From the team

Commercial real estate teams choosing underwriting tools in 2026 face three options: established legacy software, general purpose horizontal AI assistants, and vertical AI platforms built specifically for commercial real estate. Legacy tools offer depth but little native AI, horizontal tools offer flexibility but no domain grounding, and vertical AI combines domain specific modelling with AI across the whole workflow.

The three approaches explained

Almost every underwriting tool a commercial real estate team could adopt falls into one of three categories, and naming them clearly makes the choice much easier.

Legacy software means the established modelling suites the industry has used for years. They are deep and trusted, and they predate the current wave of AI.

Horizontal AI means general purpose assistants that can help across many tasks. They are powerful and flexible, and they know nothing about commercial real estate beyond what is in their training data.

Vertical AI means platforms built specifically for one industry, with the domain modelling embedded and AI running through the workflow.

Where legacy software wins, and where it holds you back

Legacy suites win on depth and familiarity. They model complex structures, they are embedded in existing workflows, and a generation of analysts already knows how to use them. For firms that value continuity above all, that is a real advantage.

They hold you back on speed and automation. Most were designed before modern AI and still expect heavy manual input, so the analyst spends time keying data rather than making decisions. Bolting AI onto a tool built for a different era is not the same as building around it.

Why horizontal AI struggles with CRE underwriting

General assistants are genuinely useful for thinking, drafting and summarising. The problem in underwriting is that they have no ground truth. Ask one to run a development appraisal or model a waterfall and it will produce something that reads fluently and may be quietly wrong, because it has no embedded model to check itself against.

For work where a plausible answer and a correct answer are not the same thing, and in underwriting they rarely are, that gap matters.

What vertical AI does differently

Vertical AI starts from the domain. The appraisal logic, the debt structures, the tenancy schedules, the waterfalls and the market conventions are built in, so the model is correct out of the box. AI then runs on top of that foundation, reading brochures, populating assumptions, forecasting scenarios and drafting commentary, while the analyst keeps control of the judgement.

The effect is that you get the automation and speed of AI with the reliability of a purpose built model.

Moving on from an older tool? What to consider

Horizontal AI will tell you anything. Legacy software will make you type everything. Vertical AI is the bet that you can have the intelligence and the rigour in the same tool.