First Secretary of State Louise Haigh is set to state that the UK government must listen to the concerns raised by leading artificial intelligence developers regarding safety risks. Speaking at the Trades Union Congress (TUC) on Tuesday, she emphasized that the government is prepared to collaborate with international partners to prioritize public safety and national security.
Her comments follow a surge of alarm from executives and former staff within the AI sector, who have publicly warned that the technology poses a potential threat to humanity. Prime Minister Andy Burnham welcomed the open dialogue about these dangers, noting that it is positive that companies developing advanced AI systems are discussing risks alongside opportunities.
Haigh acknowledged the dual nature of the technology, stating that while AI has the potential to transform public services, strengthen businesses, and drive scientific breakthroughs, the warnings from those at the forefront of its development cannot be ignored.
The political response comes as UK lawmakers push for new legislation following a report linking AI to numerous human rights abuses. Alex Sobel, a Labour MP and committee chair, criticized the current regulatory framework, stating that no country, including the UK, has a legislative approach fit for purpose.
Tensions have escalated after Anthropic researcher Evan Hubinger suggested there is greater than a 10% chance that AI could kill all humans within the next decade. This followed claims by former Anthropic employee Jacob Coxon, who told the BBC that failure to slow progress could lead to immediate existential danger. Anthropic CEO Dario Amodei has called for stricter monitoring and a slowdown in model development, a stance supported by OpenAI’s Sam Altman and Elon Musk.
While some industry figures argue these fears are exaggerated or designed for hype, proposals for an AI “kill switch” have gained traction. Anthropic co-founder Jack Clark suggested that lawmakers may need to enforce mechanisms to shut down dangerous software, though he noted existing companies already have internal methods to do so.
However, the UK government has rejected the idea of a legal kill switch. A Cabinet Office spokesperson explained that blocking access to models in the UK would not prevent their development or misuse abroad, adding that the government “cannot simply turn AI off.” Meanwhile, US President Donald Trump has dismissed AI safety concerns as a “hoax,” criticizing the narrative on his social media platform Truth Social.
10 percent chance of extinction in ten years is terrifying. We need regulations now, not empty promises from ministers.
How do we enforce safety if other countries don’t cooperate? International collaboration sounds good, but will it actually happen?
I can’t believe Trump called safety concerns a hoax. These are legitimate existential risks that politicians can’t just ignore.
It is smart that the government isn’t bowing to panic. A kill switch is impractical when code spreads globally.