AI now standard in CRE, but trust still lags behind

By mid-2026, the question in commercial real estate was no longer whether firms used AI, but how deeply it had been woven into routine work. Adventures in CRE’s Summer 2026 edition said AI had become “the expected baseline in a growing number of firms” across acquisitions, development, management, investor relations, and brokerage, a shift that helps explain why industry surveys could show adoption clearing the 65% mark even as confidence in AI-led deal judgment remained far more limited.

That framing matters because it shows AI was not sitting at the edge of CRE organizations as a pilot or novelty by late 2026; it was already embedded in operating practice. Adventures in CRE described AI as having moved from an optional technology to “a fundamental part of the industry's operations” and said it was essential for commercial real estate acquisitions and management, underscoring the central contradiction of the moment: widespread use had arrived before broad trust in AI for high-stakes decisions did.

The core failure is not that CRE lacks AI tools, but that firms are aiming them at information estates that were never standardized enough to support decision-grade outputs. Commercial Observer described companies trying to automate work still held together by scattered data, inconsistent definitions and spreadsheet-driven workflows, while Deloitte’s 2026 Commercial Real Estate Outlook found the share of executives reporting transformative AI impact fell from roughly 12 percent to about 1 percent in a single year, even as use spread—evidence that organizational readiness, not model capability, is the binding constraint.

That bottleneck becomes concrete when systems, metrics and ownership do not line up: JLL says real estate data often sits across disconnected tools, and Angela Barwick, executive director of Real Estate Technology Services, says 89% of Asia-Pacific companies are failing to get expected results from at least three technology platforms because poor planning creates compatibility and governance problems. Commercial Observer reports that when an owner or lender asks a new question, an analyst spends three hours stitching together a one-off presentation from disconnected workbooks and copy-pasting, while JLL’s 2026 findings say space data accuracy is the second-highest priority, with reporting quality close behind.

Across current CRE deployments, the most concrete payoff from AI is compression of analyst time, not delegation of the investment call. In Adventures in CRE’s Claude-plus-Excel demo, the speaker says that after pulling in deal inputs, “over the next 10 15 minutes… this is going to do the first pass underwriting,” while RXR similarly describes work that “used to take them 10 to 12 hours” now coming back ready for review, and a practitioner writes that “What used to be ninety minutes of typing into a template is now a few minutes of reviewing someone else's arithmetic instead of doing my own,” even as they still “adjust assumptions by hand.”

That pattern holds beyond underwriting: AI is proving itself by speeding preparation, explanation, and execution around decisions rather than replacing judgment itself. In trailing-twelve-month review, one user says the system “returns a narrative report on what moved and why… written in plain language instead of a spreadsheet full of red cells,” and they “run it before a T12 goes to a lender or an investor,” while Brixmor says it has cut lease-negotiation time by 10 to 15%; even more ambitious claims, like algorithms modeling cap-rate moves up to 24 months ahead or technology helping deliver India’s first 3D-printed post office in 43 days, still point to faster foresight and delivery as the clearest realized value.