The AI Safety Summit will feature 100 attendees split roughly between international cabinet ministers, CEOs from top AI companies, academics and civil society representatives.
Matt Clifford, Entrepreneur First boss and a leading figure in planning the summit, revealed a handful of details about who will be in attendance at the world-first summit on AI safety.
Writing on X, formerly Twitter, Clifford said that given the small size of the Bletchley Park venue and the goal of holding a tightly focused summit, the invite list would be far smaller than major conferences such as COP or WebSummit.
“There’ll be around 100 attendees,” Clifford wrote. “This means making very tough decisions about invitations!”
Clifford said that a large volume of startups and investors were interested in attending, however, the corporate presence of the summit would be limited to companies that are “building the most powerful models”.
Rumoured AI company attendees include OpenAI, Anthropic, Stability AI and Graphcore.
Clifford claimed that the decision was not to “pull up the drawbridge”, but instead to ensure the “companies building systems with potentially dangerous capabilities” were properly scrutinised.
First announced in June, the AI Safety Summit will take place on 1 and 2 November in Bletchley Park, the former headquarters for World War II codebreakers.
The government has described the objectives of the summit as agreeing on a shared understanding of risks with international stakeholders and establishing a “forward process” for international collaboration.
Prior to the summit, a handful of AI events run by civil society groups and tech bodies are expected to take place. There have been minimal details released about these events, however, hosts will include the Alan Turing Institute, Royal Society, techUK and British Academy.