Use is rising faster than trust. In AppFolio’s 2026 benchmark survey, 44% of respondents said they use AI in their roles (AppFolio blog), yet 78% said they “cannot yet rely on the AI features in their legacy property management software” (AppFolio). This guide covers the tool types, the adoption data, and the federal Fair Housing guidance and texting rules that changed in 2025 and 2026.
What is AI property management software?
It is software that does conversation and coordination work a leasing agent or office manager would otherwise do by hand. The tools differ most in how much of the job they finish on their own.
- Chatbot. Answers scripted questions from a decision tree. It cannot look up a unit or book a time.
- Answering service. Takes calls and messages, with live staff or AI, and passes them to you.
- Built-in PMS AI. Features inside your property management system, limited to the data that system holds.
- AI agent. Reads the inquiry, pulls listing data from connected systems, asks qualifying questions and books the showing.
Agents are still rare. The NARPM and AppFolio 2026 report found “just 7% have deployed agentic AI at all” (AppFolio blog). See the answering-service comparison and the built-in PMS AI comparison.
How many property managers use AI, and what results are they seeing?
Reported AI use ranges from 44% to nearly 80%, depending on the survey, but very few have measured what it does for them. All three surveys below were published by AppFolio, a software vendor, and rely on self-reported answers.
AppFolio 2026 benchmark. AppFolio “surveyed 1,617 U.S. based residential property management professionals from September 26, 2025 to November 3, 2025” (AppFolio). It reports 44% of all respondents using AI in their roles, a 29% increase year over year (AppFolio blog). AI adopters “expect an average portfolio growth of 31% in 2026” (AppFolio), which is an expectation, not a result.
IREM and AppFolio 2026 study. Fielded May 7 to June 7, 2026 among 3,662 real estate management professionals, it reports “AI adoption has nearly tripled since 2023, rising from 21% to 58%” (AppFolio). “Among AI users, 85% report productivity gains, but only 48% report a positive impact on net operating income.” Governance lags too: “Only 8% of organizations have a formal written AI policy, and 74% of professionals have received no formal AI training.”
NARPM and AppFolio 2026 report. “Nearly 80% of the fee managers we screened reported that their companies use AI,” mostly for “owner communication, leasing inquiries, marketing content, and maintenance coordination” (AppFolio blog). It reports that “44% say time to lease improved, 42% say maintenance resolution time improved” and “64% of owners prefer a property manager who uses AI.” But “Only 9% of managers have measured a quantifiable result from AI.”
Fraud is part of the picture. AppFolio reports “56% of managers report facing application fraud in the last year,” with “65% of those victims facing multiple types of fraud” (AppFolio). In the IREM study, 93% “are concerned about AI-generated fraudulent applicant documents” (AppFolio).
What kinds of AI property management tools are there?
There are four broad groups: AI built into a property management system, AI leasing agents, AI answering services and maintenance AI. The examples below are described only in their makers’ own words.
- Built-in PMS AI. AppFolio and Buildium both add AI features to their property management systems. The 78% figure above is a reason to test them on real inquiries before relying on them. See also RealPage Lumina alternatives.
- AI leasing agents. EliseAI offers LeasingAI, ResidentAI and EliseCRM, and says its “VoiceAI handles calls around the clock with zero hold time, offering both inbound and outbound calling capabilities” (EliseAI). It also claims “90% of our prospect workflows being automated,” a vendor claim. Haven “Responds to calls, texts and emails from Zillow, Apartments.com and every other listing site” (Haven).
- AI answering services and receptionists. Super “Takes and makes calls” and says “Two-way integrations allow Super to have context before the call even begins” (Super).
- Maintenance AI. Mason is “The AI-powered maintenance orchestration platform built for Rentvine customers” and “picks up the phone, troubleshoots issues with residents, dispatches your vendors” (Rentvine). Haven also says it “Answers every call, triages emergencies, dispatches the right vendor from your preferred list.”
None of the EliseAI, Haven or Super homepages lists a price. EliseAI and Haven ask you to book a demo. For side-by-side reviews, see EliseAI alternatives, Zuma AI leasing, answering services compared and the answering service guide for property management.
What can AI handle in leasing, and what stays human?
