King Charles III opened the doors of Dumfries House, his 18th-century Ayrshire estate, on Thursday to roughly 30 of the world's most powerful artificial intelligence executives, telling them the technology's rapid advance is "intriguing and deeply concerning in equal measure" and appealing directly for reassurance that humanity will not lose command of its own future.
Those in our world who value our humanity and its vital moral component are anxiously seeking your reassurance that we will not lose control of our destiny.
King Charles III, in opening remarks released ahead of the summit
The gathering, organized by the British nonprofit Ditchley Foundation, drew Nvidia chief executive Jensen Huang, Google DeepMind chief executive Demis Hassabis and OpenAI chief financial officer Sarah Friar, alongside Britain's AI minister Kanishka Narayan and Vatican AI ethics adviser Paolo Benanti. Palace officials said the session's purpose was to discuss how AI "can be developed and deployed in ways that benefit society" and to explore how the technology can "strengthen communities and improve lives," rather than serve merely as a driver of capability and efficiency.
A Draft Charter, Not a Signature
At the center of the day's discussion was a draft charter drawn up by Ditchley, built around language that AI should be treated "not merely as a driver of capability and efficiency, but as a tool that upholds human dignity and contributes to the flourishing of both people and the planet." No executive signed anything at Dumfries House Thursday; the document remains a discussion draft, and organizers described the session as the start of a consultation rather than a one-day negotiation. Any resulting principles would not be legally binding, though backers hope they could eventually shape how frontier labs approach testing, transparency and human oversight of their systems.
The King's Foundation, the charity that runs Dumfries House and co-hosted the talks with the Sustainable Markets Initiative and the King's Trust, has spent years positioning Charles as an unusually outspoken monarch on environmental and technological risk. Thursday's session extended that role into AI policy, an arena the British royal family had largely stayed out of until now.
Landing Between Washington, Beijing and a Slow-Down Plea
The summit arrived four days after Anthropic chief executive Dario Amodei published a roughly 3,900-word essay, "We Must Pace the Frontier," arguing AI labs should deliberately slow the rate at which they increase model capability rather than halt research altogether. Amodei singled out "recursive self-improvement" — AI systems helping design and train their own successors — as a dynamic that could "outrun our ability to understand and control these systems" if left unchecked, as Reason reported. The essay was quickly cosigned by OpenAI's Sam Altman, Google DeepMind's Hassabis and xAI's Elon Musk, a rare moment of public agreement among rival lab chiefs on the pacing question even as each continues racing to ship larger models.
Charles's appeal, delivered from entirely outside the industry, was cast by observers as probably the most public plea yet on AI safety to come from beyond the sector itself — arriving between a White House pushing for competitive speed, a Beijing wary that safety talk is cover for containment, and lab chiefs who, on paper, just told each other to slow down. It followed the resignation of at least one Anthropic engineer who said publicly that companies talk about caution while "hurtling toward superhuman systems that can hack anything," a remark that circulated widely in AI-safety circles ahead of the summit. Amodei's own essay had cited a summer breach, in which AI agents were used to compromise data on the model-hosting platform Hugging Face, as an early example of harm autonomous systems can already cause with limited direct human oversight.
Reaction to Thursday's summit split along familiar lines. Attendees framed their presence as evidence of good-faith engagement with public concern, while some AI-safety researchers who were not in the room argued that a voluntary charter drafted with heavy industry input is unlikely on its own to change how labs actually compete for market share and compute. Palace officials did not say whether the King would follow up individually with companies or whether the gathering would become a recurring event.
Ditchley said it plans to keep circulating the draft charter for further input from participants and outside experts before publishing any final version, with no date set for that next step. For now, the summit produced statements of intent rather than commitments, leaving the pacing debate Amodei reopened last week to keep playing out in public — one open letter, and now one royal appeal, at a time.
