As many companies race to adopt artificial intelligence (AI) tools globally, a new report from Toronto venture capital (VC) firm Georgian Partners suggests most Canadian companies are still learning to walk as they grapple with barriers and mixed efficiency gains.
About 48 percent of Canadian technical executives in the study said an absence of technical talent was a challenge impeding AI adoption.
Canadian firms with annual revenue greater than $5 million are trailing compared to their global peers in implementing AI for go-to-market and product strategies, according to the report.
Georgian classifies companies based on a scale of “AI maturity,” with the categories of crawl, walk, jog, and run. Only seven percent of Canadian technical respondents identified their companies as “runners” with a sophisticated portfolio of AI projects and a significant budget to match, compared to 17 percent globally.
The report, done in collaboration with market research company NewtonX, surveyed 634 business executives globally, 201 of whom were Canadian, via a 15-minute online survey.
The findings come as Canadian companies face increasing pressure to adopt AI, both from leaders at big tech firms investing heavily in the technology, and government leaders such as Canadian AI minister Evan Solomon. At a Toronto Tech Week panel hosted by the Toronto-based Vector Institute and moderated by BetaKit reporter Josh Scott, Georgian machine learning engineer Royal Sequeira previewed the results of this study alongside Craig Stewart, executive director of applied AI programs at Vector Institute.
“At Vector, core to our principles are AI safety and trust,” Stewart said. “At the same time…our adoption rates in this country are too slow, so I would encourage us to get to market much, much quicker and sell these products and services.”
Stewart said that most companies focus on “back-office” functions as a starting point for introducing AI, but argued they should accelerate AI adoption in consumer-facing functions. According to the Georgian report, Canadian technical teams are 12 percentage points below the global average in implementing AI for customer service functions.
“We want to move that bar and get to the podium,” Stewart said.
The report noted that Canadian companies face unique barriers to AI adoption. About 48 percent of Canadian technical executives in the study said an absence of technical talent was a challenge impeding AI adoption. This is four points higher than the global average, according to Georgian.
The next most important barrier was the high cost of model training and deployment, Georgian added. Canadian technical teams were also less likely than their global counterparts to build their own models. Instead, they’re buying solutions from third parties—the top way these respondents said they were managing costs associated with AI.
To offset the costs of investments in AI, the report said, 34 percent of Canadian technical executives surveyed said they are reducing headcount. This follows recent “AI-first” company-wide mandates from prominent Canadian tech firms such as Shopify. Kitchener-Waterloo-based firm OpenText laid off 1,600 employees in May as part of an AI-first policy. Vancouver-based software-as-a-service (SaaS) startup Klue cut 40 percent of its workforce last month, citing internal and external pressures from AI and a bid to increase efficiency.
At Startupfest in Montréal, Georgian lead investor Evan Kerr said that return on investment (ROI) is a key concern for AI adopters, according to the report’s global and Canadian-specific data. Kerr encouraged founders selling AI products to have a “firm perspective” on defining ROI for their clients.
“We’re seeing a lot of inconsistency in the way companies talk about [ROI],” Kerr said. “Some of the responses say [they] haven’t translated specific revenue or cost targets to AI that [they’re] putting into production.”
The best of Toronto Tech Week
BetaKit is Toronto Tech Week’s official media partner. Read all of our coverage here.
Part of the challenge is determining a reliable way to measure the direct impact of new AI tools. Fifty percent of technical respondents to the Georgian survey found it difficult to connect AI projects to revenue, Sequeira said during Toronto Tech Week.
“This could be because they already are adopting AI within their teams without really thinking about how it connects to business,” Sequeira added.
Some companies selling AI products for enterprises are hearing mixed results from client firms piloting AI projects. Cohere co-founder Ivan Zhang told BetaKit at Web Summit Vancouver that many of his client firms have not seen the return on investment from the spending they have put into AI proof-of-concept projects. Zhang said cost is a barrier, as AI models are expensive to run.
The Georgian report found uneven efficiency gains from AI across the surveyed Canadian executives. By some self-reported metrics, such as product release quality and technical debt reduction, roughly half of Canadian respondents reported benefits from implementing AI. But when it came to business metrics, Canadian respondents reported lower impacts of AI on net retention and cost savings compared to their global peers.
Other independent studies have seen similarly mixed results. A recent National Bureau of Economic Research working paper surveyed 7,000 workplaces to determine if AI chatbots impacted their bottom line and found “no significant impact on earnings or recorded hours in any occupation.” Another study from Boston Consulting Group found only a quarter of the 1,800 executives surveyed have seen significant value from AI so far.
Feature image courtesy Jennifer Jenkins for the Vector Institute.