LONDON — As ministers from major economies and the heads of top AI firms assemble this week at the U.K.’s landmark AI Safety Summit, Canada is looking to play a leading role at an event designed to shape the international conversation on the technology’s risks. 

The gathering is a showcase for Prime Minister Rishi Sunak’s pitch to make the U.K. a world leader in the development and governance of AI. The two-day event, which begins Wednesday at Bletchley Park, consists of roundtables with officials, executives and researchers, as well as closed-door meetings between government representatives. The program is focused mainly on so-called frontier AI, a category of models that can be put to many uses, including in generative tools like OpenAI’s viral ChatGPT. 

Innovation Minister François-Philippe Champagne will represent Canada, and the government’s aim is to “leave [Canada’s] fingerprints on this Bletchley Park summit,” said a source with knowledge of Ottawa’s approach, who requested anonymity to discuss the plans. Champagne is scheduled to chair a roundtable on “risks to global safety from frontier AI misuse” on the first day of the event, and join another on what national policymakers should do about its risks and opportunities.

Ottawa doesn’t expect the summit to produce definitive answers on those issues, but sees its participation there as an extension of the role it’s been playing abroad on responsible use of the technology, the source said. For example, Canada is part of the G7 Hiroshima AI Process, launched in May. On Tuesday, the group called for firms developing and deploying advanced systems to follow risk-management and transparency principles. Canada and France began working together on AI in June 2018, leading to the establishment of the Global Partnership on Artificial Intelligence (GPAI) two years later. The organization now has 29 members.

The government will also tout actions it has taken at home. Champagne has repeatedly said Canada will be the first jurisdiction with a framework for responsible use of the technology, once his proposed Artificial Intelligence and Data Act passes. Ottawa’s generative AI code of conduct, unveiled in September, is novel in setting requirements for not only developers but also users of such tools, the source said.

Governments around the world are taking different tacks to AI policy. Despite his warnings about the technology’s dangers, Sunak’s government has promised a “pro-innovation approach,” in which the government delegates rulemaking to sector regulators. The EU’s AI Act uses a sliding scale of risk, outright banning some applications while setting registration and other requirements for high-jeopardy systems. The U.S. has rolled out a set of voluntary guardrails for generative AI, and on Monday President Joe Biden ordered developers to share safety and other information with authorities. 

At the summit and in other policy discussions, Ottawa aims to act as “a bridge between all these different approaches,” said the source. The U.K. has similar ambitions to be an intercontinental AI policy broker; in recent months, British diplomats have made that pitch in major capitals around the world. 

In Canada, the British High Commission has held meetings with sector players in AI hubs across the country including Montreal, Toronto and Vancouver. “We have a very similar perspective on the world, into how these technologies ought to be governed and accessible [while considering] the risks that might be introducing,” said David Barnes, the high commission’s head of science and innovation. 

British firms are also drawn to Canada because of its supply of AI talent, he said, citing the effect of a $287-million national strategy to attract researchers who in turn would train students. On Tuesday, the U.K. allocated £118 million ($198 million) for labs and scholarships.

Sunak has also announced a new AI Safety Institute to study the technology’s risks and impacts. The U.K. has sought Canada’s support for the initiative, and Ottawa is considering how it might contribute, perhaps by seconding researchers, the source said. 

Canada will also participate in a scientific advisory group that the U.K. is looking to launch out of the summit. While Champagne has touted the economic opportunities of AI, Ottawa sees a need to redress a funding imbalance that favours advancing the technology over developing safety measures, the source said. “The era of, ‘Just trust us’—that’s over.” 

Some scholars, policy professionals and tech executives have questioned the summit’s focus, guestlist and even the need for it to happen at all. On the first day of the AI Fringe—a series of unofficial events this week in London and elsewhere across the U.K. —several panelists cautioned that dire warnings of AI rendering humans extinct were pulling focus from current problems. “What I think is catastrophic is we just see an immense skewing of [the] portfolio of mitigation techniques [for AI] towards risks which are more long-term horizon,” said Sara Hooker, director of Cohere for AI, the Toronto-based startup’s non-profit research lab.

About 100 guests are scheduled to attend the summit. Other notable Canadian participants include Geoffrey Hinton and Yoshua Bengio, the deep learning pioneers who have called for more safety measures and funding; and Cohere CEO Aidan Gomez, who’s said existential fears are a distraction from more pressing challenges like bias and misinformation. OpenAI’s Sam Altman, Google DeepMind’s Demis Hassabis and Elon Musk, in his capacity as head of X.AI, will also attend.