The UK’s AI minister has warned we must make sure computers allow humans to stay in charge – as he said bots are likely to breach our cyber defences at some point.
At a press conference in London, Kanishka Narayan flagged an incident earlier in the year where an AI agent broke into an Australian government site.
The bot was ‘looking for information about public spending on medicines’, he said, and ‘found another way in’ when it arrived on a website where its access was restricted.
Narayan said: ‘Its task was benign… but without being asked to do so, the agent gained access to information that [it] was not authorised to have.
‘The Australian government has said that no personal medical records are known to have been accessed, and there is no evidence of wider compromise, but the incident should concern us.’
He added: ‘It concerns me. It seems likely, if not inevitable, that our own cyber defences will be breached.’
Make sense of the mad world of British politics
Want to understand how what goes on in Westminster actually affects your life? Get analysis of what the biggest stories mean straight to your inbox every week. Sign up here.
The minister urged those creating the tech to ‘build machines that serve us’ and ‘leave us human beings in charge’.
Narayan also confirmed that scientists at the UK’s AI Safety Institute had access to the latest generation of models and were conducting safety tests.
It represented a reversal from last month, when the White House blocked companies including Anthropic and OpenAI from sharing their models with the facility.
The government is currently trying to decide which role the UK will play in a rapidly changing world where AI models from China and the US dominate.
King Charles convened a private summit of top AI bosses at Dumfries House in Ayrshire last month, and Prime Minister Andy Burnham has vowed to make it the main topic of next year’s G20 in Manchester.
Speaking on podcast The Rest is Politics, Burnham suggested the UK can be an ‘honest broker’ between those calling for greater regulation and those wanting to move at a rapid pace.
He said: ‘I see an opportunity here […] to use the G20 presidency to sit in the middle of that and say, look, AI for good, absolutely. You don’t want to put the brakes on something that could also bring huge human benefit, which it can particularly in the sphere of health.
‘But at the same time, you have to have the guardrails, and you have to have the principles that underpin it. And that’s what I will seek to do in the role of UK presidency.’
The PM suggested the summit could be used to develop a ‘global set of standards, a global code’ that leaders could take back to their own countries to guard against the risks.
Among those warning about the lack of protections in place was president of Microsoft Brad Smith, who told BBC News that milk and washing machines are subject to more safety rules than AI.
Smith described the technology as both a ‘tool and a weapon’, and called for companies to install ’emergency brakes’ to stop it getting out of control.
Get in touch with our news team by emailing us at webnews@metro.co.uk.
MORE: PS5 can now run physical PS2 game discs with vibe-coded emulator
MORE: Andy Burnham says he will release all details of meetings with Man City owners
MORE: Andy Burnham confirms Russia interfered in Brexit vote