AI in Property Marketing: Listings, Virtual Staging and the New Disclosure Rules

In the first three months of 2026, nearly 11% of the main listing photos on four leading US property portals had been digitally altered, and more than 90% of those altered images carried no visible disclosure. Buyers still pick homes from photos before anything else, so AI in property marketing is now as much a question of trust as of speed.

What the data shows

• 43% of US buyers took looking online for properties as their very first step, and 51% found the home they eventually bought on the internet (NAR, 2025 Home Buyers and Sellers Generational Trends).
• 83% of buyers who searched online rated photos "very useful", ahead of detailed property information (79%), floor plans (57%) and virtual tours (41%) (same NAR report).
• 46% of Realtors use AI generated content such as listing descriptions, 20% use AI daily and 32% have not actively tried it for their business (NAR 2025 Technology Survey).
• Nearly 11% of roughly 40,000 primary listing photos on Zillow, Redfin, Realtor.com and Homes.com showed signs of alteration, and over 90% of those had no visible disclosure (Coraly study, reported by Real Estate News).
• Real estate firms plan to spend less than 1% of revenues on AI in 2026, against about 2% for tech companies and financial institutions (BCG AI Radar 2026).

Buyers shop with their eyes first, so listing photos carry the most weight

The National Association of Realtors' 2025 Generational Trends report is blunt about where the journey starts. In every age group the most common first step was looking online, and buyers found the home they purchased on the internet more often than through an agent (51% against 29%). 69% searched on a mobile or tablet device.

Inside that search, pictures dominate: 83% of internet searchers called photos very useful, rising to 86% among buyers aged 35 to 44.

Sellers have noticed. Zillow's December 2025 report found 78% of sellers were more likely to hire agents who offer high resolution photography, and 75% were more likely to hire agents who provide virtual tours and interactive floor plans. NAR's 2025 Profile of Home Staging, published in May 2025, found sellers' agents said photos (88%), videos (47%) and traditional physical staging (43%) in listings were more important to their clients. 83% of buyers' agents said staging makes it easier for a buyer to picture the property as their future home, and 29% reported that staging lifted the dollar value offered by 1% to 10%.

That is why AI arrived in property marketing through the image first: digital staging is the cheapest way to help a buyer imagine a life in the home. Agents remain cautious, though. 34% of sellers' agents considered virtual staging less important to their clients, while 24% felt it was equally important.

Virtual staging moved from agents' tools into the portals themselves

The biggest change of the last year is who controls the edit. On 10 September 2025, Zillow added Virtual Staging to its premium Showcase listings. Shoppers tap an icon on selected room photos, choose one of seven styles (modern, Scandinavian, industrial, midcentury, luxury, coastal or farmhouse) and swipe between the staged and the original photo. The feature uses technology from Virtual Staging AI, a company Zillow bought in October 2024. Zillow says Showcase listings sell for about $7,000 more than comparable listings, a figure from its own analysis.

In the UK, Rightmove announced two AI tools on 7 November 2025. "AI Keywords" lets people search for features such as "exposed brick", "river view" or "underfloor heating", scanning both images and text in listings, trained on 25 years of Rightmove data. "Style with AI", an experiment with a set group of users, lets buyers remove furniture, adjust lighting and restyle a room, for example to a Scandi or Art Deco look. Rightmove said it had 27 AI initiatives in development.

In both cases the buyer asks for the restyle and the real photo stays one swipe away. That is very different from an agent quietly uploading a staged image as the main photo.

AI now writes the listing, but most agents are still at the start

Text is where agents use AI most. NAR's 2025 Technology Survey, fielded in July 2025 with 1,241 usable responses, found 46% of Realtors use AI generated content such as listing descriptions, 38% use virtual tours and 7% use chatbots for lead capture or client communication. ChatGPT was the most used tool (58%), followed by Google's Gemini (20%) and Microsoft Copilot (15%).

The results are modest. 46% of Realtors saw a neutral or no noticeable impact from AI on their business, 33% saw a moderately positive impact and 17% a significantly positive one. BCG's AI Radar 2026, a survey of nearly 2,400 executives across 16 markets published in January 2026, found real estate firms plan to spend less than 1% of revenues on AI this year, roughly half of what tech and financial companies plan.

