AI for property management: what it can run and what a person decides

AI is reliable in a property management company wherever the work follows a rule: checking that an application is complete, reading documents so a person can review them, sorting maintenance requests, chasing renewals and assembling the monthly owner report. It should not approve or decline an applicant. Screening decisions, adverse action notices and anything that touches fair housing stay with a person on your team.

That split matters most at a 5 to 30 person firm, where nobody in compliance catches a bad call and the people answering phones also run screening, maintenance and owner statements.

How many property managers use AI in 2026?

Somewhere between a third and a half, depending on who asks.

AppFolio surveyed 1,617 US residential property management professionals between September 26 and November 3, 2025. Multifamily Executive's write-up of that report puts AI use at 44% of respondents.
Buildium's May 2026 survey of nearly 1,000 property management professionals found that 33% of companies already use AI or automation.

The more telling number is how far that use goes. In Buildium's earlier 2026 report, just 8% of respondents said they had been able to fully automate any workflow. The same report says property managers most often use AI to write property descriptions and customer communications.

So the most common use is drafting text. At most firms, the chasing, checking and reporting around that text is still done by hand.

Where is AI reliable in a property management company?

Five jobs, all with the same shape: a clear rule, a lot of repetition, and a person who signs off before anything consequential goes out.

1. Checking that an application is complete

An application arrives without the second pay stub, or with an ID photo nobody can read, or with a co-applicant who never filled in their part. Someone finds out three days later.

This is a checklist, and a checklist is the easiest work to hand off. AI compares each application against your written list of required items the moment it arrives, and asks the applicant for what is missing. A person sees the file only when it is complete, or when the applicant has stopped answering.

2. Reading documents, with a person deciding

Fraud is what makes this job hard. Checkr surveyed 1,400 landlords and property managers from August 13 to 18, 2026. 67% said AI-generated fraud is harder to catch than it was a year ago, and 37% named fake or altered pay stubs and bank statements as their biggest concern. In AppFolio's survey, 56% of managers reported facing application fraud in the last year.

AI can read a pay stub and lift gross pay, deductions, net pay, year to date and pay dates, then redo the arithmetic a payroll system would have done. We walk through those checks in how to verify a pay stub. The output is a marked-up file: these figures add up, these do not, this net pay never shows up in the bank statement. A leasing manager reads the file and decides what to ask for next.

3. Sorting maintenance requests

Buildium's October 2026 research on where property managers have started using AI puts resident communication and support first at 48%. Rent collection, maintenance triage, and leasing and renewals each land between 34% and 36%.

Most requests sort themselves once someone reads them. Water coming through a ceiling is urgent. A sticking cabinet is not. AI reads the request and any photo, labels it emergency, urgent or routine under your rules, matches it to the right vendor for that property, and drafts the work order. A person approves any dispatch above a spending limit and anything the rules do not cover.

4. Renewal follow-up

A lease that ends on June 30 should have a renewal conversation in March or April, not in late May. The work is a calendar plus a sequence of messages.

Ninety days before the end date, the sequence starts: a reminder to the property manager to set terms, then a drafted offer to the resident, then follow-ups at set intervals, then an alert when the resident has not answered. The person decides the rent, the term and whether to renew at all.

5. Owner reporting

Only 38% of owners in Buildium's 2026 report said they are consistently profitable. When most owners are not reliably in the black, the monthly statement has to explain the month, not just list it.

The statement is built from data you already hold: rent collected, open work orders, vacant days, renewals pending, bills paid. AI pulls it together, writes a plain summary of what changed, and flags anything unusual, such as three plumbing calls at one unit in a month. A person reads it and sends it.

Where must a person decide?

Approving, declining and setting conditions on an applicant. Those are fair housing and consumer reporting decisions, and the law ties them to you, not to your software.

Fair housing

The Fair Housing Act bars discrimination based on race, color, national origin, religion, sex, familial status and disability. HUD has withdrawn a number of fair housing guidance documents since 2025, but its April 2026 withdrawal notice is direct about the law itself: "Any actions that do not comply with the text of the Fair Housing Act continue to be subject to enforcement by the Department." The same notice points out that people can still file a civil action in federal or state court within two years.

