Who’s going (and who’s not) to the AI Safety Summit at Bletchley Park?

Ahead of the AI Safety Summit starting tomorrow morning taking place outside of London in Bletchley Park, today, the U.K. government has confirmed more details about who is actually going to be attending the event. The list’s publication comes after weeks of speculation and criticism that the event’s line up — both in terms of topics and attendees — would fall short of giving a full representation of the different stakeholders and issues at play.

Organizers have said that some of the headline conversation topics will include the idea of catastrophic risk in AI; how to identify and respond to it; and establishing an agreed concept of “frontier AI”.

Depending on how close you think those risks are to reality, some of the ideas might appear more abstract, and less about some of the more specific and pressing worries people have voiced about the role AI is playing right now, for example in furthering misinformation, or offering a helping hand to malicious hackers looking for ways to break into networks.

As we wrote yesterday, the U.K. is partly using this event — the first of its kind, as it has pointed out — to stake out a territory for itself on the AI map — both as a place to build AI businesses, but also as an authority in the overall field.

That, coupled with the fact that the topics and approach are focused on potential issues, the affair feel like one very grand photo opportunity and PR exercise, a way for the government to show itself off in the most positive way at the same time that it slides down in the polls and it also faces a disastrous, bad-look inquiry into how it handled the Covid-19 pandemic. On the other hand, the U.K. does have the credentials for a seat at the table, so if the government is playing a hand here, it’s able to do it because its cards are strong.

The subsequent guest list, predictably, leans more towards organizations and attendees from the U.K.. It’s also almost as revealing to see who is not participating.

The 46 academic and civil society institutions include national universities such as Oxford and Birmingham (but not Cambridge); alongside international institutions like Stanford and several other U.S. universities (but not some you might have expected, like MIT); China’s Academy of Sciences will be present. Groups like the Alan Turing Institute, the Ada Lovelace institute, the Mozilla Foundation and the Rand Corporation will also be present.

Countries participating include the U.S. (represented by Vice President Kamala Harris); a number of European countries (but curiously none from the Nordics); Ukraine but not Russia (yes, it’s sanctioned, but that hasn’t stopped individuals from the country competing in sports, and AI academics are, basically, athletes of another kind). Brazil is the sole representative of Latin America among a smattering of countries from the Global South.

The 40 businesses include several heavy hitters like Google, Meta, Microsoft and Salesforce but not Apple nor Amazon (AWS will be there, however); OpenAI and Elon Musk’s X AI will also be there; players in the world of processors including ARM, Nvidia and Graphcore are participating; along with several startups. And it will also include a handful of multilateral organizations, including the United Nations and some of its agencies.