The big names of AI and government were there, but will it make a difference?James O’Malley explores the impact of the first AI Safety Summit
At the start of November, Britain was briefly the centre of the tech world as high-ranking politicians, Silicon Valley luminaries and respected academics all descended on Buckinghamshire for what was billed as the world’s first intergovernmental summit on AI safety.
The meeting at Bletchley Park was the brainchild of Prime Minister Rishi Sunak, and was his dual attempt to establish Britain as a major player in the AI revolution and cement himself a political legacy.
On the strength of the guest list, it certainly appeared to be a success. Representing tech was OpenAI founder Sam Altman, X CEO Elon Musk, and Mark Zuckerberg’s top lieutenant and former Deputy PM Nick Clegg. The political attendees included US Vice-President Kamala Harris, EU President Ursula von der Leyen and, most controversially, China’s vice-minister of science and technology, Wu Zhaohui.
With so many important people in the same room, the summit surely couldn’t fail to deliver meaningful outcomes. Could it?
The Bletchley Declaration
Experts in the security industry claim the government made a rod for its own back with the way the gathering was billed. “The word ‘summit’ is a little problematic,” said Professor Ciaran Martin, chairman of cybersecurity firm CyberCX UK. “A ‘summit’ means a bunch of leaders staying up to four in the morning... and these exhausted people producing a document that has legal force.”
Alas, nothing quite so dramatic happened at Bletchley. Instead, it was more like an academic conference, a talking shop that concluded with the PM interviewing Elon Musk. That doesn’t mean it was a waste of time, however. “Overall, it was a very, very good thing to do because the alternative was nothing at all,” said Martin, who was the founding CEO of the British government’s National Cyber Security Centre and head of cybersecurity at GCHQ.
The fact that the summit got nations talking about the threat posed by AI is, in his view, a good thing. “Bureaucracies respond to the need for activity,” said Martin.
He said a community has been established that includes the “world’s major governments, it’s got most of the world’s major tech companies, and it’s got a bunch of academic experts and so forth who have looked at what the para
