Published 6 minutes

How to use ChatGPT in property management

For property managers and landlords in Canada and the US. Seven prompts you can copy for real property management work, why each one works, what never to paste in, and where a chatbot stops and an AI agent takes over.

AI in property management
A person in a mustard sweater typing on a laptop at a wooden desk

This guide is for property managers and landlords in Canada and the US who want to use ChatGPT, or a tool like it, in their day-to-day work. It’s good at the writing and thinking around the job: replies to residents, listings, plain-language explanations, summaries, comparisons and owner updates. You get the most from it by giving it the facts, the tone and the format you want, and by reading every answer before it goes anywhere.

Below are seven prompts you can copy and adjust. Each one is written the way that gets a usable answer the first time, with a note on why it works. After them come the things never to paste in, and the point where a chatbot hands over to something that can do the work itself.

1. Reply to a resident who’s behind on rent

I manage a rental apartment. A resident is 9 days late on October rent of $2,150 and has texted to say they lost shifts at work and can pay $1,000 now and the rest on the 25th. Write a short, friendly reply that thanks them for letting us know, accepts the plan, and confirms both dates and amounts in writing without sounding harsh. Plain language, no more than 120 words.

Every fact it needs is in the prompt: the amounts, the dates and what you want to happen. Without them, ChatGPT fills the gaps with guesses, and that’s when replies go wrong. The word limit keeps it from writing a letter when a text is what the resident sent.

2. Turn your notes into a rental listing

Write a rental listing from these notes. Headline first, with area, rent and move-in date. Then two short paragraphs: the home, then what’s included and the rules. Facts only, no words like “stunning” or “must see”, and describe the home, not the kind of person who should live there. Notes: 2 bed, 2nd floor duplex, near Ossington station Toronto, $2,450, Dec 1, ~850 sq ft, kitchen redone 2025, dishwasher, in-suite laundry, heat and water incl, hydro extra about $60/mo, cats ok no dogs (shared yard), street parking with permit, 1 year lease.

The instructions do as much work as the notes. Telling it to leave out the adjectives and to describe the home rather than the tenant avoids the two most common problems with AI-written listings: vague copy, and wording that can break human rights or fair housing rules. Check every fact in the result against your notes before you post it. How to write a rental listing that gets the right inquiries goes through what a good one includes.

3. Explain a lease clause in plain language

A resident has asked what this clause in their lease means. Explain it in plain, friendly language at a grade 8 reading level, in three or four sentences, without changing its meaning or giving legal advice. Clause: “[paste the clause]”

Residents often ask what a clause means, and the honest answer is usually simpler than the clause. Asking for a reading level and a length gets you an answer you can send. Asking it not to change the meaning is a reminder you should check for yourself, because a plain-language version is easy to make slightly wrong.

4. Summarize a long email thread with an owner

Here is an email thread between me and a property owner about a roof repair. Summarize it as: decisions made, open questions, and who owes what next, with dates. Bullet points, nothing else. [paste the thread]

A thread that’s grown to twenty messages hides what’s actually been agreed. Asking for decisions, open questions and next steps, in that order, turns it into something you can act on, or forward to a colleague who’s taking over.

5. Compare vendor quotes

Put these three roofing quotes into a table with columns for price, what’s included, what’s excluded, warranty, start date and anything unusual. Then, in two sentences, say which questions I should ask each vendor before choosing. [paste the quotes]

Quotes are hard to compare because every vendor lays them out differently. A table makes the gaps obvious, like one quote that leaves out disposal or another with a shorter warranty. Asking for questions rather than a pick keeps the decision with you.

6. Draft the monthly owner update

Write a short monthly update for the owner of a fourplex from these notes. Warm and professional, in the order: rent collected, repairs done, anything coming up, anything we need from them. Under 200 words. Notes: all 4 units paid, unit 2 paid on the 8th. Replaced the hot water tank in unit 3, $1,850, approved Oct 4. Unit 1 lease ends Jan 31, resident wants to renew. Need owner to decide on 3% increase vs keeping rent the same.

Owners read updates quickly, so put money first, then what happened, then what’s next, then what you need from them. Putting the decision you need at the end, clearly, is what gets it answered.

