AI for Realtors: What Agents Actually Use It For (and the Rules)

AI for realtors mostly means using generative AI tools like ChatGPT to draft the writing that eats up an agent’s week: listing descriptions, social posts, and client emails. It can save real time, but you are still responsible for every word and image it produces, including under fair housing law and your state’s and MLS’s advertising rules.

Below is what agents are actually doing with AI as of November 2026, a simple workflow to start with, and the compliance points that trip people up.

How agents are using AI right now

The best current snapshot is NAR’s 2026 REALTORS® Technology Report, released September 22, 2026. According to NAR’s release and coverage from Inman and RISMedia:

• 23% of agents use AI daily, and another 25% use it weekly.
• Among agents who use AI, 55% say it has had a positive impact on their business.
• 81% say saving time is their main reason for adopting new technology (up from 66% a year earlier), and 71% cite a better client experience (up from 64%).

The most common uses among agents who use AI:

So, despite all the talk about AI “agents” and automated lead machines, the everyday reality is writing help. NAR deputy chief economist Jessica Lautz framed it this way: when routine work gets handled faster, agents have more time for guidance, negotiation and advice.

Be careful with adoption numbers you see elsewhere. Several vendor blogs quote much higher figures (one claims 82% daily use), and they do not match NAR’s published report. Stick to the primary source.

Where AI helps most

Listing descriptions

This is the obvious starting point. Give the tool the facts (beds, baths, square footage, upgrades, lot details, neighborhood features you can verify) and ask for a draft in your voice. Then edit. AI tools will happily invent features or exaggerate (“chef’s kitchen” for a basic galley), and a misstatement in a listing is your problem, not the software’s.

Social posts and email

AI is good at turning one piece of content into several: a listing description into an Instagram caption, a market update into a newsletter paragraph, a showing recap into a follow-up email. Ask for a few variations and pick the one that sounds like you.

Consumer search is moving into chatbots

It is also worth knowing where buyers are. In October 2025 Zillow launched an app inside ChatGPT, described as the first and only real estate app there at launch. Consumers can search listings, explore neighborhoods and affordability, and connect with an agent from within the chat. Zillow said it planned to add new construction listings and 3D tours over time. If you work with Zillow’s agent programs, that is one more place leads may come from; how much it matters in your market is not something the public sources quantify.

A simple workflow to get started

1. Check your brokerage’s AI policy first. Find out which tools are approved and what client data, if any, you may put into them.
2. Pick one task. Listing descriptions are the easiest win because the inputs are factual and the output is short.
3. Build a fact sheet, not a vague prompt. List the verified features, your target tone, length limits for your MLS, and words you never use.
4. Generate two or three drafts. Ask the tool to stick to the facts provided and not to add features.
5. Edit line by line. Verify every claim against the listing data and your own walkthrough. Cut anything you can’t back up.
6. Run a fair housing pass. Remove language about who the home suits (families, singles, specific groups) or neighborhood descriptions that hint at protected characteristics. Describe the property, not the buyer.
7. Save the prompt that worked. Reuse it, so each listing gets faster without getting sloppier.

Once that feels routine, extend the same approach to social posts and follow-up emails.

The compliance points you cannot skip

Fair housing applies to AI output and ad targeting

In 2024 HUD issued guidance explaining how the Fair Housing Act applies to algorithm-driven and AI-assisted advertising, warning that targeted ads could unlawfully exclude protected classes even when the discrimination is unintended. In 2025 HUD withdrew that and several other fair housing guidance documents as part of a deregulatory review, as NAHRO reported. Sources give slightly different dates for the withdrawal notices, so check the Federal Register if you need the exact timeline.

The key point: withdrawing guidance does not change the law. The Fair Housing Act still applies, and so do state fair housing laws and the NAR Code of Ethics. If you run ads with automated audience targeting, review who the ads actually reach.

AI-edited and virtually staged photos

This is where the rules have changed fastest. California’s AB 723 took effect January 1, 2026. Per Inman’s explainer and the San Diego MLS compliance page, it requires:

• A conspicuous label on digitally altered listing images, whether edited with photo software or AI.
• Access to the unaltered original images.

SDMLS and CRMLS have adopted matching MLS rules, and SDMLS notes that brokers remain responsible for compliance in advertising outside the MLS too. Some sources describe criminal or license penalties for willful violations; we could not confirm those against the statute text, so read the official bill if this affects you.

Outside California, there is no single national rule. MLSs set their own policies: some require watermarks, some a caption or remarks, and some want the original next to the staged image. NorthstarMLS, for example, publishes its own guidelines for virtual staging and AI-enhanced photos. The safe practice everywhere is the same: label every altered image, keep the originals, and never use AI to hide defects or add features that don’t exist.

Client data and accuracy

General chatbots are not designed for confidential transaction details. Keep names, financial information and contract terms out unless your brokerage has approved a tool for that purpose. And remember that AI can state wrong facts confidently. Anything about schools, zoning, taxes or square footage needs to come from an authoritative source, not a chatbot.

The bottom line

AI is most useful to realtors as a drafting assistant for listings, posts and emails, which is exactly how most agents who use it already work. Start with one task, keep a human edit on everything, and know your local rules on photo disclosure and fair housing. Tools and rules are both changing quickly, so treat the specifics here as accurate as of November 2026 and check your state commission and MLS for updates.