UK government rejects calls for AI ‘kill switch’ amid growing fears of its risks

UK government rejects calls for AI ‘kill switch’ amid growing fears of its risks Britain ‘cannot simply turn AI off’, the Cabinet Office said on Friday - Bookmark - CommentsGo to comments Proposals to introduce a so-called "kill switch" to protect against an attack from rogue artificial intelligence have been turned down by Andy Burnham’s government. On Friday, the Cabinet Office voiced its resistance to establishing a new "last-resort power" that would permit the state to shut down an autonomous model during an emergency, stating the UK "cannot simply turn AI off". The position follows efforts by peers in Parliament in recent weeks to argue for such a safeguard amid escalating alarms over technological risks, including warnings from experts that the technology could wipe out humanity within the decade. Earlier in September, Liberal Democrat Lord Clement-Jones submitted an amendment to the Cyber Security and Resilience Bill creating a legal mechanism for state intervention should a model malfunction. He said this would allow for "a necessary emergency kill switch" in situations where the Government "currently lack clear statutory power to direct an emergency shutdown". But in a statement on Friday, a government spokesperson said: “Britain cannot simply turn AI off and 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. “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.” Officials said the Government is exploring whether further targeted interventions may be needed to address the most significant risks linked to the technology, but that any measures would need to be proportionate. Earlier this week, the Prime Minister acknowledged “risks to national security” posed by AI after a top safety researcher at Anthropic said he believed the technology could “kill all humans” in the next decade. Mr Burnham has boosted the role of AI safety minister Kanishka Narayan by giving him a seat at the Cabinet table, a move which he said on Wednesday had “lifted the level of the conversation about artificial intelligence within Government”. “The offer is open to colleagues on all sides of the House to be part of that crucial discussion,” he told the Commons at Prime Minister’s Questions. It came after Evan Hubinger, alignment science lead at Anthropic, laid out his concerns about the potential advancement of AI in a post on social media. “We really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade,” he wrote on X. “I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to.” His post came in response to another Anthropic researcher, Jacob Coxon, warning that a race between companies towards self-improving superintelligence was “gambling with our lives”, as he resigned from his role at the firm. Announcing his departure, Mr Coxon said that Anthropic and OpenAI, its main competitor, were not “acting responsibly” and that people outside of AI labs were underestimating the power of the technology. “The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt,” he said. “No other human activity poses this level of danger.” It comes after the Financial Times reported the US tech giant had withheld its latest model from the UK’s AI Safety Institute, one of the world’s leading bodies for testing the risks of the technology. Claude Mythos 5.1 was only made available to vetted American organisations, according to the paper. Anthropic and OpenAI have been contacted for comment. Join our commenting forum Join thought-provoking conversations, follow other Independent readers and see their replies Comments