AI can run the conversation from first inquiry to booked showing. People should keep lease decisions, screening decisions, in-person tours and unusual cases.
| Leasing step | What AI can do | What stays human |
|---|---|---|
| Listing photos | Retouch listing photos | Take the photos and choose which to publish |
| Answering calls, texts, emails, chat | Reply on every channel, any hour | Review flagged conversations |
| Qualifying prospects | Ask move-in date, budget, bedrooms and the same set questions for everyone | Decide when context calls for a different path |
| Matching units | Find the unit that fits among your listings | Set prices and availability |
| Scheduling showings | Book open slots in a connected calendar | Conduct the showing |
| Confirming showings | Ask renters to confirm, and free unconfirmed times | Handle last-minute changes |
| Following up | Write to renters who did not book | Decide when to stop |
| Screening | Pass renters who want to apply to your team | Send the application, review the report and decide |
| Maintenance requests | Take the request and flag urgent ones | Approve repairs and respond to emergencies |
The tool handles volume and speed. A person handles decisions and exceptions. See the leasing automation guide and the after-hours guide.
Why does response speed matter?
Renters contact several properties, so the first clear answer often wins the showing. Every inquiry left waiting overnight is a chance for another listing to book that renter first.
Vacancy makes each lost lead cost more. The US rental vacancy rate was 7.6% in Q3 2026, up from 7.3% in Q2 2026 and 7.1% in Q3 2025 (FRED). In AppFolio’s survey, elevated vacancy was “cited as the top threat by 55% of respondents” (AppFolio).
Staffing is harder too. NAA reports rising leasing salaries: “Leasing professionals led the market with a sharp 28.4% year-over-year increase,” while “Unique job postings across core apartment roles declined 8.1% year-over-year” (NAA). See the missed-call cost guide, the inquiry response guide and the guide to listing-site leads after hours.
How does AI reduce showing no-shows?
It asks renters to confirm before the showing and frees the time when they don’t. One case study reports a large jump in show-ups after turning this on.
In a ShowMojo case study, Jarod Puckett of Arkansas Homes and Rentals in Little Rock said that before automated confirmations, “somewhere between 30% and 40% of prospective renters who scheduled a showing actually showed up” (ShowMojo). After, he said: “I think we’re up there in the 80, 90% of the people show up for their appointments.” That is one manager’s estimate, not a study.
The sequence in that case is simple to copy:
- “24 hours in advance of the showing,” send a confirmation.
- If there is no confirmation, try again “a few hours before the showing.”
- “If there’s still no response, the showing is canceled,” and the time opens up.
See the no-show reduction guide.
Does the Fair Housing Act apply to AI leasing assistants?
Yes. HUD said in 2024 that the Fair Housing Act applies to tenant screening and the advertising of housing, including when artificial intelligence is used, even though HUD has since withdrawn some of its 2024 guidance.
On May 2, 2024, HUD said “the Fair Housing Act applies to tenant screening and the advertising of housing, including when artificial intelligence” is used (HUD). It spoke of “ensuring that tenant screenings are transparent, accurate, and fair,” and said the screening guidance “suggests best practices for fair, transparent, and non-discriminatory tenant screening policies.”
That guidance has since changed:
- Advertising guidance withdrawn. HUD withdrew its April 29, 2024 guidance on the Act and the advertising of housing, with an “Effective date of withdrawal: September 17, 2025” (Federal Register).
- Screening guidance taken offline. GAO reported: “In February 2025, we observed that HUD had removed the tenant screening guidance from its public website.” HUD told GAO it was “taken down as part of an agencywide review,” and also that “the guidance remained in effect” (GAO). No later HUD statement was found.
- Criminal-records guidance withdrawn. HUD’s 2016 guidance on the use of criminal records was withdrawn “effective September 25, 2025” (Federal Register).
The Act itself did not change. HUD’s withdrawal notice says “Any actions that do not comply with the text of the Fair Housing Act continue to be subject to enforcement” (Federal Register). The Act prohibits discrimination because of race, color, national origin, religion, sex, familial status and disability (HUD).
Courts have acted on AI screening too. SafeRent Solutions reached a $2.3 million class settlement with final approval in the District of Massachusetts, and for five years won’t include “a SafeRent Score or an accept/decline recommendation based on a tenant screening score” (Bloomberg Law).
What should a fair housing check of an AI leasing assistant include?
Check what it says, whether you can read every conversation, and whether every renter gets the same treatment. No vendor label replaces that review.
- Same questions for everyone. Qualifying questions should be the same for every prospect for the same unit. Scripts and templates in this guide never screen people by a protected class.
- Off-limits topics. Ask the vendor what the AI does when a renter mentions children, a disability, religion or national origin, then test it.
- Readable logs. You should be able to read every message and call summary, by property.
