Commercial real estate could be both a builder and a beneficiary of the artificial intelligence (AI) ecosystem envisioned in the Canadian government’s newly released national AI strategy. A combination of spending promises, policy prompts and expectations for private capital investment underpin ambitions to boost AI competence and uptake in the general population, nurture, entice and retain AI innovators and entrepreneurs, build out and reinforce Canadian-controlled AI infrastructure, enhance productivity and spur economic growth.
Five economic sectors — health and life sciences; energy and natural resources; transportation; agriculture; and manufacturing and robotics — are identified as having a strong foundation of data and/or a convergence of scientific, economic and industrial depth where strategic investment in AI is projected to “deliver both commercial success and sovereign resilience” that could anchor Canada as a global leader. The federal government also has an objective to streamline administrative tasks and improve service delivery within its own departments through the implementation of AI technology.
“Canada’s new AI for All Strategy is about putting artificial intelligence to work for Canadians. It will give people the confidence to use AI safely, help businesses adopt it, and ensure more of the value is created here at home,” maintains Evan Solomon, Canada’s Minister of Artificial Intelligence and Digital Innovation.
Sovereign infrastructure aspirations
Agendas within the commercial real estate sector are somewhat more scoped than nation building, but robust investor appetite for the data centre asset subclass could align with the government’s target to secure 850 megawatts (MW) of AI compute capacity in hyperscale facilities by 2030. Potential further demand of up to 2.3 gigawatts (2,300 MW) is foreseen, along with requirements for high-capacity fibre-optic cables.
That would forge what’s described as a “sovereign alternative” as one component of the larger base of AI computing capacity. An injection of “crowd-in private capital” will be pursued to help finance a proposed public supercomputer to be made available for Canadian research, development and commercialization efforts. It’s expected to become operational by 2031.
The strategy reiterates that foreign hyperscale developers/operators will likewise be welcome to invest in the projected 5.5 gigawatts of AI computing capacity commercial users will need to employ by the end of this decade if their projects can “deliver clear benefits for Canadians”. On that front, the strategy pitches the competitive advantages of Canada’s largely low-emission electricity supply and climatic conditions that reduce the operational costs of cooling (energy and water) relative to more southerly locations. It also references the target to double the capacity of Canada’s electricity grid by 2050, which is set out in the national electricity strategy that was released earlier this spring.
“As demand for AI compute grows, Canada’s approach will be to link new data centre development with clean energy expansion, robust environmental standards and tangible benefits for local communities, ensuring that Canada remains at the forefront of sustainable high-performance computing infrastructure,” the AI strategy states.
The productivity super-deduction, introduced in the 2025 federal budget, is highlighted as an incentive for investors. It allows them to write off 100 per cent of the eligible capital costs to acquire, build or expand a facility used for manufacturing or processing in the tax year it is acquired or becomes operational, provided that occurs between Nov. 4, 2025 and Dec. 31, 2029. After that, a 75 per cent deduction will be available for applicable claims in the 2030 and 2031 tax years, and a 55 per cent deduction for 2032 and 2033.
Business and training stimuli
The strategy cites evidence of lower AI adoption levels in Canada than in many other member nations of the G20 and Organisation for Economic Co-operation and Development (OECD). Some of the promised federal investment in training and uptake is to be channelled directly to businesses, but it’s also intended to flow indirectly to employers as they hire from a pool of job candidates with augmented skills. Commercial real estate firms could presumably be on the receiving end of both benefits.
Statistics Canada data shows that about 12 per cent of total Canadian businesses used AI in the production of goods or services as of mid-2025, but only about 8 per cent of small and medium-sized enterprises (SMEs) were doing so. That compares with adoption rates of 26 per cent in Germany, and ranging from 29 to 42 per cent in the Nordic nations.
Findings from a KPMG-University of Melbourne survey of AI training and literacy rank Canada 44th out of 47 countries. Fewer than 24 per cent of Canadian respondents reported they had undertaken any formal AI training, while 36 per cent viewed AI as harmful to society.
“Initial adoption has happened; deeper, confident integration has not, and low literacy and low trust are the binding constraints,” the strategy concludes.
In response, the Canadian government pledges support to get a targeted 60 per cent of businesses on board with AI within the next eight years. In tandem, it promises to activate up to 90,000 work placements and jobs for young career candidates, who would be embedded in SMEs and non-profit organizations. That would be a significant portion of 250,000 new AI-related jobs projected to be added to the economy by 2031 — contributing to a projected 3 per cent gain in gross domestic product (GDP) that would reflect a $200 billion surge in labour productivity.
