Technology, AI & Cyber
King Charles warns AI giants of 'existential dangers'
King Charles III warned senior technology figures in Scotland on Thursday of the “existential dangers” of artificial intelligence falling into the wrong hands.
The conference at Dumfries House brought together representatives from Nvidia, Google DeepMind, OpenAI and Anthropic, along with Britain's minister for artificial intelligence.
Charles told guests there was urgency in considering the dangers of such technologies being used “in potentially catastrophic ways”. He called on tech leaders to keep technology in the service of “humanity, community and the natural world”.
Buckingham Palace expects the conference to start a process to draw up common principles for the development and use of AI. No formal agreement is expected at the end of the meeting, which was set to last four to five hours. Charles was not taking part in the discussions.
Guests included Google DeepMind chair Demis Hassabis, Nvidia head Jensen Huang, OpenAI chief financial officer Sarah Friar, Vatican AI adviser Paolo Benanti and author Yuval Noah Harari.
Hassabis said the risk of something bad accidentally happening if AI development goes wrong was “definitely non-zero”. He cited recent cyber incidents involving rogue AI agents and warned of possible more serious risks, including biological and nuclear threats.
Huang emphasised AI's possible benefits in research and medicine and its potential to create jobs in new industries. He called for “responsible optimism”.
The meeting comes amid wider warnings and disputes over AI safety, job losses and the energy demand of data centres. Anthropic chief executive Dario Amodei called on companies on Saturday to slow development to better understand risks, with support from OpenAI chief executive Sam Altman and Elon Musk.
US President Donald Trump and several members of his government have said they do not want to constrain the sector, citing concern that China could gain an advantage. Meta head Mark Zuckerberg has also rejected calls to slow development.
Update
King Charles III delivered the warning at the opening of the conference.
The conference included representatives from Nvidia, Google DeepMind, OpenAI and Anthropic, as well as Britain's AI minister.
Buckingham Palace expects the meeting to begin a process to draw up common AI principles, but no formal agreement is expected at the end of the talks.
DeepMind's Demis Hassabis warned of possible cyber, biological and nuclear risks, while Nvidia's Jensen Huang called for responsible optimism.
Source note: AFP news report published on 17 September 2026 at 15:01:31 UTC.
Update
Buckingham Palace responded to Charles Spencer's claim about remarks by the then-prince after Princess Diana's death.
The item adds comments by Microsoft AI head Mustafa Suleyman, Pope Leo XIV and others to the wider AI-safety debate.
The Ditchley Foundation drew up a charter ahead of the summit, according to the Sunday Times.
Source note: AFP news report published on 17 September 2026 at 09:19:47 UTC.
Uncertainty notes
The meeting was not expected to produce a formal agreement, so any common principles remain prospective.
Source
AFP news report published on .