10
Loading video...
By Edward Sapp
Founder
0
Share
Newsweek is a Trust Project member
See more of our trusted coverage when you search.Prefer Newsweek on Googleto see more of our trusted coverage when you search.
Artificial intelligence (AI) is making commercial real estate underwriting faster, but faster is not the same as better.
As AI moves from experimentation into investment workflows, the greatest danger may not be that the technology produces obviously wrong answers.
It may be that it produces convincing answers that cause investors to stop questioning the assumptions underneath them.
Read More on Leadership Insights
That distinction matters in an industry where a seemingly minor assumption can materially change the economics of a deal.
...
A different expense forecast, a missed lease provision, an incorrect financing assumption or an overly optimistic exit valuation can alter an investment thesis long before anyone realizes the model was wrong.
The technology is moving quickly. Deloitte's 2027 Commercial Real Estate Outlook found that 92 percent of surveyed CRE organizations were still in the research or piloting phase of AI adoption, while only 8 percent had integrated AI solutions.
At the same time, fewer than half reported implementing more advanced controls such as challenger models, exception handling or detective controls.
Deloitte's survey covered 950 executives and their direct reports at commercial real estate owners and investment organizations.
That tells me something important. The question is no longer whether AI will enter real estate. It already has. The question is whether the industry will learn how to use it without confusing computational speed with investment judgment.
False Confidence
AI is exceptionally good at processing information. It can review hundreds of documents, organize lease data, identify discrepancies, compare assumptions and test scenarios far faster than any person working manually.
PricewaterhouseCoopers (PwC) and the Urban Land Institute's Emerging Trends in Real Estate 2026 report found that AI use is expanding across real estate, including in research, underwriting and reporting.
Those capabilities can be enormously valuable. But they also introduce a subtle risk: when an output arrives quickly, neatly formatted and supported by thousands of data points, it can feel more authoritative than it deserves to be.
I call that false confidence.
An underwriting model has always been a representation of reality, not reality itself. AI does not change that.
It can make the process of building and interrogating a model dramatically more efficient, but it cannot automatically determine whether the assumptions being fed into that model accurately reflect what is happening at a specific property.
That becomes particularly important in commercial real estate because properties are not interchangeable datasets.
Human Judgment
A retail asset might look attractive based on comparable rents and historical occupancy, while years of tenant turnover reveal a parking constraint that is affecting leasing.
An office building might appear inexpensive relative to comparable sales, while its deferred capital expenditures tell a different story.
A property may have a ground lease that resets under terms that materially change the economics. Property taxes may change following a transaction. Expenses previously absorbed by an owner may become the responsibility of the next owner.
Some of the most important information in a deal is therefore not necessarily the information that is easiest to structure.
This is where I believe human judgment remains indispensable.
As the founder of Sapp Capital Advisors, I have spent years working through the assumptions behind real estate investments. That experience has taught me that good underwriting is not simply about producing a model.
It is about understanding why the numbers look the way they do, testing the assumptions behind them, and identifying which details could materially change the outcome.
Accelerating Questions
AI can be a great multiplier on human judgment, but it is not a replacement for human judgment.
Consider a relatively straightforward example: reviewing leases. An AI system can potentially process hundreds of documents and build a matrix showing rents, expiration dates, options and other provisions. That can save an enormous amount of time.
But finding a clause is not the same as understanding its significance.
A model may identify a lease expiration. An experienced underwriter needs to ask what happens next. Is the tenant likely to renew? What does the local market suggest? Are there competing properties offering better terms? Does the tenant have unusual negotiating leverage? Is the space physically suited to another user if the tenant leaves?
The technology can accelerate the questions. It does not eliminate the need to ask them.
The same principle applies to market data. When transaction activity is limited, comparable sales can become less informative. Historical data may not fully capture current conditions, particularly in markets undergoing structural change.
Relationships with brokers, owners, lenders and other market participants can provide context that does not necessarily appear in a spreadsheet.
The Great Potential
That does not make human judgment infallible. People have biases, make mistakes and can become attached to their own assumptions.
AI can actually help expose those weaknesses by making it easier for investors to test scenarios more extensively and examine information that might otherwise have been overlooked.
That is where I see the technology's greatest potential.
Instead of asking AI to make an investment decision, investors can use it to expand the amount of analysis they are capable of performing.
A professional who once had time to review 20 leases might be able to review hundreds. Someone who previously tested three scenarios might be able to test dozens.
A discrepancy between seller materials and underlying documents can potentially be identified much faster.
The value comes from extending human analytical capacity, not eliminating human accountability.
The Changing Real Estate Investment Environment
Real estate firms adopting AI should therefore build verification into the investment process: question the assumptions, test the scenarios, identify what the model cannot see, and independently validate the information that could materially affect the decision.
The goal should not be to make investment decisions faster simply because AI can. It should be to make them more thoroughly informed while keeping responsibility for the final judgment firmly with the people making the investment.
This becomes especially important as the commercial real estate investment environment evolves.
CBRE expects U.S. commercial real estate investment volume to increase 16 percent in 2026, with investment activity increasing across the major property types.
As transaction activity picks up, there will be more opportunities to apply AI across acquisition, underwriting and asset management.
That makes disciplined implementation more important, not less.
A Starting Point
The temptation will be to treat AI-generated analysis as an endpoint. I think it should be treated as a starting point.
One way to do that is to make the investment process explicit. What assumptions are being made? What information supports them? Which assumptions are most sensitive to changes? What would have to be true for the investment thesis to fail? What information is missing from the dataset? What should be independently verified before capital is committed?
Those questions are not obstacles to AI adoption. They are what make AI adoption useful.
Firms can also learn from their own results. If an AI-assisted underwriting process identifies a particular risk, was that risk actually present?
If a projected expense differs from the eventual result, why? If a model repeatedly performs well in one area but poorly in another, that information should feed back into the process.
In other words, AI should become part of an evolving institutional knowledge base rather than a black box that produces answers.
Cautiously Optimistic
This is particularly important because real estate underwriting remains highly contextual. Two properties can have identical square footage and similar rents while representing completely different investments.
The surrounding market, physical condition, tenant relationships, financing structure, lease language, regulatory environment and future capital requirements can all change the equation.
No model can remove that complexity. What AI can do is help professionals navigate it more efficiently.
That is why I am optimistic about the technology while remaining cautious about the way it is interpreted. The industry should not reject AI because it can be wrong. Humans can be wrong, too.
The goal should be to build processes where technology and professional judgment challenge each other rather than where one replaces the other.
Speed Is Not Certainty
The biggest risk is not necessarily an AI system producing an obviously absurd underwriting model. An obviously bad model gets rejected.
The more dangerous scenario is a sophisticated-looking model that contains a flawed assumption and gives everyone involved enough confidence to stop asking questions.
Real estate has always required judgment because the future cannot be reduced entirely to historical data. AI can process more information, identify patterns and test scenarios at a scale that was previously impractical.
But when the cost of being wrong involves millions of dollars, speed should never be confused with certainty.
AI can multiply human judgment.
It cannot replace the responsibility to exercise it.
Edward Sapp is the founder of Sapp Capital Advisors, which provides paid commercial real estate underwriting, financial modelling and model-review services.
The views expressed in this article are the author's own.
View 0 Comments
Submit Correction
View Editorial & AI Guidelines