The gap is not surprising. A chat window produces a readable description in seconds, but it does not check that the floor area matches the title deed or that the view described still exists. Teams that get value treat AI drafting as one step in a process, with a named person who checks facts before anything goes live.

Dubai portals are using AI to check listings, not only to polish them

The most instructive examples in the Gulf are about accuracy. In February 2026, AGBI reported that Property Finder now pulls key details such as the permit, floor area and number of bedrooms automatically from government records when an agent uploads a listing. If the details do not match official data, the listing is blocked. Attempts to publish non compliant listings spiked at first, then fell; by the end of 2025 the company said unsuccessful attempts had dropped more than 90%. Property Finder also said agent numbers on its platform have grown roughly 30% a year since 2022 while engagement per listing has fallen, which means more agents competing for the same attention.

Bayut has taken the search side. On 27 August 2026 it launched AI Search, a conversational tool that lets people describe lifestyle, location, budget and features in everyday language instead of clicking through filters. It builds on TruEstimate, BayutGPT and Bayut's property search inside ChatGPT.

The audience is a strong reason for Dubai teams to take multilingual content seriously. Dubai's real estate market recorded AED 917 billion in transactions in 2025, with around 193,100 investors, of whom 129,600 were new, according to figures published by the Government of Dubai's Public Debt Management Office in January 2026. Those buyers arrive from many countries and read in many languages. AI makes a first translation of a listing or a brochure almost free, but a translated floor area, service charge or payment plan that is wrong in one language is still wrong. A native speaker on the team should sign off each language version before it runs.

Buyers are ready for it. Stanford's 2026 AI Index, published in April 2026, puts the UAE's generative AI adoption rate at 64%, well above what its income level would predict, while the United States ranks 24th at 28.3%.

Disclosure and ad targeting rules are catching up

California's Assembly Bill 723 was approved on 10 October 2025 and took effect in January 2026. Any broker or salesperson who uses a digitally altered image in property advertising must say so in a reasonably conspicuous statement on or next to the image, with a link or QR code to the original. The law covers images changed with photo editing software or artificial intelligence to add, remove or change furniture, flooring, paint colour, landscaping, views through windows, utility poles and neighbouring properties. Lighting, white balance, cropping and exposure adjustments do not count.

The Coraly study, reported by Real Estate News in June 2026, shows why the rule exists. 69% of altered images involved sky replacement, and Redfin had the highest rate of virtually staged photos at 17.7%. Kitchens and bathrooms were altered in less than 1% of cases. Real Estate News also reported that frustrated buyers have started calling misleading listings "housefished", that New York's Department of State issued a consumer alert on AI images in late 2025, and that NAR's Code of Ethics already requires Realtors to "present a true picture" in their advertising.

Targeting has its own history. In the United States, a court approved a settlement between the Department of Justice and Meta on 27 June 2022 over housing ads. Meta stopped using its "Special Ad Audience" tool for housing ads, built a new Variance Reduction System to reduce disparities in who sees them, and paid a civil penalty of $115,054. AI can find likely buyers, but the rules on who is shown a property advert apply to the software too.

For a brokerage, the practical rules are short. Keep every original photo. Label any staged or altered image in the image itself. Check every AI drafted number against the permit and the title. Have a person approve each language version.

Questions people ask

Do buyers really choose homes from listing photos?

NAR's 2025 data says photos are the website feature buyers value most, rated very useful by 83% of online searchers. More than half of buyers found the home they bought online. The listing photo is usually the first viewing, so its accuracy decides whether the second viewing happens.

Should agents disclose virtually staged or AI edited photos?

In California it is now the law for any image where furniture, finishes, landscaping or views have been changed. Elsewhere it is the safer practice, because the Coraly study found over 90% of altered photos carried no disclosure and buyers have noticed. Zillow and Rightmove both keep the original photo one swipe away, which is a good standard to copy.

Is AI good enough to write listing descriptions?

It is good at a fast first draft, which is why 46% of Realtors already use AI generated content. It is not good at knowing whether a detail is true. Property Finder's permit checks show the accurate path: pull the facts from official records, then let AI help with the wording.

Where should a Dubai brokerage start with AI marketing?

Start with the tasks your team repeats every week, such as descriptions, translations and follow up messages, and decide who checks each one before it goes out.

CroVIN runs hands on AI training for property marketing and sales teams in Dubai, so your people use these tools quickly and accurately with the right checks in place.