AI and Property Tax Assessments: Proceed with Caution

By Carlo L. Batts, MAI, Rittenhouse Appraisals and The Reduxx Group

Artificial intelligence (AI) promises to shake up many industries including commercial real estate. One part of that promise would be revolutionizing property tax assessments, by improving accuracy, analyzing massive datasets, and streamlining bureaucratic processes.

But as someone who has spent decades working commercial property valuation and tax assessment, I see significant risks in how AI is being deployed in this space. The technology may actually exacerbate existing problems, causing property owners to overpay taxes on portfolios that have already lost significant value.

The Valuation Crisis

Commercial real estate valuations are currently at odds with governments’ appetite for tax revenue. Economic volatility, changing property management standards, interest rate increases and rising capital costs have depressed commercial property values across many markets. Yet assessments haven’t kept pace with these market realities.

Enter AI, which many tax authorities view as a tool to modernize assessment processes. The problem? AI learns from available knowledge, and in property taxation, that foundational knowledge is often flawed, incomplete, or outdated.

Where AI Falls Short

The limitations of AI in property tax assessment are substantial and specific:

Unique characteristics of properties or markets get lost. Properties have distinctive features, which influence value, including location nuances, tenant mix, physical condition, and market positioning. AI struggles to weight these factors appropriately, especially when many jurisdictions maintain artificially high assessments to preserve revenue.

Past data misleads. Property valuation isn’t linear. AI trained on historical data can’t adequately account for market disruptions, changing use patterns, or economic shocks. While steadily increasing assessments might serve government budgets, they don’t reflect current market conditions.

The black box problem. AI decision-making processes lack transparency. Without visibility into the logic behind valuations, identifying and correcting errors becomes nearly impossible. Property owners can’t effectively challenge assessments they don’t understand.

Methodology inconsistency. AI requires standardized frameworks to function effectively. For emerging property types like data centers, or properties with unique characteristics, lack of comparable data and standardized valuation approaches leads to unreliable assessments.

Data scarcity. Critical information such as rents, occupancy rates, operating costs, and tenant details often remain private. AI can’t accurately assess what it can’t access. Many local jurisdictions also lack the technology infrastructure to feed quality data into AI systems.

Inadequate oversight. Most local and state taxing authorities haven’t budgeted for AI training or professional oversight. Without human experts reviewing AI outputs, errors compound and inaccuracies persist.

A Tool, Not A Solution

I’m not anti-technology. AI can play a valuable role in streamlining research and identifying broad patterns. But it cannot replace the nuanced judgment that experienced professionals bring to property valuation.

As AI becomes more prevalent in property tax assessment, commercial property owners must remain vigilant. Regular portfolio reviews, understanding local assessment methodologies, and challenging inaccurate valuations become even more critical when algorithms, rather than experienced assessors, drive the process.

The goal should be accuracy, not automation. Until AI can fully account for market complexity, local nuances, and real-time conditions, human expertise remains essential to ensuring fair property tax assessments.

Carlo L. Batts, MAI is the principal of Rittenhouse Appraisals and The Reduxx Group, both based in Center City Philadelphia.