Canada’s AI reputation at risk if it doesn’t speed up adoption: tech community

TORONTO — Members of Canada’s tech community say they’re worried the country is adopting artificial intelligence too slowly, threatening its reputation as a leader in the sector, if the trend persists.

“The slow pace of adoption of Canadian-made AI technologies is one of the biggest challenges that puts Canada in danger of ceding its hard-fought leadership position in AI,” said Krista Jones, chief delivery officer at MaRS Discovery District, a tech and innovation hub.

Jones was speaking at the MaRS Impact AI conference in Toronto on Thursday, where entrepreneurs and investors discussed how to ensure Canada remains a hotbed for AI innovation and talent. Discussions also tackled how to prevent squandering the advances and interest in AI that the technology’s pioneers, Geoffrey Hinton and Yoshua Bengio, have sparked in Toronto and Montreal.

Panelists at the conference pointed to recent studies showing Canada’s AI adoption rate is lagging its neighbour’s as proof that action needs to be taken soon.

A study from consulting firm KPMG showed 35 per cent of Canadian companies it surveyed had adopted AI by last February. Meanwhile, 72 per cent of U.S. businesses were using the technology.

“That’s a startling figure to me, especially when you consider that Canadian businesses report a higher capacity of their talent to effectively adopt AI technologies versus their competitors in the U.S.,” Jones said.

Canadian companies have shied away from AI adoption for a myriad of reasons, including a lack of funding, a fear of the technology’s risks and simply, not knowing where to start.

Catherine Fortin LeFaivre has seen too many Canadian companies get “bogged down” in trying to implement AI because they want to develop large-scale AI strategies that take too much planning and involveintricate processes like proofs of concept.

Rather than never implementing the technology because these processes are so onerous, the Canadian Chamber of Commerce’s vice-president of strategic policy and global partnerships said these companies should start with small, manageable uses of AI.

“It’s OK for it to be incremental and small and not complicated,” she said.

Tomi Poutanen, the head of Toronto-based AI health company Signal 1, added adoption can also be hard because of regulatory requirements and funding.

While he’s got his patient insights technology implemented in several health care organizations, he said many hospitals shuffle all their money toward staffing and patient outcomes. 

That leaves little room for spending on technology, let alone AI, and any company wanting to undertake such endeavours winds up taking on decades-long loans, he said.

If you approach a chief financial officer at an Ontario hospital and tell them you have a tool that can move people out of acute care and back into the community, eliminating backlogs, Poutanen said you will be told “there’s no money in saving lives.” 

“All hospital beds will be filled all the time, no matter what you do for discharge. There’s no incentives for a hospital to pay money to run a more efficient, better quality health care system,” Poutanen continued.

“That’s the brutal reality of trying to sell innovation to Canadian hospitals and that’s also why most Canadian health care technology companies first establish a market south of the area before they do it.”

But many speakers at the conference agreed that the country shouldn’t see AI adoption as futile.

They said there are several things Canada can do to ensure it doesn’t fall behind.

For example, Elissa Strome of the Canadian Institute for Advanced Research said governments have to step up as procurers of AI.

“Having the government as a first customer, I think, is a really important signal … that really helps them to move forward,” she said.

She added governments should offer incentives to domestic companies who buy Canadian AI because “we’ve heard from Canadian companies that a lot of their business is outside of Canada.”

Meanwhile, Fortin LeFaivre said she’d like to see the country use some metrics to track AI adoption and set goals, which Statistics Canada could track and politicians could be offered to prioritize in ministerial mandate letters.

“We need to incentivize, we need to measure, we need to be accountable,” she said.

Should Canada not take action soon, Jones said the country need only to look back at what happened with insulin to see what could transpire with AI.

“Insulin was discovered here,” she said, referencing how the diabetes therapy was invented on the site where the MaRS building now stands.

“And, of course, the patent was sold for $1 for the good of humankind and off it went to the U.S., where it’s saving lives but for a pretty penny.”

This report by The Canadian Press was first published Feb. 22, 2024.

Tara Deschamps, The Canadian Press