Kelly Jenkins
Speaker, Innovation Experts Real Estate Institute.
Strategy and Commercial Head at SOULATICS.

Walk into any property exhibition in Dubai today and you will hear the words

“AI-powered” attached to everything from valuation tools to tenant screening. The technology has become a fixture of real estate marketing. But for investors making decisions about where to deploy capital, the more useful question is not whether AI exists in real estate—it clearly does—but what it actually does, and where it falls short.

The answer matters because AI in real estate operates very differently from AI in other industries. Understanding that difference can sharpen how investors evaluate deals, assess developers, and manage their portfolios.

The Core Difference: AI Cannot See Properties

In industries like finance or healthcare, AI systems work with structured, standardised data. A blood test returns the same format whether taken in Dubai or Delhi. Stock prices flow through regulated exchanges with millisecond precision.

Real estate does not work this way. Every property is unique. Two apartments in the same building can have materially different values based on floor height, view orientation, or the quality of finishes that no database captures. A villa’s true condition—whether the AC system is aging, whether there’s moisture behind the walls—exists in physical space, not in data fields.

This is not a limitation that better technology will solve. It is the nature of the asset class. AI can process information about properties, but it cannot inspect them. The investor or their advisor still needs to stand in the unit, assess the neighbourhood at different times of day, and make judgments that no algorithm can replicate.

Where AI Adds Real Value

That said, AI has become genuinely useful in several areas that matter to investors.

Market pattern recognition. The Dubai Land Department now provides transaction data with remarkable transparency. AI tools can identify micro-trends within this data—which unit types in which buildings are transacting fastest, how price-per-square-foot has moved across specific corridors, where rental yields are compressing or expanding. For investors evaluating multiple opportunities, this analysis that once took weeks now takes minutes.

Document processing at scale. Off-plan investors often review multiple SPAs, payment plans, and developer disclosures simultaneously. AI can extract and compare key terms across documents—completion dates, penalty clauses, handover specifications—highlighting inconsistencies or unusual provisions. This does not replace legal review, but it focuses attention where it matters.

Rental market monitoring. For landlords managing multiple units, AI can continuously scan listing platforms to flag when comparable properties change asking rents, when vacancy periods extend beyond norms, or when new supply enters the immediate competitive set. This passive intelligence keeps investors informed without constant manual effort.

The Valuation Problem

Perhaps the most oversold application of AI in real estate is automated valuation. The technology can generate estimates, but investors should understand what those estimates actually represent.

Automated valuation models work by finding comparable transactions and adjusting for differences. The challenge in Dubai’s market is that “comparable” often means something very different than in mature Western markets. A branded residence and a non-branded unit in adjacent towers may share similar specifications but trade at dramatically different premiums. A developer’s reputation, the quality of facilities management, the specifics of a master community’s covenant—these factors drive value but resist quantification.

The practical takeaway: use AI valuations as one input among several, not as the answer. They are most reliable for commodity product in mature communities with deep transaction history. They are least reliable for unique properties, new areas, or anything where qualitative factors dominate.

What Changes in 2026

The more interesting development is not smarter valuations but what might be called “AI as analyst”—systems that can synthesise multiple information sources into coherent investment views.

Consider a scenario where an investor asks: “Show me buildings in JVC where average rental yields exceed 7%, construction completed within the last three years, and service charges remain below AED 15 per square foot.” A year ago, answering this required manually pulling data from multiple sources. Today, properly configured AI can return this analysis in seconds, with source citations an investor can verify.

This capability does not make decisions. It accelerates the filtering process that precedes decisions. For investors reviewing numerous opportunities, this compression of research time is the genuine productivity gain—not artificial intelligence, but analytical leverage.

A Framework for Evaluation

When developers or platforms promote AI capabilities, investors benefit from asking three questions.

First, what data does this system actually access? AI is only as good as its inputs. A valuation model trained on off-market transaction data will outperform one relying solely on listed prices. A tenant screening tool with access to court records provides more signal than one limited to self-reported history.

Second, what decisions does this replace versus inform? The most honest AI applications position themselves as decision support, not decision makers. Anything claiming to automate the investment decision itself should be treated with scepticism.

Third, what happens when the system is wrong? In real estate, errors compound. A flawed valuation can lead to overleveraging. A missed lease clause can trigger costly disputes. Understanding how AI systems handle uncertainty—and whether they communicate confidence levels alongside outputs—separates useful tools from impressive demonstrations.

The Human Element Remains

Dubai’s real estate market continues to mature in ways that favour informed investors. Transparency has increased. Data accessibility has improved. AI tools have become genuinely useful for specific tasks.

But the fundamentals of successful property investment have not changed. Location judgment, developer assessment, timing decisions, and the physical reality of individual assets still require human evaluation. The investors who will benefit most from AI are not those who delegate decisions to it, but those who use it to make better decisions faster—while keeping their feet firmly on the ground, often literally, walking the properties they intend to own.

About the Author

Kelly Jenkins is a speaker at the Innovation Experts Real Estate Institute and founder of Soulatics, an agentic AI consultancy. He advises organisations on practical AI implementation across property and development sectors.