The short answer
For a real estate agent, AI is most useful for writing listings, replying fast to showing requests, following up with buyers and sellers, and keeping social media going. It does not price homes for you. Three rules frame its use: written buyer agreements before tours since August 17, 2024, fair housing language in every ad, and prior consent for AI-voice calls under the TCPA.
We make Extra, an AI agent platform, so this article is written by an interested party. Facts about other tools and about the law come from the sources listed at the bottom, as of the date shown.
What does AI actually do for a real estate agent?
For an independent agent or a small brokerage, AI is most useful for four jobs: writing listing descriptions, answering showing requests fast, following up with buyers and sellers, and keeping social media going. It does not price the home for you, replace the showing or know the neighborhood the way you do.
The payoff is speed. A buyer who sends an inquiry from a listing site on Sunday night often contacts several agents; the one who replies first with showing times has the edge. An AI agent that drafts that reply while you are at a closing changes the math.
Which real estate tasks should you hand to AI?
| Task | What AI does | What you check |
|---|---|---|
| Listing description | Writes it from your walkthrough notes and the property facts | Square footage, price, required disclosures, MLS rules |
| Replies to online inquiries | Answers, asks two qualifying questions, offers showing times | Real openings in your calendar |
| Buyer follow-ups | Drafts a note when a new listing fits their search | That the person agreed to hear from you |
| Seller updates | Summarizes showings and feedback each week | Your tone and your pricing advice |
| Social posts | Writes “just listed,” “just sold” and tip posts | The seller’s OK on photos |
| Buyer agreements and paperwork | Drafts plain-English explanations for clients | The actual form your brokerage and state require |
What changed with buyer agreements in 2024?
Since August 17, 2024, under the NAR settlement practice changes, MLS participants working with a buyer must have a written agreement before touring a home, and offers of compensation can no longer be published on the MLS. This matters for AI in two ways. Your inquiry replies should explain, in plain words, that a written agreement comes before the first showing. And any AI-drafted text about compensation must follow your brokerage’s approved language, not something a chatbot improvised.
Good practice: keep a short, approved paragraph explaining buyer agreements and have the AI paste it word for word rather than paraphrase it.
Can AI write listings without fair housing risk?
Yes, if you review every word. The Fair Housing Act (42 U.S.C. 3604(c)) makes it illegal to make, print or publish any notice, statement or advertisement for the sale or rental of a dwelling that indicates a preference, limitation or discrimination based on race, color, religion, sex, handicap, familial status or national origin. AI models trained on old listings can reproduce phrases like “perfect for a young couple” or “ideal for singles,” which describe people instead of the property.
- Describe the home, not the buyer: “three bedrooms near two parks,” not “great for families.”
- Avoid neighborhood descriptions that hint at who lives there.
- Check ad targeting too: an audience built on protected traits is a problem even if the copy is clean.
Can you prospect sellers by phone with AI?
Carefully. In February 2024 the FCC ruled that AI-generated voices count as an “artificial or prerecorded voice” under the TCPA, so calls that use them need the called person’s prior express consent unless an exemption applies. An AI voice cold-calling homeowners from a purchased list is out. Live calls still have to respect the National Do Not Call Registry and your state’s rules.
What works without that risk: home value requests from your website, mailers in your farm area, referrals, open house sign-ins with consent, and social media. For email outreach rules, see our post on cold email rules.
How do you put AI to work in a brokerage, step by step?
- Create a fixed property fact sheet (beds, baths, square footage, taxes, HOA) that the AI may never change.
- Write your inquiry rules: response time, qualifying questions (budget, pre-approval, timeline), showing windows.
- Connect the agent to your email and calendar, not to your transaction software unless needed.
- Review every listing before it goes live for the first month.
- Measure response time to inquiries before and after: that is the number that matters.
What does AI get wrong for agents?
- Pricing: a model does not see condition, light or a noisy street. It gives a range, not a list price.
- Puffery that becomes a claim: “guaranteed appreciation” or “zero maintenance” creates liability.
- Edited photos: removing a power line or virtually repainting a facade without saying so misleads buyers.
- Stale facts: school assignments, zoning and taxes change; verify them at the source.
How does Extra help a real estate agent?
Extra (hello-extra.com), the AI agent platform for small businesses, covers part of this work. Pam answers buyer emails and offers times in your Google Calendar; Jade writes listing descriptions and posts. Jim, the prospecting agent, is built for business-to-business outreach: he writes from your own email address and LinkedIn profile, with follow-ups and weekly numbers. For a commercial broker reaching companies about office or retail space, that is a real fit; for cold-calling homeowners, it is not.
The plan is $99 per month for all eight agents (see pricing). Extra does not replace your transaction platform or MLS. See also AI sales prospecting for small businesses.
Which message templates should you use?
Two messages matter more than the others: the reply to a showing request and the weekly seller update.
Inquiry reply: “Hi, thanks for your interest in the three-bedroom on Maple Avenue. It’s still available. I can show it Thursday at 6 p.m. or Saturday at 11 a.m. Before we tour, we’ll sign a short written buyer agreement; I’ll send it ahead so you can read it. Are you pre-approved, and what’s your timeline? The AI assistant for Rivera Realty.”
Weekly seller update: “This week: 4 showings, 2 positive comments on the light, 2 concerns about price given the kitchen. A second showing is set for Saturday. Let’s talk Tuesday about pricing strategy.”
- Always two specific times, not “when are you free?”
- Two qualifying questions at most in the first message.
- The update sticks to facts; the pricing recommendation stays yours.
Checklist before you let AI write for you
- A fixed fact sheet per property the AI may not change.
- Approved language for buyer agreements and compensation.
- A fair housing review of every listing and ad audience.
- Consent records for any automated calls or texts.
- Labels on virtually staged or edited photos.
- A response time target for inquiries, measured weekly.
Frequently asked questions
Can AI price a home?
It can suggest a range from the comps you give it. The pricing opinion is yours, after you have seen the property.
Do you have to disclose that AI wrote a listing?
There is no general federal rule for text you reviewed and approved, but check your MLS and state rules. Photos that were digitally altered or virtually staged should be labeled.
Where should you start?
With replies to showing requests: that is where speed pays most. For the bigger picture, read what AI agents do for a small business.
