Elon Musk Calls for Third‑Party Oversight at the First AI Safety Summit in the UK and Highlights the Bletchley Declaration

Elon Musk at the first AI safety summit in the United Kingdom said he hopes to establish a “third‑party referee” that can monitor what leading AI companies are doing and at least raise an alarm when concerns arise.

Musk described artificial intelligence as a “double‑edged sword,” estimating an 80% chance that the technology will benefit humanity and a 20% chance it could be dangerous.

He emphasized that AI is one of humanity’s greatest threats, noting it is the first time we face something far smarter than us and it is unclear whether we can control such technology, but he believes we can strive to guide it toward outcomes beneficial to humans.

“I’m not sure what fair rules are, but before supervision you must start with insight,” Musk said.

The United Kingdom yesterday published a declaration signed by 28 countries, including China, the United States and the European Union, known as the “Bletchley Declaration.”

This is the world’s first international statement on AI, aimed at addressing concerns about future powerful AI models posing existential threats to humanity and worries about AI currently amplifying harmful or biased information.

At the summit, China’s Vice Minister of Science and Technology Wu Zhaohui led a delegation, participated in discussions on AI safety, promoted China’s “Global AI Governance Initiative,” and will hold bilateral talks with other nations. The initiative systematically outlines China’s approach to AI governance across development, safety, and governance.

In short, the 28 nations agree that AI poses potential risks to humanity; the declaration focuses on scientifically understanding these risks and establishing cross‑national policies to mitigate or avoid them.

Musk noted that people are very concerned that governments might rush to create rules before figuring out what to do, but that does not appear to be happening now.