The writing and reading parts of the job are now fast. A listing description from the MLS facts, a buyer-friendly summary of an inspection report, a five-touch follow-up for a lead source, a market update from the numbers you pulled this morning: each takes minutes. Showing the house, reading the room in a negotiation, and knowing which block floods stay yours.
The failure modes are specific to this business. AI-written ad copy drifts into language about who should live in a home, a fair housing problem. Property facts get invented: square footage, school district, the age of the roof. And the tools do not know your market, so any number that did not come from your MLS pull is a guess.
Your clients' data is protected too. Pre-approval letters, financial statements, IDs, and anything on an application are personal information under state law and your brokerage's policies. Redact before pasting, and use only tools your broker has approved.
Quick wins this week
- Paste the MLS facts (beds, baths, square footage from the tax record, year built, updates) and ask for a 150-word description of the property, never the buyer. Verify every fact.
- Upload an inspection report and ask for a summary grouped by safety, major, and minor items, each with the page number. Read the cited pages before you send anything.
- Give last month's numbers from your MLS (median price, days on market, months of inventory) and ask for a 200-word neighborhood update in your voice.
- Ask for a five-message follow-up sequence for one lead source (open house, online lead, expired listing), spaced over two weeks, no pressure language.
- Describe a hard conversation (a seller who wants to overprice, a buyer who wants to waive inspection) and role-play it before the real one.
What AI can do for real estate agents, task by task
Listing descriptions and ad copy
Paste the verified facts from the tax record and your walkthrough notes, plus a few photos, and ask for a description under your MLS character limit that describes the property, not the buyer. Check every fact against the record, then read it once for fair housing: no references to family, religion, national origin, disability, or who would love the home.
Inspection reports, HOA documents, and CC&Rs
Upload the PDF and ask for a table of findings by severity with page references, or a list of HOA rules that affect your client's plans (rentals, pets, parking). Open each cited page. The summary is for orientation; the client still reads the report, and legal questions go to an attorney.
Lead follow-up and nurture
Describe the lead source, what you know about the person, and your voice (paste two emails you actually sent). Ask for a sequence with subject lines and send timing. Load it into your CRM as templates, not as automation you never read again. Texts and calls still need TCPA consent.
Market updates and neighborhood content
Pull the numbers from your MLS yourself: closed sales, median price, days on market, inventory. Paste them with the period and ask for a newsletter section or a 60-second video script. Do not ask the model for the data; it does not have your market and will invent it.
Pricing conversations and CMA narratives
Paste the comps you selected (address, sale price, date, size, condition notes) and your suggested range, and ask for the narrative that connects them plus the three objections the seller will raise. The pricing judgment is yours; the model is writing up reasoning you already did.
Transaction deadlines and coordination
Paste the dates and contingency periods from the executed contract and ask for a deadline calendar with reminders, then compare it line by line to the contract. Zapier can push the dates to your calendar. Never let the model interpret contract language; that is your broker or an attorney.
Prompts for real estate agents
Replace the bracketed placeholders, paste into any chat assistant, and iterate on the result.
Fair-housing-safe listing description
You are a real estate copywriter. Write a listing description of no more than [CHARACTER LIMIT] characters for [PROPERTY ADDRESS OR TYPE] using only these verified facts: [BEDS, BATHS, SQUARE FOOTAGE, YEAR BUILT, LOT, UPDATES, NOTABLE FEATURES]. Location details: [NEIGHBORHOOD AND NEARBY AMENITIES]. Describe the property and location only. Do not describe or imply the ideal buyer, family status, religion, national origin, disability, or any protected class, and avoid phrases like 'perfect for' followed by a type of person. Do not add facts I did not provide. Give two versions: one warm, one concise.
Tip: Read the output for who it describes; that is the fair housing test.
Inspection report summary for a buyer
You are a buyer's agent helping a first-time buyer read an inspection report for [PROPERTY ADDRESS]. Using only the report I uploaded, produce three lists: safety items, major systems (roof, foundation, electrical, plumbing, HVAC), and minor or cosmetic items. For each item give the inspector's wording in brief and the page number. Do not estimate repair costs. Do not say whether the buyer should proceed. Finish with five questions for the inspector or a contractor.
Tip: Open every cited page before forwarding it.
Follow-up sequence for a lead source
You are a real estate agent writing follow-ups in my voice. Samples of my writing: [PASTE TWO EMAILS YOU SENT]. Lead source: [SOURCE, E.G. OPEN HOUSE VISITOR OR ONLINE INQUIRY]. What I know: [NOTES]. Write five messages over 14 days with a subject line and send day for each, under 90 words apiece. One message should offer something useful with no ask. No urgency tricks, no pricing predictions. Mark any message that needs consent to send as a text.
Tip: Edit until it sounds like you talking.
Neighborhood market update from my MLS numbers
You are a local real estate agent writing a monthly update for [NEIGHBORHOOD OR ZIP] readers. Use only these figures from my MLS for [MONTH]: closed sales [NUMBER], median sale price [PRICE], median days on market [DAYS], active inventory [NUMBER], comparison period [PRIOR FIGURES]. Write 200 words in plain language: what changed, what it means for a seller, what it means for a buyer. No predictions and no figures beyond those given. End by inviting readers to reply with a question.
