Faster, Cheaper, Riskier? What Brokers Should Know Before Putting AI to Work

October 23, 2026

d3sign / Getty Images

AI has moved from curiosity to an everyday tool in real estate faster than most of us ever expected. Brokers and agents are using it to write listing descriptions, draft emails, build websites, schedule social posts, automate follow-up and even produce video, often for a fraction of what a traditional tech stack costs.

For many brokers, especially the tech-forward ones, that is an easy trade: fewer subscriptions, faster output and more time with clients.

But the qualities that make AI so appealing—speed, low cost and the ability to handle tasks that once required multiple tools—can also introduce risks. The key is to capture the benefits while protecting your clients, your brokerage and your license.

The Upside Is Real With AI

First, let's look at where AI can make a real difference. Here are the key benefits for brokers:

Time savings: Drafting blog posts, captions, listing descriptions, emails and follow-up messages is the kind of repeatable work AI does well.
Lower cost and fewer logins: Consolidating a patchwork of subscriptions into one assistant simplifies the stack and the budget.
Consistency: A weekly blog and a steady social calendar are easier to keep up when a first draft takes minutes instead of hours.
Reach: Tools that translate content or turn one piece of content into many formats help small teams compete with larger ones.
Leverage for small brokerages: Capabilities that once required a marketing department are now within reach of a solo broker.

With all those potential business benefits in mind, the question isn’t whether to use AI but how to use it so that speed does not outrun judgment.

Where the Risks Live When Using AI

AI’s speed and convenience can come with trade-offs, like errors, privacy concerns and compliance missteps that can create real risks for brokers, their clients and their businesses. Here’s what to consider.

1. Client data and confidentiality

The most important rule is also the simplest: be deliberate about what you type into an AI tool.

Not every plan treats your data the same way. Anthropic, for example, says that for its consumer plans (Free, Pro and Max), chats may be used to improve its models if the user allows it, while its commercial products, such as Claude for Work and the API, do not use data for training by default. Other providers have their own terms. Whatever tool you use, check its data policies and privacy settings before entering client or business information. The plan you choose and the settings you enable can determine how the information Is handled.

NAR's AI Policy Template for Brokers is direct on this. It says only approved AI tools should be used for tasks involving client, transaction or brokerage data, that data entered should be "the minimum necessary to accomplish the task," and that "if a tool has not been approved, assume it is off-limits for client or brokerage data." It also flags that consumer-grade tools should not be used for work involving personal information, client names, financial details, transaction documents or MLS data.

Treat this as a bright line: no social security numbers, no financial records, no settlement statements, no access codes and no confidential transaction documents in a tool you have not vetted. MLS data deserves the same care. The same guidance warns that MLS data must not be scraped, redistributed or repurposed in violation of copyright, MLS rules or licensing agreements.

2. Fair housing

AI is very good at persuasive copy, and persuasive copy drifts toward describing the ideal buyer which can raise fair housing issues. Phrases like "perfect for young families," "quiet, mature community" or references to nearby houses of worship can imply preference or steer buyers. The NAR policy kit lists these as examples of prohibited AI-generated content.

"The software wrote it" is no excuse either. When using AI in real estate advertising, focus on the property’s features, not the people you think should live there. The same principle applies to automated ad targeting and lead scoring. HUD's guidance makes clear the Fair Housing Act applies to housing advertising and tenant screening when AI and algorithms are used for these purposes.

3. Accuracy: AI sounds confident even when it is wrong

AI can invent amenities, overstate features or get details wrong, and it will do so in a polished voice. Under Articles 2 and 12 of the Code of Ethics, REALTORS®—members of the National Association of REALTORS®—must avoid exaggeration and misrepresentation and present a true picture in advertising and marketing. NAR guidance warns that AI output may be inaccurate but still must meet those duties.

State regulators agree. The California Department of Real Estate's licensee advisory says responsibility for inaccurate or misleading AI-generated advertising rests with the licensee and their responsible broker, not the technology provider. It also reminds licensees that required advertising disclosures, including license and broker identification, still apply to AI-generated content.

