Cheaper home markets lead Canada in AI-written property listings

Canadians searching for homes in the country's more affordable markets are the most likely to encounter real estate listings written by artificial intelligence and mortgage brokers helping clients in those cities may need to add a new layer to the due diligence conversation.

That finding emerges from a study by Originality.ai, a Canadian AI detection firm that scanned 72,047 listings, including 15,735 rentals and 56,312 sales, across 20 cities during the first two weeks of May 2026.

Its tools flagged 37% of all listings as likely AI-generated and 30% as likely human-written, with the remaining 33% inconclusive, falling below the firm's 100-word minimum threshold for reliable detection.

The geographic breakdown tells a pointed story. Calgary, one of Canada's more accessible major housing markets, topped the national list, with approximately 70% of its listings flagged as likely AI-generated, while Montreal registered the smallest share at just 7%.

In English Canada, Vancouver, where average home prices remain among the highest in the country, recorded only 31% of listings as likely AI.

The pattern suggests that in markets where margins are tighter and listing volumes are higher, realtors may be leaning more heavily on automation to manage workload, a dynamic with direct implications for buyers who cannot afford mistakes in a competitive search.

What AI listings look like and why it matters

For-sale listings showed a higher rate of AI adoption than rentals, with 41% of sales listings flagged as computer-written compared to 21% of rental ads.

Michael Fraiman, a researcher at Originality.ai who co-authored the study, noted that AI writing tends to follow a recognisable pattern.

Listings flagged as computer-generated were associated with words like "exceptional" and "convenience," each of which appeared in AI-flagged ads roughly two-thirds of the time, according to Originality.ai's analysis.

By contrast, casual shorthand such as "br" for bedroom, multiple exclamation marks, or terms like "must-see" and "gorgeous" strongly correlated with human authorship.

The study also found a measurable price gap between AI-assisted and human-written listings.

Homes described by likely AI-generated copy sold for an average of $1,100,906, compared with $1,180,910 for listings assessed as likely human-written — a difference of roughly $80,000.

The gap in rental listings was negligible: human-written ads averaged $3,037 per month versus $2,997 for likely AI-generated ones.

Brokers and the broader AI moment

The Originality.ai findings land at a moment of genuine reckoning for AI's role in Canadian real estate and financial services.

The Office of the Superintendent of Financial Institutions (OSFI) flagged AI-driven fraud and third-party technology risk as top supervisory priorities in its 2025–2026 risk outlook, with superintendent Peter Routledge describing AI as "a real amplifier" of cyber and third-party risk.

Vancouver-based mortgage broker Sharon Davis previously told Canadian Mortgage Professional that human brokers retain a structural advantage when it comes to evaluating complex scenarios that automated tools cannot replicate.

"AI doesn't think that way. AI doesn't say, 'Well, let's look at everything,'" Davis said.

"We have to be a hybrid. We need AI to keep us moving forward and help us be more efficient."

David Clarke of TMG The Mortgage Group says borrowers still prioritize trust and human relationships in complex mortgage decisions even as AI tools become more commonhttps://t.co/Ep2hYmvpdK

— Canadian Mortgage Professional Magazine (@CMPmagazine) May 20, 2026

Tim Rye, senior vice president of commercial solutions at Teranet, offered a similar view on the need for a balanced approach.

"AI is coming. Tools and technologies are coming faster, and we're seeing it more and more," Rye told CMP.

"I don't think we can avoid it. We have to embrace it, and we have to make sure we put protections, tools and partnerships in place to ensure a trusted ecosystem remains for the housing environment of Canada."

For now, that ecosystem includes real estate listings that may or may not have been written by a person, a reality that makes understanding AI's impact on the Canadian mortgage process increasingly relevant to every transaction.

Brokers advising buyers would do well to encourage clients to read listings critically, verify property details in person, and consult directly with realtors on any claims that seem unusually polished or unusually vague. 

Make sure to get all the latest news to your inbox on Canada’s mortgage and housing markets by signing up for our free daily newsletter here.