Canadian businesses are ramping up investment in artificial intelligence, but much of the spending is unlikely to translate into near-term economic growth because the equipment is being imported rather than produced domestically, according to a new analysis from RBC Economics.
In a report published Nov. 10, RBC senior economist Claire Fan says the economic benefits of AI investment in Canada are expected to emerge more gradually, with the greatest potential coming from productivity improvements as businesses adopt the technology and use it to increase output.
The analysis suggests that while rising investment in AI infrastructure and equipment is beginning to contribute to business spending, Canada's economy may not experience the same immediate boost seen in the United States. The difference largely reflects the domestic production of the technology and equipment being purchased.
AI investment gaining momentum
The report identifies three main categories of AI-related business investment: software, non-residential structures such as data centres and new power facilities, and machinery and equipment, including semiconductors and data servers.
In the United States, where AI investment is further advanced, computer equipment purchases have accounted for the largest share of business investment growth, followed by software spending. Investments in structures, including data centres, have made a smaller contribution, according to modelling by the U.S. Federal Reserve cited in the report.
Canada is beginning to see similar investment activity, including major data centre projects breaking ground across the country. RBC's analysis of national accounts data also indicates that businesses are increasing purchases of computer equipment as they expand their AI capabilities.
Purchases of computer and peripheral equipment increased 28 per cent in volume during the first half of 2026 compared with the same period a year earlier. The report says spending was expected to increase further in the third quarter.
However, the composition of that investment is limiting its immediate contribution to Canadian economic growth.
Imports limit immediate economic benefits
A significant proportion of the computer equipment being purchased by Canadian businesses comes from outside the country, meaning the spending does not generate the same domestic economic activity as investment in locally produced goods and services.
Canadian imports of computers and peripheral equipment were 75 per cent higher in September than a year earlier, driven by a sharp increase in high-value processing units used in data centres, the report says.
Because imports are deducted when calculating gross domestic product, the increase in equipment purchases has so far had a relatively neutral net effect on economic growth.
RBC also notes that imports of computer equipment have modestly exceeded related investment spending in Canada, suggesting some inventory accumulation that is not reflected in the investment figures.
Software investment has increased more modestly in Canada in recent years compared with the United States, with fewer indications of AI-related spending. The report attributes this partly to the continued dominance of U.S. companies in developing foundational AI models and generative AI software.
Investment in AI-related structures is more difficult to measure because Canadian economic data does not have dedicated categories for spending on data centres or power facilities.
Monthly production figures have pointed to some momentum in non-residential building and engineering construction since the spring. However, RBC expects the contribution from structures investment to remain smaller than the gains associated with machinery and equipment.
Productivity expected to deliver larger gains
While the immediate contribution of AI-related investment to GDP growth may be limited, RBC says the longer-term economic benefits could be more significant as businesses integrate the technology into their operations.
The report argues that the economic value of new technology has historically extended beyond the initial spending on construction and equipment. Greater benefits tend to emerge later, when businesses use new technologies to increase their capacity and improve productivity.
For Canada, that could mean enabling employees across different sectors to produce more output within the same number of working hours, or accomplish more with fewer hours worked.
RBC says Canada's productivity performance remains weak despite a modest improvement over the past year. GDP per hour worked increased just over one per cent, a slight improvement from previous years but still consistent with the relatively slow productivity trend that characterized the decades before the pandemic.
The report says broader AI adoption, implemented safely with appropriate oversight, could help improve that performance and support wage growth.
Productivity improvements could also become increasingly important as demographic changes put pressure on Canada's available workforce.
More than one in three Canadians is now older than 55, according to the analysis. Although immigration caps could ease from current levels, RBC says rising retirements will continue to reduce the working-age population and constrain the economy's potential growth.
The report argues that increasing output through productivity improvements will therefore be important in offsetting the impact of an aging population and supporting longer-term GDP growth.
Fan concludes that the economic payoff from AI investment in Canada is more likely to come through the technology's eventual adoption across businesses than from the initial purchases of equipment and infrastructure.