Sweden is home to some of the world’s most comprehensive regulations on artificial intelligence (AI). Canada, meanwhile, has sought to strike a balance between safety and promoting innovation. Yet both countries are now pushing widespread adoption so they can join “the A team.”
“The A team will build and shape the tools. The B team will use whatever tools others provide… Sweden cannot afford to be a B nation.”
On Thursday, Sweden’s deputy prime minister and head of industry — accompanied Sweden’s King Carl XVI Gustaf and his wife Queen Silvia, as well as many academics and industry leaders, on a visit to Mila, the federally funded AI research hub in Montréal. Attendees included Mila scientific director Hugo Larochelle, Québec’s economy minister Christine Fréchette, Cohere chief AI officer Joelle Pineau, and AI safety organization LawZero CEO Sam Ramadori, who came together to discuss the importance of AI research collaboration and developing sovereign AI technology.
Busch told the crowd that “not deploying AI is a very real risk.”
“It means slow growth; it will mean weaker public services and dependency on others,” Busch said.
She added that while Sweden has been cautious when it comes to AI, she also recognizes the need to move ahead.
“The A team will build and shape the tools. The B team will use whatever tools others provide, with weaker competitiveness and weaker sovereignty,” Busch said. “Sweden cannot afford to be a B nation.”
Sweden is a European Union (EU) country, meaning that it’s subject to the EU AI Act, which came into effect last year. These regulations prohibit manipulative or subliminal techniques, social scoring, and real-time facial recognition in AI tools. Regarded as one of the most comprehensive regulatory frameworks on AI in the world, it has drawn criticism from the tech industry for its perceived constraints on innovation, leading the European Commission to roll back some of its measures.
Busch’s speech mirrors rhetoric from Canadian AI minister Evan Solomon. He has said Canada’s AI industry is at a “crisis moment” and that he will pursue “light, tight, right” regulations that don’t stifle innovation. At The Logic Summit, Solomon said AI adoption is just as important as AI sovereignty, or control over the technology’s supply chain and data.
Like Canada, Sweden has sought to adopt a national AI strategy and support homegrown talent. The country of 10 million people has produced some notable tech players. Buy now, pay later FinTech startup Klarna went public this year in the US, and AI coding startup Lovable said it has hit $200 million USD in annual revenue.
RELATED: Evan Solomon teases new AI laws as experts warn Canada is behind international peers
Under former Prime Minister Justin Trudeau, Canada tried to strike a balance between AI innovation and regulation by signing international treaties on AI safety. However, it failed to pass laws on regulating AI.
The Swedish visit to Mila was meant to introduce the Scandinavian country to collaboration opportunities with Québec’s AI research hub. Since its founding in 1993 by Turing Award winner Yoshua Bengio, Mila has grown into a world-renowned AI research hub. Mila and its sister organizations, Toronto’s Vector Institute and Edmonton’s Alberta Machine Intelligence Institute, are funded by the federal government through the Pan-Canadian AI Strategy.
The Mila event also featured a panel of industry and academic AI experts. In response to a question from BetaKit about what Canada’s AI regulation should ideally look like, Elena Fersman, vice-president and head of AI innovation and incubation at Swedish telecom company Ericsson, said, “Don’t regulate technology; regulate the use of technology.” Pineau, with Cohere, agreed that Canada should adopt this approach.
Solomon has said his office will table an AI strategy in the new year. It’s currently reviewing submissions from its AI task force made up of industry players and researchers, as well as over 11,000 public submissions — the largest number of public submissions to a task force in Canadian history.
In an interview at Mila, Pineau, who was on the task force, told BetaKit that her submission focused on the elements needed to achieve trust in AI.
“I really pulled out the fact that to achieve trust beyond safety, you need to think about literacy. You need to think about transparency. I think we need to have a holistic approach to build trust,” Pineau said.
Cohere is a Toronto-based enterprise AI scaleup and has been crowned as Canada’s “AI champion” according to some government ministers. Pineau added that the Sweden visit was “exciting” as Cohere looks to explore partnership opportunities in Europe, and doesn’t yet have a presence in Sweden.
Feature image courtesy Mila. Photo by Maryse Boyce.