Agentic AI’s Impact on Commercial Real Estate Goes Beyond Time Saved

Five years ago, underwriting a mixed-use deal took a commercial real estate analyst more than a week. Today, an agentic AI system can run the full analysis in about 90 minutes, with the analyst spending another 20 minutes reviewing the output, according to Leni’s head of industry Marcio Sahade, who previously spent 14 years at firms such as Tishman Speyer and Hines. That is the difference between bidding on three deals per quarter and bidding on 15.

The same purpose-built agentic system can read three complex retail leases in under seven minutes, producing a structured comparison of uses, rents, escalations and renewal options, while flagging unusual clauses and drafting language to address the biggest risks. This is the kind of work that can easily take up an afternoon.

Extraction has been treated as clerical for decades, and the firms pulling ahead are the ones that stop treating it that way. 

Nearly half of CRE investors still aren’t using data science in any meaningful way, according to Altus Group’s most recent research. The gap is narrowing, but slowly. By 2028, about a third of enterprise applications will incorporate agentic AI, up from less than 1 percent in 2024, according to market research company Gartner. In commercial real estate, where the most expensive labor is the time an analyst spends reading PDFs — not a broker commission or closing costs — this will be an important shift.

Today, the CRE industry still largely runs on paper: purchase and sale agreements, offering memoranda, trailing 12-month financials, rent rolls, environmental site assessments, property condition assessments, American Land Title Association surveys, mortgage docs, tenant estoppels, common area maintenance reconciliations, invoices. A staggering 80 percent of enterprise data lives outside databases in PDFs, scans, and email threads. 

Reading a complex commercial lease and pulling out the terms that matter takes an estimated four to eight hours and costs between $150 and $350, according to research by CBRE. Multiply that across a typical portfolio or an acquisitions package, and the real price of moving slowly becomes obvious.

In an attempt to increase efficiency and therewith effectiveness, some 92 percent of CRE firms have piloted artificial intelligence, but only 5 percent say they’ve achieved all of their AI goals, according to JLL’s 2025 Global Real Estate Technology Survey. The share of executives reporting a “transformative impact” from AI dropped to roughly 1 percent from about 12 percent a year earlier, according to Deloitte’s 2026 Commercial Real Estate Outlook. 

Adoption is rising faster than outcomes, and the reason is consistent across workflows: Firms have mostly pointed AI at the surface layer, at dashboards and chatbots and summary emails, when the work that actually determines whether a deal closes sits underneath — reading the documents, pulling the terms, running the math, and producing a defensible model. An acquisitions analyst who spends Monday through Thursday rekeying an offering memorandum will never be able to screen the next five deals. 

Then there’s asset management, where reporting is the tax every operator pays to own the asset. A typical manager reconciles the rent roll to accounting, pulls comps, writes a narrative, and pushes updated projections to limited partners. Much of that work is document archaeology: tracking down amendments, confirming common area maintenance exclusions, and reconciling variances buried in someone’s inbox.

In due diligence, the document problem can become a deal killer. The 30- to 90-day window requires underwriting the asset, reconciling the rent roll, commissioning key surveys, chasing tenant confirmations, and reviewing leases, amendments, contracts and title exceptions. On lean teams, that often means sampling 20 to 30 percent of leases, flagging what seems material, and hoping nothing critical gets missed. It ends up closer to triage than diligence.

But the potential impact extends well beyond individual workflows. Generative AI could unlock $110 billion to $180 billion in value for real estate if the industry actually goes after this, consulting firm McKinsey has estimated. That’s the size of a new market.

Arunabh Dastidar is the co-founder and CEO of real estate investment platform Leni.