A model shown your past approvals can repeat past patterns. Give it the written criteria, not your history, and keep it out of the decision.

Adverse action

The FTC's guidance for landlords is specific. A consumer report includes a "risk score or recommendation from a tenant screening company." If you deny an application, require a co-signer, or ask for a deposit you would not ask of another applicant, based even partly on a report, you must send an adverse action notice. That holds even when the report was not the main reason.

The notice names the screening company and how to reach it, says the company did not make the decision, and tells the applicant they can get a free copy of the report if they ask within 60 days and dispute anything inaccurate. If a credit score was used, the notice includes the score and the key factors that hurt it.

Screening reports can be wrong

In 2023 the FTC and the CFPB reached a $15 million settlement with Trans Union over tenant screening reports with inaccurate and incomplete eviction records: a filing reported without the dismissal, or a judgment reported when the case was actually dismissed. A score built on a bad record is a bad score. A person who reads the report can ask the applicant about it. A rule that auto-declines below a number cannot.

This is not legal advice. State and city rules on screening, income standards and notices vary, so check yours, or ask counsel, before you change your criteria.

Job, what AI does, what a person still decides

Job	What AI does	What a person still decides
Application completeness	Checks each file against your list and asks for missing items	Whether to extend a deadline or accept a substitute document
Income documents	Reads stubs and statements, redoes the arithmetic, marks mismatches	What to ask for next, and which income figure to use
Screening report	Attaches it to the file and summarizes it	Approve, decline or set conditions, and send the adverse action notice
Screening criteria	Applies the written criteria the same way to every file	What the criteria are, and every change to them
Maintenance requests	Labels urgency, picks the vendor, drafts the work order	Any dispatch over the spending limit, any unclear safety issue
Renewals	Starts the sequence 90 days out, drafts the offer, follows up	The rent, the term and whether to renew
Owner statements	Assembles the numbers, drafts the summary, flags unusual items	What goes to the owner, and what to recommend

How do you automate rental application screening without automating the decision?

Automate everything before the decision and everything after it. Leave the decision to a person.

Write the criteria down. Income ratio, credit standard if you use one, rental history, and how you treat criminal records under your state's rules. One document, applied to every applicant.
Check completeness on arrival. Ask for missing items automatically, on a schedule, until the file is complete.
Read the documents. Lift the figures, redo the arithmetic, compare stated income with what the documents support.
Order the screening report once the file is complete, not before.
Assemble one file with every mismatch marked and the criteria beside it.
A person decides and records the reason in plain words.
Draft the adverse action notice whenever a report played any part in a denial or a condition, and have the person who decided send it.
Keep the record. Who decided, when, on which file, and what was sent.

Steps 2 to 5 and 7 are where the hours go. Step 6 is where the liability sits.

What should you measure?

Each is a timestamp or a count. When the answer sits in an email thread, nobody can add it up.

Hours from application received to complete file.
Hours from complete file to decision.
Days from maintenance request to vendor dispatch, by urgency.
Share of leases with a renewal offer out 90 days before the end date.
Owner statements sent on time each month.
Applications flagged for a document mismatch, and how many of those cleared.

Pick one job, measure it for a month, hand it off, and measure again.

What dailybuilt does for property managers

dailybuilt is the thinking layer for your company. For a property management firm, our team maps the admin work, sets up AI to do it, and builds the workflow on the dailybuilt platform. A person on your team approves before anything goes out, and each month you get a report of the work handled and numbers you never had.

The platform underneath is software your team logs in to: custom collections we set up for units, leases and applications, with 15 field types, automations with 31 event triggers plus schedules, waits and branches, documents and e-sign, an inbox for website forms and email, invoices and card payments with no platform fee, and an audit log of every change. You can export all of your data at any time.

It is a monthly retainer, scoped on an intro call. See how it works for property managers.