Dozens of countries vow to share notes on AI

World leaders from every corner of the globe agree: AI can cause “catastrophic harm” if not managed correctly.

Representatives of 28 countries signed on to a document called the Bletchley Declaration this week, laying out how they’ll collectively navigate the risks and opportunities of the next generation of AI.

The signing took place as international politicians and technologists convened to discuss AI safety at Bletchley Park, the English country estate that famously served as home base for British World War II codebreakers like computer pioneer Alan Turing.

“Many risks arising from AI are inherently international in nature, and so are best addressed through international cooperation,” the declaration reads. “We resolve to work together in an inclusive manner to ensure human-centric, trustworthy, and responsible AI that is safe, and supports the good of all.”

The symbolic gesture comes as worldwide conversations about AI regulation over the past several months recently began to bear fruit in the form of more concrete actions. Earlier this week, President Biden signed a sweeping executive order attempting to better guide responsible AI development; last week, the UN introduced its long-awaited AI advisory body; and the European Union is reportedly in “touching distance” of passing its own laws.

In the Bletchley Declaration, governments including the US, China, and the EU agreed to identify shared risks and build a shared scientific understanding around them. Then, each country will create its own policies around managing those risks, “collaborating as appropriate while recognizing our approaches may differ based on national circumstances and applicable legal frameworks.”

It’s a bit light on tangible specifics for now, but the point is to lay the beginnings of a groundwork for international cooperation on the issue, UK technology secretary Michelle Donelan told reporters, according to The Guardian. It could also serve to carve out a prominent seat at the table for the UK itself in this new global regime.

“For the first time we now have countries agreeing that we need to look not just independently but collectively at the risks around frontier AI,” Donelan said.

Forrester VP and principal analyst Martha Bennett said in an emailed research note that many of the delegates likely wouldn’t have agreed to sign onto a document that stipulated more detailed commitments.

“This declaration isn’t going to have any real impact on how AI is regulated,” Bennett said. “We’ll have to wait and see whether good intentions are followed by meaningful action.”

Among those who flocked to the English countryside for the summit were world leaders like European Commission President Ursula von der Leyen and United Nations Secretary-General António Guterres, as well as tech industry A-listers such as Elon Musk and OpenAI CEO Sam Altman, according to Reuters and the Associated Press. The guest list included dozens of companies and organizations from the AI world, like the Ada Lovelace Institute, IBM, Hugging Face, and Databricks. Vice President Kamala Harris gave a speech from London.

While there’s not always agreement on what exactly AI regulation should look like, there’s clearly no shortage of important people the world over who wouldn’t mind being seen discussing it.