Always verify the accuracy of any AI output: Don't skip this step with a weekly blog or an automated social calendar. Volume makes mistakes easier to miss.

4. Altered images and AI-generated video

If your AI workflow touches listing photos, or turns still photos into video, disclosure rules apply and they are getting stricter. And recently, state laws addressing deceptive advertising and consumer protection may also apply to AI-altered images and videos.

California requires a statement that an image has been digitally altered, plus a link or QR code to the original, under AB 723 (Bus. & Prof. Code section 10140.8), effective Jan. 1, 2026.
Wisconsin's Act 69 takes effect Jan. 1, 2027, with New York and New Jersey considering similar measures.
MLS rules vary. NorthstarMLS, for example, requires AI-enhanced images to be identified and paired with an unaltered "before" image, and it prohibits changes to permanent features.
AI-generated video that simulates a drone flyover or walkthrough from still photos should say so clearly.

The safest habit is to assume that any materially altered or AI-generated media needs a label, keep the originals, and check your MLS and state rules before you publish.

5. Avatars, synthetic media and trust

Avatar video is one of the most tempting features, and it deserves special consideration as well. Real estate is a relationship business, and trust is the product.

A few guardrails to consider:

Never use AI-generated testimonials or fake reviews. The FTC's consumer reviews rule bans testimonials from people who do not exist or who have no real experience.
Do not create a voice or likeness of anyone without consent. The NAR policy kit calls for written approval before generating synthetic voices, images or likenesses of agents, staff or clients.
Disclose when it matters. Few laws specifically require disclosing an avatar of yourself, but advertising rules still apply, and some states require disclosure when a bot interacts with consumers in the sale or advertising of real estate. One clear line, such as, "This video was produced using an AI avatar," costs nothing and protects your credibility.
Keep your real face and voice for the moments that matter, such as consultations, negotiations and anything sensitive.

6. Automations and agentic AI

Automations are where small mistakes scale quickly. An automated message to the wrong person, an autoreply that sounds like advice, or an AI that answers questions about price or value can create real exposure.

NAR's policy kit states that "fully autonomous execution without human review is prohibited," and that AI may not carry out activities requiring a real estate license or create obligations on a brokerage's behalf. For valuations and CMAs, it says agents must use independent professional judgment and verify comps and market data rather than rely solely on an AI-generated number.

7. Accountability always lands on you

This is the thread running through every point above. The policy kit puts it plainly: agents and staff are "fully accountable for any AI-generated content they use, share or publish." Managing brokers have supervisory responsibilities, too. If your brokerage doesn’t yet have an AI policy, creating one is a good place to start.

A Practical Checklist

Before you replace a tool with AI, or let AI loose on a workflow, ask yourself:

Is this tool approved for the data I am about to enter? If not, leave out client names and other personal information, financials and transaction documents.
Have I checked my plan's privacy and training settings?
Has a human reviewed this before it goes out? Listings, ads, emails, CMAs and anything with a client's name on it.
Does the copy describe the property, not the people?
Is every fact verified against the MLS record or source documents?
Is anything altered or AI-generated, and is it labeled? Keep the originals.
Does my brokerage’s name, license number and required disclosures still appear?
Could this automation say or do something I would not approve of in person?
Does my managing broker know what I am using?

Learn More About Putting AI to Work Responsibly

AI can help brokers save time, streamline everyday tasks and stay competitive. But those benefits come with responsibilities. These tools do not carry your license, your Code of Ethics obligations or your clients' trust. You do.

Use AI to work more efficiently but make sure your judgment—not the technology—guides the decisions that matter.

Want to go deeper on this topic? Join us at NAR NXT, Nov. 6–8, 2026, in New Orleans, for our session on the responsible use of AI: Protecting Clients AND Yourself in the Age of AI

This article is for general information and is not legal advice. Advertising, disclosure and privacy rules vary by state and MLS, so confirm requirements with your managing broker, your state licensing authority, your MLS and legal counsel.