Tip: If any number in the draft is not one you typed, delete it.
Prep for a pricing conversation
You are an experienced listing agent coaching me before a pricing meeting. Seller's expectation: [PRICE]. My comps: [PASTE COMPS WITH ADDRESS, SALE PRICE, DATE, SIZE, CONDITION]. My recommended range: [RANGE]. Write the narrative that connects the comps to my range in under 200 words, then list the three objections this seller is most likely to raise and a two-sentence response to each. Do not change my range. Flag any comp that weakens my case.
Tip: The flagged comps are the ones the seller will find on their own.
Want a prompt for something else? Use the Prompt Builder or browse customer communication prompts, sales and outreach prompts.
Skills to build
Fair housing review of everything AI writes
Why: Models learned from decades of ad copy that describes buyers, and that language is a violation whether a person or a machine wrote it.
How: Keep the protected classes and common trigger phrases next to your screen. Read every description and ad for who it describes.
Feeding the model facts instead of asking it for facts
Why: AI does not know your market, the tax record, or the school boundary. Anything it produces without your input is invented.
How: Pull the numbers first, then write. Every figure in an AI draft traces to your MLS, the county record, or your own notes.
Writing in your own voice
Why: Clients hire the person they met at the open house. Generic AI copy sounds like every other agent's, which is worse than sounding like you.
How: Paste two or three real emails into every prompt or a saved Project, and edit the output until it sounds like you talking.
Reading long documents with citations
Why: Inspection reports, HOA packets, and disclosures are long, and clients expect you to know what is in them.
How: Use NotebookLM or Claude for page-referenced summaries, then open every page it cites.
Systemizing follow-up
Why: Most lost deals are lost to silence. AI plus a CRM turns follow-up from willpower into a system.
How: Build one sequence per lead source, load them as CRM templates, and review replies daily. Automate delivery, never the judgment about what to say next.
Tools worth knowing
ChatGPT
The general-purpose AI assistant most of your coworkers already use.
Claude
A careful writing and analysis assistant that shines on long documents.
Gemini
Google's assistant, strongest when your work already lives in Google Workspace.
Canva
Design for non-designers, with Magic Studio AI for text, images, video, and resizing.
NotebookLM
A research notebook that only answers from the sources you give it, with citations.
Zapier
Connect thousands of apps with no-code workflows, now with AI steps, agents, and chatbots
Cautions for real estate agents
The Fair Housing Act and state laws prohibit advertising that indicates a preference based on race, color, religion, sex, familial status, national origin, or disability, and many states add more classes. AI drafts drift into 'perfect for a growing family' or 'walking distance to the church' without meaning to. Every description, ad, and targeting choice gets a fair housing read first.
Only a licensee can perform licensed activity, and only an attorney can give legal advice. Do not let AI negotiate for you, interpret contract terms for a client, or draft addenda. State rules also require brokerage identification in advertising; check that AI-generated ads include it.
Ask AI for partnership or lead-generation ideas and it will suggest arrangements with lenders, title companies, or inspectors that amount to referral fees or unearned marketing payments, which RESPA prohibits. Run any co-marketing or referral arrangement past your broker before acting on it.
Invented square footage, school assignments, permit history, or flood status in a listing is misrepresentation, and MLS rules and your E&O carrier treat it that way. Verify facts against public records and disclose virtual staging where your MLS requires it. Client financial documents and IDs are personal information; redact before pasting.
Your 30-day plan
- Week 1: Ask your broker which AI tools are approved and get the office's advertising and fair housing checklist. Save two of your own emails as voice samples.
- Week 1: Rewrite one active listing description with the fair-housing prompt and verify every fact against the tax record.
- Week 2: Summarize the last inspection report you handled with page citations and compare it with the notes you actually sent the buyer.
- Week 2: Build a five-touch sequence for your weakest lead source and load it into your CRM as templates.
- Week 3: Write next month's market update from your own MLS pull, turn it into a short video script, and record it.
- Week 4: Role-play your next pricing conversation, set up one Zapier flow from lead form to CRM, and save your prompts and voice samples in a Project or custom GPT.
Frequently asked questions
Will AI replace real estate agents?
Can I use ChatGPT to write listing descriptions?
Is it a fair housing violation if AI wrote the ad?
Can AI summarize an inspection report or HOA documents?
Which AI tool is best for real estate agents?
Terms used on this page
Related roles
- AI for Sales ProfessionalsSpend your hours in conversations instead of research tabs and CRM fields. Here is how sellers use AI for prep, follow-up, and proposals, and where the legal lines are on outreach and recording.
- AI for Insurance AgentsAI explains replacement cost versus actual cash value to a client in two paragraphs, drafts the rate-increase email you dread, and builds a quote comparison from three dec pages. The advice, the license, and the E&O exposure stay yours.
- AI for MarketersDraft campaigns faster, mine customer feedback for messaging, and turn reporting into a ten-minute job, without drifting off-brand or into a CAN-SPAM problem.
- AI for Small Business OwnersAI answers the one-star review calmly, drafts the month of posts, turns your voice memo into an SOP, and preps the questions for your CPA. You still make the calls, sign the checks, and own what goes out under your name.