- Human decisions. The AI passes renters who want to apply to your team; a person runs the screening and decides.
See the AI disclosure guide, the tenant screening guide and the state screening law pages.
What texting and calling rules apply to AI that contacts renters?
Federal TCPA rules require consent before automated marketing texts and calls, and they require you to honor opt-outs quickly. Rules on opt-outs changed again in 2026.
| Federal rule | What the FCC says | Source |
|---|---|---|
| Marketing robotexts and robocalls | Written consent is required for marketing messages | FCC FAQ |
| Artificial or prerecorded voice calls | Non-emergency calls need “the prior express consent of the called party” | FCC FAQ |
| AI-generated voices | Held “to those same standards” as other artificial voices | FCC, Feb 8, 2024 |
| One-to-one consent rule | Repealed July 14, 2025 | FCC DA 25-621 |
| Opt-out timing | “within a reasonable time not to exceed ten business days from receipt of such request” | FCC 26-67 |
Consent. FCC rules require prior express written consent for marketing calls using an artificial or prerecorded voice, and robotexters need written consent for marketing messages (FCC). In February 2024 the FCC said its ruling “makes voice cloning technology used in common robocall scams targeting consumers illegal” (FCC). An AI voice that calls renters is covered.
One-to-one consent. In Insurance Marketing Coalition v. FCC, the Eleventh Circuit vacated the rule requiring consent for each seller “just days before” its January 27, 2025 effective date (Morrison Foerster). The FCC’s July 14, 2025 order then “repeals the revised version” of the rule and reinstates the earlier version (FCC).
Opt-outs. The FCC’s revocation rules applied from April 11, 2025, with a waiver for one part, section 64.1200(a)(10) (FCC DA 25-312). FCC 26-67, adopted September 30, 2026, says a revocation in response to an informational call or text may be read as covering only that category, while revoking marketing covers “all future calls or text messages containing an advertisement” from that caller (FCC). Its amendments take effect “30 days after Federal Register publication.”
Ask any vendor how it collects consent, stores it and processes opt-outs. See the texting rules guide.
Is there free AI for property management?
Buildium’s own blog lists a 14-day free trial with no credit card required (Buildium). None of the EliseAI, Haven or Super homepages lists a price. EliseAI and Haven ask you to book a demo.
General chat tools can help you draft a reply. They don’t connect to your listings, calendar or message history, so you can’t easily check what each renter was told, which matters for the Fair Housing and consent checks above. See the AI leasing agent cost guide.
How do I evaluate and choose AI property management software?
Start with your current tools, lead volume and team size, then test each product on the same inquiry. Judge the finished work, not the category name.
- Integration. Does it connect to your property management software, calendar and listing sites?
- Channels. Does it handle calls, texts, email and website chat, or only some?
- Fair Housing checks. Can you read every conversation? Does it ask every prospect the same questions?
- Consent and opt-outs. How does it collect consent and honor opt-outs within ten business days?
- Scheduling and reminders. Does it book in your calendar, confirm before showings and free unconfirmed times?
- Reporting. Can you see leads, response times, showings and no-shows by property?
- Cost structure. Compare per-door, per-listing, per-minute and flat pricing against the same workload.
- A written AI policy. Only 8% of organizations have one, per the IREM study (AppFolio). Write down what the AI may do and who reviews it.
Send each tool the same test inquiry. For example: “Hi, is the two-bedroom still available? I’d like to move in next month and see it this week.” Then check that it answers, asks about move-in date and budget, offers a time, and asks nothing about a protected class.
Nolan answers calls, texts, email and website chat, books showings in your Google Calendar, sends reminders and follow-ups, and has connectors for property management software and CRMs. Its plan is $399 a month for 10 active listings, then $15 a month for each active listing after the first 10, with a 14-day free trial and no setup fee. For more options, see the AI leasing agent guide and the rental property management apps guide.
How do I test whether AI saves time and money?
Measure one property before and after. The NARPM report found “Only 9% of managers have measured a quantifiable result from AI” (AppFolio blog), so your own numbers will tell you more than any survey.
- Pick one property with steady inquiry volume.
- Record a baseline month: first response time, showings booked, share of showings attended, leases signed and staff hours on inquiries.
- Turn the tool on for a test month and record the same numbers the same way.
- Read a sample of conversations each week for accuracy, tone and fair treatment.
- Compare cost per lease for both months, including the software and staff time.
Keep everything else steady during the test, such as rent and listing sites. If the numbers improve and the conversations hold up, roll it out to the next property.






