The artificial intelligence boom presents Canada with unique opportunities and risks as we seek to benefit from a technology that could reshape how we live.

In this special series, Canada’s AI Advantage, The Logic examines how Canadian companies, investors, institutions and workers can gain from the country’s early lead in AI, even as Canada’s pioneers in the field become the world’s most powerful voices of caution.

When the release of ChatGPT kicked off an artificial intelligence craze a year ago, Shelby Austin, the CEO of Toronto-based Arteria AI, received a flood of calls from international executives seeking advice on how they could get ahead of the curve on the nascent technology. 

Showing notably less interest were Canadian CEOs of organizations that, despite operating in some of the same markets as Arteria partners like Citigroup and Goldman Sachs, appeared to have less curiosity about what the rapid advancement of AI technologies like large language models meant for their businesses. 

Austin, who founded Arteria AI in 2020, said the calls with executives are just one anecdote about Canadian companies’ posture toward AI technology. But her experience also echoes more macro-level findings that suggest Canada is lagging its peers when it comes to AI adoption. Some industry insiders see low AI take-up as part of a deeper risk aversion within the Canadian private sector, one that threatens to put the country at a disadvantage during a moment of rapid advancement in artificial intelligence. 

The Dais, a Toronto Metropolitan University think tank, released a study in September that found just 4.7 per cent of Canadian companies with 10 or more employees had integrated AI into their business in any capacity as of 2021. The problem underscores “a trend that has been plaguing Canada for a long time,” study author Angus Lockhart told The Logic in an interview. 

Canada’s slow AI adoption comes at a time when the technology looks poised to disrupt entire industries, potentially carrying wider implications for the country’s competitive position relative to other nations. 

Austin, like others interviewed for this story, said Canadian firms in general tend to prefer waiting until technologies are proven before investing in them. 

“There’s just not the same demand and hunger to be leading edge,” she said. 

“At a corporate level, there’s a much more conservative ethos and reluctance to go further or be an early adopter in the space, which we think can be challenging.” 

Arteria offers an AI-driven platform that financial companies use for contracting purposes. By sorting through reams of otherwise hard-to-organize data, the technology provides companies with sharper information that can assist with a range of tasks from client onboarding to loan processing, Austin said. 

Jas Jaaj, managing partner for AI and data at consulting firm Deloitte, said Canada’s record on AI adoption is mixed, depending on the sector. In research, financial services and telecoms, Canada is leading the way in many respects, he said, while it trails in sectors like construction and real estate. 

In academia, the number of AI-related patents generated in Canada leapt 57 per cent in fiscal 2023, the second-fastest of any country. Many companies are also experimenting with AI, Jaaj said. (Unlike Dias’s findings, Deloitte estimates that 40 per cent of firms are exploring AI at least in minor ways.) 

Actually implementing those technologies at commercial scales will require considerably more thought and investment. Deloitte estimates in its report that just 26 per cent of companies have adopted AI technology, compared with 34 per cent globally, and that “corporate Canada is already falling behind its global peers” in using the technology. 

“The leap that we have to take now is to take these prototypes and then deploy them into production and scale them across the enterprise,” Jaaj said in an interview.

Sheldon Fernandez, CEO of Waterloo, Ont.-based DarwinAI, said there is a marked difference in the approach to technological adoption between Canadian and U.S. firms. The company sells a microwave-sized device that uses AI to scan for potential faults on printed circuit boards, automating a visual factory-floor inspection process that was previously done by humans.

Roughly half of DarwinAI’s dozen clients are Canadian, but Fernandez said U.S. and other foreign firms tend to have a higher appetite for risk and investing in new technologies like AI. 

“In general, Canadian companies are a lot more risk averse than their American counterparts,” he said. 

When DarwinAI launched a separate product during the pandemic—a platform that monitored AI programs and offered insights into how they ultimately came to decisions—just one of the company’s 10 clients was Canadian. 

“I used to lament that here we have cutting-edge IP invented by a Canadian company, connected to the University of Waterloo, doing wonderful things,” he said. “And the companies that were benefiting from this were all American corporations.” 

Even so, Fernandez and others said there are signs Canadian companies might be improving their uptake. Canadian firms including Shopify, Coveo Solutions and many others have embraced generative AI technology as a way to drive down costs, automating customer services and countless other aspects of their businesses.

Deloitte’s Jaaj agrees, pointing out that the technology has potential to help companies reap savings by improving operations or developing new products for a fraction of the previous cost. 

“This trend with generative AI is going to help us because you can now subscribe to a service from these large language model providers—be it through Microsoft, be it through Google, etc.—to be able to use their APIs, their interfaces, and be able to build solutions by tapping into the models that they have trained.”  

Sam Ramadori, CEO of Montreal-based Brainbox AI, said a major hindrance among would-be AI adopters is a lack of consistent data to feed into software systems. Data is often siloed between various divisions within a company, he said, or stored in inconsistent formats or with different naming conventions. Adopting AI often requires the terribly tedious task of streamlining those disparate data pools. 

“You have that layer that you have to go and solve, and you have to solve that before getting to the benefits of AI,” he said. “There is a bit of a leap there.” 

Brainbox sells an AI-driven software system that monitors and adjusts heating and cooling functions in commercial buildings. The system ingests information on a room-to-room basis from a building’s furnaces, thermostats, moisture sensors and other applications, then uses AI to make constant adjustments aimed at lowering heating costs and increasing efficiency. Brainbox has installed the technology in over 1,000 commercial buildings, and has serviced major big-box store chains including mattress maker Sleep Country. 

While many of the company’s clients are Canadian, Ramadori said Canadian companies tend to be more hesitant to adopt the technology. Rather than an AI-specific problem, he says the trend appears to be due to a broader aversion to risk that has permeated other industries across the country.

“It’s an unfortunate, recognized state of affairs: that in Canada, we’re just generally more conservative and slower to adopt new technology.” 

Tardier AI uptake is likely to have broader implications for Canada’s private sector as the technology is increasingly integrated across the economy. 

Austin of Arteria AI said artificial intelligence is set to become a dominant technology in coming years—influencing anything from revenue to productivity and costs—regardless of whether Canadian companies choose to join the fray. 

“It’s undeniable,” she said. “And as a result, if you’re not using [AI], and everyone else has it as a tool in their toolbox, it begs the question: where are you going to be in one year, two years, five years?”