AI agents for real estate: faster lead response, less admin, and the fair housing line

In real estate, an AI agent earns its keep in two places: replying to every new lead within minutes, at any hour, and taking over the administration that surrounds each deal. Pricing advice, negotiation and anything about the character of a neighborhood stay with a licensed person. Fair housing law, consent rules for calls and texts, and your MLS's data rules all apply to what an AI agent says and does on your behalf.

A note on wording: this guide is about AI agents that work for real estate professionals. It is not about software replacing the agent.

#Where an AI agent helps most

#Lead response and nurture

• Instant replies to new leads, from portals, your website, sign calls and social media. The lead gets a useful response in minutes, not the next morning. In a business where the first agent to respond often wins the client, this is usually the most valuable single job.
• Qualifying. Timing, price range, area, whether they are pre-approved, whether they have a home to sell. Asked conversationally and recorded in your CRM.
• Booking showings and calls. The agent offers times from your calendar, confirms, reminds and reschedules.
• Long-term nurture. Most leads are months from moving. The agent keeps in touch with something relevant, such as new listings that match what they asked for, and tells you when someone becomes active again.
• Open house follow-up. Same day, to every visitor, with the questions they asked answered.

#Listings and marketing

• Listing descriptions, drafted from the property facts and your notes, in your voice, then checked against fair housing wording before you approve them.
• Social posts, email announcements and flyers derived from the listing.
• Market updates for your database, drafted from the figures you supply.

#Transaction administration

• Deadline tracking. Inspection, appraisal, financing and closing dates from the contract, with reminders to every party.
• Document chasing. Who still owes a signature, a disclosure or a proof of funds.
• Status updates to clients, so nobody has to call to ask what happens next.
• CRM upkeep. Notes from calls and showings written into the right record.

#Property management

• Rental inquiries, answered at once with availability, criteria and viewing times.
• Pre-screening against written criteria, applied the same way to every applicant.
• Maintenance requests, sorted by urgency, with the details a contractor needs collected up front.

#What stays with a licensed person

• Pricing opinions and offer strategy.
• Negotiation.
• Anything that amounts to legal, tax or financing advice.
• Describing neighborhoods, schools and safety. Point people to official third-party sources instead.
• Decisions on rental applications. The agent can collect and organize. A person decides, using your written criteria.
• Difficult conversations. A sale falling through is not a job for software.

#The rules that apply

Real estate carries obligations that most AI tools know nothing about. An AI agent acting for you is covered by the same rules you are.

• Fair housing. In the US, the Fair Housing Act prohibits discrimination in the sale, rental and advertising of housing. For an AI agent, the main risks are steering, which means suggesting areas based on who lives there, and discriminatory wording in listings and replies. Regulators have made clear that the law applies to automated tools used for advertising and tenant screening. Build the rules in: describe property, not people; refuse questions about the make-up of a neighborhood and point to public sources; apply the same criteria to everyone.
• Consent for calls and texts. The TCPA requires the right consent for automated calls and texts, covers AI-generated voices, and sits alongside Do Not Call rules. A form fill is not automatically consent to automated texts.
• Licensing and disclosure. The agent should say it is an AI assistant working for you, give your brokerage details where your state's advertising rules require them, and never present itself as a licensed person.
• MLS and listing data. Your MLS's rules govern how listing data can be used and displayed. Check them before feeding listing data into any tool.
• Wire fraud. An AI agent should never send, confirm or discuss wiring instructions. Say so in its instructions, and tell clients the same.
• Personal and financial information. Pre-approval letters, identity documents and bank details belong in your transaction system, not in a chat log.

This is general information and not legal advice. Your broker, your state regulator and your attorney are the authorities on what applies to you.

#CRM assistant, lead tool or custom agent?

For a solo agent, the built-in tools are nearly always enough. Custom work starts to make sense at team and brokerage level, where a rule such as "portal leads over a certain price in these ZIP codes go to these three agents in rotation, and get a call within five minutes" is how the business actually runs.

#A sensible first project

Lead response and showing booking. It is measurable within two weeks, the risk is low because the lead asked to hear from you, and it removes the most common way to lose business, which is replying too late.

Set it up in draft mode for the first week, where you approve each reply. Collect the replies you changed, and turn them into rules. Then let it send on its own for first responses and scheduling, and keep everything else with you.

#What to measure

• Time to first response, by lead source.
• Lead-to-appointment rate.
• Appointments held, not only booked.
• Replies from long-term nurture, and how many became active clients.
• Missed deadlines. This should be zero, and the agent's reminders are there to keep it there.
• Hours of administration saved per transaction. Ask your transaction coordinator.

#What it costs

• CRM assistants and lead tools typically cost tens to a few hundred dollars per user per month.
• Listing content tools cost less, and the AI assistant in your office suite may already be enough.
• A custom agent for a team or brokerage typically costs $8,000 to $40,000 to build, depending on how many lead sources and systems are involved. A single automation, such as lead routing with an instant reply, costs less. Running costs are usually small. See how much does an AI agent cost.

For reference, our own automation projects start from $4,800 and take about two weeks, and custom AI agents start from $9,800 and take three to five weeks. Each is quoted as one fixed price before work begins, with a 30-day fix window.

#Where to start

Check how long your last twenty leads waited for a first reply, and how many of them you ever spoke to. If the wait is measured in hours, fix that first, and your CRM's own assistant may be all you need. If you run a team with routing rules and several lead sources, book a free 20-minute call and we will tell you what a build would involve. The general principles are covered in AI agents for sales.

This guide is part of our series on AI agents for business.

Common questions.

• How are real estate agents using AI agents?

  Mostly for speed and admin. An AI agent replies to portal and website leads within minutes at any hour, asks about timing, budget and area, books the showing, and keeps following up with people who are months away from moving. It drafts listing descriptions and social posts, tracks contract deadlines, chases documents and keeps the CRM tidy.

• Will AI replace real estate agents?

  It is unlikely to replace the parts clients pay for: pricing judgment, negotiation, local knowledge and steadying people through a stressful purchase. It does replace a lot of the work around those things, such as first replies, scheduling, reminders and paperwork tracking. Agents who hand that over can serve more clients properly, which is where the advantage lies.

• Is it legal to use AI to call or text real estate leads?

  In the US, automated calls and texts require the right consent under the TCPA, AI-generated voices are covered by the same rules, and Do Not Call lists still apply. A lead who filled in your form has not necessarily agreed to automated texts. Get consent wording reviewed, keep records of it, and check your state's rules. This is general information, not legal advice.

• Can AI write listing descriptions without breaking fair housing rules?

  Yes, with a check in place. Describe the property, not the ideal buyer or the people nearby. Avoid phrases that signal a preference for or against any protected group, such as references to families, age, religion or ethnicity. Have the agent screen its own drafts against a list of risky phrases, and have a licensed person approve every listing before it goes out.

• Which AI agent is best for real estate?

  Start with the assistant in the real estate CRM you already use, since it already holds your leads and your calendar. Add a dedicated lead-response tool if speed to reply is your main problem. A custom agent makes sense for teams and brokerages with their own lead routing rules, several lead sources, or a property management system that off-the-shelf tools do not connect to.