TORONTO — Canada’s economy would grow by $26.5 billion via increased productivity if businesses in key sectors fully adopt generative AI to complement their workers, suggests a new analysis from Scale AI.

The math: Deloitte, acting for the Montreal-based cluster backed by the federal government, forecast the gains in output for 13 major industries of deploying every available generative AI tool, then totted up the financial impact.

For example, the consulting firm estimated that about 38 per cent of the 2.39 million workers in health care and social assistance could use AI, driving productivity gains of between 4.7 per cent and 8.4 per cent. That would increase the sector’s collective economic impact by $5.3 billion. The manufacturing, education, finance and insurance industries would also see significant boosts if employees had access to generative tools, Deloitte projected.

The backdrop: Canadian businesses have been notoriously slow to adopt new digital technologies, contributing to what the Bank of Canada has called a productivity “emergency” for the country. 

Firms will reach full usage of generative AI one way or another by 2044, the Deloitte analysis forecast. But Canada will leave money on the table if it waits for businesses to adopt the technology at their natural pace, said Scale AI CEO Julien Billot. “The sooner you can get that productivity, the better for the country.”

In a scenario where adoption goes slowly, Deloitte estimated the finance, manufacturing and retail sectors would collectively gain $18 billion in extra output over the next five years. Speed things up, and those productivity gains could total $33 billion. 

The middle layer: Scale AI is calling for Canadian businesses to buy “homegrown AI.” 

That doesn’t just mean products like large language models from Toronto’s Cohere, or search tools from Quebec City’s Coveo. To roll generative AI out across firms of all sizes and sectors, “you need a lot of integration or customized solutions,” said Billot. 

Service providers like consulting companies or digital agencies can then resell what they build to other clients, or turn it into products of their own. Billot compared AI today to the early days of the web, when pages were built one at a time at great cost. Now, most businesses launch their sites using the same tools from tech firms like WordPress or Shopify. In AI, “the goal for Canada [is] to have the right players in these platform games,” he said. 

Like the other clusters, Scale AI funds projects that bring together large firms, startups and academic labs to develop technologies that solve specific problems. Last month, it unveiled 23 projects worth $98.6 million that applied AI to sectors like mining, e-commerce and biotech. 

While project participants contribute their own resources, some Canadian tech executives have criticized the involvement of multinational firms, which they say use the program to harvest IP subsidized by public funds.

The budget question: The federal government has committed $125 million to the five clusters to support AI commercialization, money that’s due to expire in March 2027. Ottawa’s $750-million renewal of the main cluster program ends in March 2028.

Scale AI plans to seek further funding from the Liberal government, which under Prime Minister Mark Carney is looking to make cuts to federal spending. The cluster’s goal is eventually to “stop existing, because nobody needs us anymore,” Billot said. “The feeling we have today is we’re not there.” 

To bank the productivity benefits of generative AI, Canada still needs programs and policies that encourage businesses to adopt the technology, he added. “It’s time to invest now, because that’s where maximum gains can be achieved.”