Funding for businesses is to be channelled through the Business Development Bank of Canada’s (BDC) $500-million LIFT (lead with innovation and focus on technology) program, a $500-million injection into the Regional Artificial Intelligence Initiative (RAII) and other “targeted support” through programs geared to small businesses.
Yet-to-be-developed free resources to help businesses assess their preparedness for AI adoption and potential paths for doing so are also promised. Similarly, AI literacy content is promised for students, educators and the broader public.
“Canada will create a National AI Literacy Initiative that will offer entry-level AI training accessible to all Canadians,” the strategy advises. “Through the National AI Literacy initiative, Canada will empower public libraries and community organizations — long trusted as hubs for learning — as natural partners to bring AI literacy initiatives into every community, especially those in rural, remote and northern regions.”
Applications for investment asset management
Observers within the commercial real estate sector suggest it reflects some of the general population’s hesitancy, but they also detect openness that’s in line with the Canadian government’s AI aspirations. Practitioners of the valuation and asset management disciplines voice enthusiasm for AI’s ability to quickly scan and identify patterns, correlations and anomalies in the vast reams of data that the industry and the general economy engender. However, that’s balanced with caution about the depth and congruence of the data from which those insights are gleaned.
Participants in a recent webinar discussion about AI’s contribution to investment decision-making described it as a tool to support discerning human judgement. Carl Gomez, chief economist and executive vice president of research with Centurion Asset Management, slotted it at the generative, information distilling stage of its evolution, with more iterations of advancement required before it is ready to take — and/or human analysts are prepared to cede — authoritative action.
“A lot of real estate firms are using AI to integrate very rote sorts of practices. Deal-screening, data synthesis, reading over reports, checking for small errors, that’s, practically, where we’re utilizing the tool today,” Gomez reported. “It can distill information very quickly into something that is usable. The problem right now is, depending on what data sources you’re feeding in, the outcomes that come out can be contradictory. Or maybe they don’t feel right to the experts in the field who have observed things, and that challenges the old order of how to make decisions.”
Christina Gratrix, senior director, product management, with Altus Group, concurred that much depends on consistent terminology and categorizations across datasets and on the informed direction of the humans who are prompting AI to mine data. With those in place, she maintains historical data can yield valuable insights, help identify catalysts of market swings and parse out factors that may have been overlooked previously.
“There’s a huge difference today in the data points that clients are asking for and collecting and using to build context for their assumptions from even 10 years ago. A lot of firms have data from the past five, 10, 20 years, but that historic record doesn’t have all the data points they need today,” Gratrix observed. “There are variables that we didn’t consider five, 10 years ago that may be embedded in the historic data, but they may not know how to break it out yet.”
From her perspective as a service provider to the industry, she stressed AI adopters must have the acumen defend what the technology delivers because clients with their own investment committees to convince will be asking for that assurance. That’s also premised on data management standards and transparent methodologies.
“Making sure that we can explain every aspect of why the AI tool that we use came up with the answer it did, that’s key,” Gratrix asserted.
In future, Gomez speculates AI could replace some jobs that have traditionally been entre-level roles for analysts. For now, he perceives humans are still largely performing those functions, but required skill sets are evolving, demanding AI competence along with real estate knowledge.
“If you look at our workflow over the last few years, we’re doing a lot more on an individual basis. We’re being a lot more productive with fewer people,” agreed Ray Wong, vice president, valuation and advisory, with Newmark Canada. “From that perspective, it’s going to make the grunt work a lot easier. Instead of spending hours on a spreadsheet or a model, having AI generate that for us and allowing us to focus on the value-add.”
That complements the national AI strategy’s projections for unleashing productivity, while Gomez hypothesizes that improvements in data consistency and synthesis alter competitive jockeying in an industry that has traditionally measured performance with a lot of different metrics recorded in different places. Although still lacking the consistent clarity that the repository of data on record at exchanges provides for public equities, it is becoming easier to capture the panorama of details.
“Historically, part of real estate’s appeal is that it’s not very transparent and sometimes you can earn excess returns because you have a better handle on the market than others — because you have boots on the ground, different resources and things like that,” Gomez mused. “That is the challenge, but it’s also the opportunity now as the data becomes more refined and AI comes into the picture to allow us to get better clarity on the market.”
