerns this morning by saying the dangers are on the same scale as ‘nuclear war’.
‘People developing this technology themselves have raised the risk that AI may pose and it’s important to not be alarmist about this,’ he said.
‘There’s debate about this topic. People in the industry themselves don’t agree and we can’t be certain.
‘But there is a case to believe that it may pose a risk on a scale like pandemics and nuclear war, and that’s why, as leaders, we have a responsibility to act to take the steps to protect people, and that’s exactly what we’re doing.’
Giorgia Meloni is one of the few national leaders to be attending Mr Sunak’s summit, with Joe Biden, Justin Trudeau, Emmanuel Macron and Olaf Scholz all among those staying away.
The Italian PM said she was ‘proud’ of her ‘friendship’ with Mr Sunak and expressed hope they would work together on AI to ‘solve the biggest challenge that maybe we have in this millennium’.
As she met the PM for a bilateral meeting at the Bletchley Park summit, Ms Meloni told Mr Sunak ‘your priorities are also mine’, citing migration and AI as among key issues
Mr Sunak said he was ‘delighted to be working so closely’ with European Commission president Ursula von der Leyen in discussing AI
The PM praised Ms Harris, who he hosted in Downing Street last night, for the executive order on AI made by the White House this week.
As she met the PM for a one-one-one meeting at the Bletchley Park summit, Ms Meloni told him ‘your priorities are also mine’, citing migration and AI as among key issues they are discussing.
She said the UK summit would be used as a ‘base’ for the G7 presidency held by Italy next year, where ‘artificial intelligence will be one of the main topics’ on the agenda.
Michelle Donelan, the Science, Innovation and Technology Secretary, earlier said that a sci-fi style loss of control of AI was her biggest concern.
‘That is a risk that is much more hypothetical in nature, that naturally is the one that I am most concerned about because it is the one that would result in the gravest ramifications,’ she said.
Put to her that was the ‘Terminator scenario’ – a reference to the Arnold Schwarzenegger film where machines take over the world – she said: ‘Well, that is one potential area where it could lead but there are several stages before that.’
Source: Read Full Article