British Prime Minister Andy Burnham has called on world leaders to establish global regulations governing the development and deployment of artificial intelligence (AI) to mitigate its potential risks. Speaking at the United Nations General Assembly in New York, Burnham emphasized the need for a coordinated international approach to ensure AI technologies are developed safely and ethically, while unlocking their economic and social benefits.
Burnham outlined his vision for the UK to play a central role in facilitating this effort, comparing the current moment to the global financial crisis when nations cooperated to stabilize the banking system. He proposed bringing together governments, private sector leaders, and experts to create a unified set of global principles and standards aimed at transparency, safety, and preparedness in AI development. The Prime Minister mentioned that AI would be a primary focus of the UK's upcoming presidency of the G20, offering a platform to advance these initiatives ahead of a planned summit in Manchester in November next year.
Highlighting Britain's position as an “honest broker,” Burnham noted the country’s expertise and existing collaborations in AI safety, including work by the UK’s AI Security Institute alongside American research institutions. However, he stressed the need to strengthen international cooperation, particularly to prevent hostile states from exploiting AI technology to spread disinformation.
Burnham’s call for regulation contrasts sharply with views expressed by former US President Donald Trump, who dismissed warnings about AI's risks as a “hoax” intended to slow US technological leadership. Meanwhile, several tech industry leaders and experts have expressed concerns about AI’s potential dangers. The head of alignment at Anthropic has suggested there is a notable chance that AI could pose existential risks within the next decade, a point underscored by ongoing debates on the need for strict controls.
Discussions around specific regulatory measures such as mandatory testing and “kill switches” for advanced AI systems have been contentious. A proposal backed by 20 countries, including Germany, Canada, and Ireland, urges that AI remain under human oversight and calls for compulsory evaluation of frontier models. The UK, along with the US and China, has yet to endorse this declaration. Critics, including experts from Britain’s Alan Turing Institute, have highlighted challenges in implementing emergency shutdown mechanisms for AI, noting potential difficulties in safely deactivating highly integrated systems.
Attempts to legislate AI “kill switch” laws have faced setbacks both in the UK Parliament and in US political arenas, despite ongoing advocacy from certain lawmakers and state officials. As the technology continues to evolve rapidly, Burnham’s urging for a collective and principled global framework reflects growing international concern about balancing AI’s transformative potential against its risks.