7. Check a message before you send it

Here’s a message I’m about to send a resident about entering their unit for a repair. Check that it’s clear, polite and complete: the date, the time window, the reason and who will be there. Point out anything missing or anything that could come across badly. Don’t rewrite it unless I ask. [paste the message]

This is a useful habit that’s easy to skip. A second read of an important message catches a missing time window or a sentence that reads as cold. Asking it to point things out rather than rewrite keeps the message in your own voice.

Notices like this one have rules about timing and content that vary by province and state, and some notices must be on an official form, such as the Landlord and Tenant Board forms in Ontario. ChatGPT can help with the wording around a notice, but not replace the form.

What never to paste in

ChatGPT only needs the facts the task needs, and most tasks don’t need much personal information. Leave out:

  • Social Insurance or Social Security numbers
  • Bank account and card details
  • ID numbers, like driver’s licence or passport numbers
  • Credit reports and background check results
  • Anything a resident told you in confidence that the task doesn’t need

Where you can, use a unit number or initials instead of a full name. Then turn off training. On the personal plans (Free, Plus and Pro), your conversations can be used to train OpenAI’s models unless you switch it off. Go to Settings, then Data controls, and turn off “Improve the model for everyone”. On the business plans, such as ChatGPT Business and Enterprise, company content isn’t used for training by default, according to OpenAI’s help centre. If your team uses it for resident or owner work regularly, a business plan with a clear policy on what goes in is worth having.

Where a chatbot stops

Look back at the seven prompts, and each one ends the same way. ChatGPT hands you text, and then you copy it, paste it into your email or your software, send it, and update the record. It can’t see the rent ledger, book the plumber, send the reply or remember what the resident said last month, unless you tell it every time.

That’s the difference between a chatbot and an AI agent. An agent works inside your systems and finishes the job itself. Omnia by Tuze is an example built for property management. Its AI teammates answer every call, text and email from renters, residents and owners in your company’s voice, use your property management software the way your team does, and book real appointments in the right calendar. The late-rent reply in prompt one becomes a conversation Omnia has, with the payment plan recorded and the reminder set for the 25th.

The two work well together. ChatGPT is a good writing partner for the work you do yourself, and an agent takes the repetitive work off your desk entirely. What AI can take off a property manager’s desk in 2027 goes further into what agents can do today, and what should stay with your people.

Where to start

Pick the prompt closest to something you did this week and try it on your next real case. Keep the version that works in a shared document, so your team starts from a good prompt instead of a blank box. After a few weeks, you’ll have a small library that fits how your company writes.

General information, not legal advice. Rules change, so check the current rules where your homes are.

Quick answers

How can property managers use ChatGPT?

For writing and thinking work around the job, such as replying to a resident, turning notes into a rental listing, explaining a lease clause in plain words, summarizing a long owner email thread, comparing vendor quotes and drafting a monthly owner update. Give it the facts and the tone you want, and read every answer before it goes out.

Is it safe to put tenant information into ChatGPT?

Only what the task needs. Leave out Social Insurance or Social Security numbers, bank details, ID numbers and credit reports, and use a unit number or initials instead of a full name where you can. On a personal plan, turn off "Improve the model for everyone" in Settings, then Data controls, so your conversations aren't used for training. Business plans don't train on company content by default.

Can ChatGPT write a legal notice to a tenant?

It can help with the wording of a letter, but many notices have to be on an official form or include specific information to be valid, such as Ontario's Landlord and Tenant Board forms. Use the official form where one exists, and check the rules where the home is.

What's the difference between ChatGPT and an AI agent for property management?

ChatGPT writes what you ask for, and you copy it into your other tools and send it. An AI agent works inside your systems and finishes the job itself, such as answering the resident, booking the vendor and updating the record. Omnia by Tuze is an example built for property management, with AI teammates that answer every call, text and email.

About the author

Jazib Ahmad, Founder of Tuze

Jazib is an entrepreneur who loves technology, economics and solving hard problems. Before Tuze, he built safety-critical obstacle avoidance systems for autonomous indoor drones, earned his master's at the University of Toronto and published AI research with IEEE. He started Tuze to help small and midsized property managers grow to new heights. Among other things, Tuze gives property managers AI they can trust with real work, so they spend less time on busy work and more time with the people they look after. He believes in a future where housing, and the other things people need to live well, are within everyone's reach.

Questions about your own rentals?

Book a call with our team. We’ll talk it through with you, and if Tuze can help, we’ll show you how.

Book a call