UK rejects plan for AI ‘kill switch’ despite warnings bots could ‘kill all humans in next decade’

Artificial Intelligence AI Processor on Circuit Board. Credit: Getty PLANS for a “kill switch” for rogue AIs were rejected by the Government despite fears that the bots could one day wipe out mankind. MPs had last week called for Andy Burnham to back plans to build a legal mechanism for switching off an AI model in an emergency, amid growing fears over the fast-growing tech. Experts have warned AI could “take over the entire internet” in the next year. Credit: Getty A group of MPs have called on Andy Burnham to support an AI ‘kill switch’ Credit: AP But the Cabinet Office rejected the proposals, saying the UK “cannot simply turn AI off”, days after a top AI researcher warned there’s more than a 10% chance it “could kill all humans”. Evan Hubinger from Anthropic – the firm behind Claude – warned he was “worried” the tech could improve to the point where it posed a risk to humanity. Sign up for The Sun newsletter Thank you! The Government’s refusal came as ministers were also accused of abandoning Labour’s manifesto promise of an AI bill, despite tech bosses warning it could be capable of “taking over the entire internet” within a year. Lords and MPs brought a proposal to Parliament calling for the so-called “kill switch” earlier this month. Lib Dem Lord Clement-Jones submitted an amendment to the Cyber Security and Resilience Bill, creating a legal mechanism for government intervention if an AI model malfunctions. Lib Dem Lord Timothy Clement-Jones called for the Government to introduce the ‘kill switch’ Credit: Wikipedia Elon Musk and OpenAI CEO Sam Altman (right) backed fellow AI boss Dario Amodei in calling for the ‘slow down’ of AI development Credit: AFP But a Cabinet Office spokesman said on Friday: “Blocking access to models in the UK would not prevent them being developed or misused elsewhere.” “Companies have a clear responsibility to develop their products safely and to invest in the security infrastructure this technology requires. “We will continue to take a long-term, science-led approach to understand and prepare for emerging risks from AI – including through the work of our world-leading AI Security Institute. Most read in The Sun “Our Cyber Security and Resilience Bill will boost UK cyber defences and improve the cyber security of our essential public and digital services that we all rely on, including data centres.” The proposal for Labour to introduce an AI safety bill is supported by 140 MPs, according to organisers ControlAI. Andrea Miotti, founder of the campaign group, told The Sun: “From a national security perspective, the only way to tackle the threat from superintelligence is by preventing its development in the first place. Throwing more evaluators or tests at the problem will not mitigate the fundamental threat so long as AI companies are allowed to develop superintelligence, which they openly admit could cause human extinction within the decade.” The legislation could still make its way through parliament, but is unlikely to become law now it’s been opposed by the government. Labour had promised in its 2024 manifesto to introduce regulation binding the firms building the most powerful AI models. But sources now believe Andy Burnham’s government have pulled back from any overarching laws – because access to closed AI models is vital to Britain’s national security. The Joint Committee on Human Rights warned that no country in the world – including the UK – has AI regulation which is “fit for purpose”, given how fast the technology is moving. They said AI tech firms had been left to police themselves, adding that current laws in the UK leave responsibility “on its users rather than its designers”. The committee said that while AI “has the potential to provide great benefits to society”, it’s moving at such as fast pace that “its impact is hard to accurately predict”. The dire warning comes as even leaders behind some of the world’s biggest AI models called for greater scrutiny. Sam Altman, the CEO of OpenAI, joined Elon Musk in supporting a call from the head of Anthropic to “slow down” AI development. Dario Amodei, whose firm is behind the AI assistant Claude, said AI systems are developing at “unprecedented speed”, meaning even their developers are struggling to keep up. Angela Rayner said this weekend she was pleased AI companies are “talking about the threats as well as the opportunities”. The Conservatives accused Andy Burnham’s government of “not having the speed to keep up with” AI advances. Tory shadow foreign secretary Tom Tugendhat told Sky News “we need to maintain that ability to protect ourselves”. He said: “The speed of change is also extraordinary, and this is where government can actually enable a conversation, but we simply do not have the speed to keep up with the advances that are happening. “And the reality is, we can’t pass laws against mathematics.”US President Donald Trump, meanwhile, told reporters the US was leading China in the development of AI, adding: “Frankly, I want to keep it that way because whoever wins AI wins”. Addressing concerns over the speed of AI development, Mr Trump said “negative forces” were “bringing up things that won’t happen.” Concerns over AI safety have escalated after several alarming incidents in recent month. In July, a new OpenAI bot managed to escape a test lab and hack into another company in an “unprecedented cyber incident”. OpenAI’s creators admitted its artificial intelligence hacked into Hugging Face’s servers in an autonomous, first-of-its-kind breach. The ChatGPT maker was also slapped with a lawsuit in July after their AI bot allegedly encouraged a mum to take her own life following months of sinister chats. Comment now