AI Is Changing Valuation, But the Critical Dataset Is Still Human Judgment
Commercial real estate is entering a new phase of artificial intelligence adoption, and the most important signal is not speed. It is accountability. In a recent piece for the New England Real Estate Journal, Josephine Aberle, MAI, argues that AI can help organize valuation work, but it cannot sign the certification. For KG Data readers, that distinction is more than professional philosophy. It defines the boundary between automation and reliable property intelligence.
Valuation has always been a data-intensive discipline. Appraisers work with sale comparables, rent comparables, expense benchmarks, cap rate surveys, operating statements, absorption data, construction budgets, financing assumptions, and market studies. AI can summarize these inputs faster than any human analyst. It can draft narrative sections, compare language across reports, flag inconsistencies, and convert unstructured documents into searchable information. Those are real productivity gains.
But productivity is not the same as credibility. The current commercial real estate market is fragmented by sector, location, capital availability, lease structure, tenant risk, and asset condition. A multifamily development, an office tower, a data center, an industrial facility, and an affordable housing asset may all sit inside the same broader market report, but they do not share the same risk profile. AI can process the words around those risks. It cannot independently decide which risks deserve the most weight in a value conclusion.
The data problem is becoming more subtle. Real estate no longer suffers from a lack of information. It suffers from uneven information quality. Closed sale data may be stale, incomplete, or distorted by unusual financing. Rent data may reflect asking rents rather than executed leases. Cap rate surveys may lag market movement. Offering memoranda may present optimistic assumptions. AI systems trained or prompted on these materials can reproduce their weaknesses with confidence and polish.
AI can accelerate valuation workflows, but it cannot convert weak inputs into reliable conclusions.
This is where appraisal becomes a useful model for broader proptech adoption. The appraiser’s role is not merely to collect data, but to test relevance, reliability, and context. Is a sale truly comparable? Is an income assumption supportable? Does a cap rate reflect the specific asset, or only the market mood? Would a lender, regulator, investor committee, or court accept the reasoning? These questions are analytical controls. They are also governance controls.
For technology teams building valuation tools, the lesson is clear: the next generation of AI products should not only generate text or summaries. They should create transparent audit trails. Users need to know which data was used, where it came from, how current it is, what assumptions were applied, and what confidence limits surround the output. Explainability is not a luxury in valuation. It is part of the risk infrastructure.
For lenders and investors, the practical question is changing from “Was AI used?” to “How was AI controlled?” Institutions should track whether AI-assisted reports include source verification, assumption testing, exception flags, human review, and clear responsibility for final conclusions. The market will increasingly reward firms that can combine faster analysis with stronger validation.
The signal to watch is not whether AI enters commercial real estate valuation. It already has. The more important indicator is whether firms treat AI as an assistant to professional judgment or as a substitute for it. In uncertain markets, the advantage will belong to teams that can move quickly without weakening trust. Data improves judgment only when someone qualified remains accountable for what the data is allowed to mean.
Source: New England Real Estate Journal
