How Commercial Real Estate Firms Deploy AI in Practice

Real-world AI applications emerge in commercial property

While most commercial real estate firms claim to be exploring artificial intelligence, a smaller group is now demonstrating specific use cases that deliver measurable business value. From identifying off-market opportunities to calibrating investment hold periods, these early adopters are moving past pilot programs into operational deployment.

The shift comes as industry surveys reveal uneven progress. JLL reports that 88% of investors and landlords have started AI pilots, yet Deloitte's 2026 Commercial Real Estate Outlook found 27% of respondents cite implementation challenges including technical barriers, expertise gaps, and organizational resistance.

Why it matters

These concrete examples provide a roadmap for firms still determining where AI can generate returns versus where it adds complexity without commensurate benefit. As adoption spreads, early movers may lose competitive advantage—making speed of implementation increasingly critical for firms seeking differentiation through technology.

Investment strategy and market timing

Miami Beach-based BGO, an institutional investor owned by Sun Life, has developed a proprietary algorithm that evaluates real estate performance across approximately 2,500 U.S. markets. Co-CEO John Carrafiell explained the system helped recalibrate the firm's apartment strategy in Miami, with analytics identifying second-tier submarkets positioned for growth.

The algorithm also prompted BGO to exit a $1 billion industrial portfolio in Southern California's Inland Empire despite the region's 4.5 million population and two decades of strong fundamentals. "Our analytics signaled that returns were about to decelerate," Carrafiell noted. "Getting out ahead of that—before others saw it—was key."

Toronto-based Hazelview Investments uses AI models to process real-time market signals combined with proprietary data on rents, vacancy rates, and tenant satisfaction. "When we evaluate an acquisition, we use our AI models to score each opportunity against assets we've actually owned and operated," said Strachan Jarvis, co-head of private real estate investments. "That carries straight through into underwriting, where our assumptions are benchmarked against real outcomes, not market averages."

Underwriting and tenant analysis

Bryn Feller, a senior vice president at brokerage Northmarq's Chicago office, highlighted how predictive AI transforms lease renewal forecasting. Rather than relying on investor perception or historical averages, platforms can now calculate renewal probabilities for individual tenants using correlated data points. Feller cited CFS, a predictive analytics platform, as one tool converting available data into actionable leasing intelligence.

"AI went from a buzzword for a lot of organizations to being front and center," Feller said. "They're moving from checkers to chess and moving into this 4-D multi-timeframe kind of calculus that humans just can't do to this degree of analysis."

Development and regulatory tracking

For developers, AI applications extend to regulatory monitoring. Cityscrape, a platform that tracks planning code changes, helps builders anticipate zoning modifications before implementation. Noah Shechtman, a Cityscrape co-founder and development director at Toronto-area builder Brightstone, explained the system aggregates municipal data in real time. "If I know there's a change when it's first proposed, I can pivot elements of my project six months or a year before it's implemented, whereas having to react to change always causes project delays," he said.

Ontario-based Dancor Construction recently partnered with developer KED Ltd. specifically for KED's AI capabilities in project development. KED founder Dario Zulich uses AI to rapidly generate custom proposals for prospective clients, inputting requirements like ceiling heights and column spacing to produce quick estimates.

Skepticism remains

Not all industry executives view AI as universally beneficial. Some argue that as adoption becomes widespread, competitive advantages erode. Others maintain that market experience can match or exceed AI-generated insights for certain applications.

These details were first reported by CoStar News.