What should a seller or listing agent include on an AI staging disclosure checklist before publishing a virtual-staged real estate photo?

Direct answer

An AI staging disclosure checklist is a short approval record for every property image that has been changed by generative or image-editing software. It should state that the image is virtually staged, identify the main edits, confirm that the listing description and other photos preserve the actual condition, and record who reviewed the result. The checklist is not a substitute for a local disclosure rule, MLS rule, contract term, or professional code, so the listing agent should verify those requirements before release. The standard wording for AI staging is: “This image has been virtually staged with artificial intelligence. Furniture, décor, and minor cosmetic improvements are digital additions and do not represent items included in the sale.” That sentence is honest, readable, and normally sufficient for ordinary room staging, although a stricter local rule may require different wording. The key is to disclose the medium and separate digital additions from physical fixtures.

Also worth reading: Do AI home staging tools meet ethics disclosure requirements? · What is the definitive post-quantum cryptography migration checklist for enterprise architectures? · Is AI Virtual Home Staging Helping Sellers or Cursing Renters With Impossible Homes?

Why disclosure exists

The need for disclosure comes from the difference between presentation and misrepresentation. A sofa, rug, lamp, artwork, or plant added to an empty room can help a buyer understand scale and possible use, but it can also suggest that a feature already exists. A wall color, flooring repair, garden, view, or outdoor furnishing may change the impression of condition, maintenance, or included property. Buyers rely on listing photographs when deciding whether to tour a home, so an AI-generated feature should not be allowed to look like a photograph of the physical listing. This is why the disclosure should be visible near the image and supported by an unedited reference or written note when the edit is material. It is not enough to say “AI used” if the buyer still cannot tell what changed.

Core checklist

The checklist should be completed once per image, not once per campaign, because a hallway, kitchen, and backyard can require different notes. The first field should identify the property address or listing ID, image filename, date, creator, and software used. The next field should classify the edit as virtual staging, cosmetic retouching, repair simulation, floor-plan generation, rendering, or a combination. The editor should then list every material addition or removal, including furniture, appliances, paint, flooring, landscaping, lighting, sky, views, and structural elements. A simple rule is that anything a buyer could mistake for a permanent feature, included item, or present condition belongs in the disclosure. Minor denoising, exposure correction, or crop work usually does not need the same level of detail, but the reviewer should still know that editing occurred. The final fields should record the reviewer, approval time, disclosure text, and any MLS or platform notes.

Start with the original image and create a dated working copy rather than replacing the source file. Ask the editor to keep a short change log for each output, including the prompt category, selected result, and any manual correction. For a staged living room, the disclosure should cover the added sofa, tables, lamps, rugs, and wall art, while the listing description should avoid claiming that the room is furnished. For a kitchen, be especially careful with appliances, countertops, cabinets, fixtures, and lighting because buyers may assume those items are included or already installed. For landscaping, disclose generated trees, gardens, patios, fences, and sky changes because they can affect the impression of maintenance and usable outdoor space. Before upload, compare the edited image with the original at 100% zoom and ask whether a reasonable buyer could confuse a digital addition with the physical property.

Legal and platform reality

New York City has moved this issue from general ethics into a specific disclosure environment. The city’s landlord-discrimination law, Local Law 153 of 2022, took effect on 1 January 2024. It requires a landlord or housing provider to disclose to a prospective tenant, before the person pays a fee or signs a lease, whether an advertising image of a property was created or altered using artificial intelligence. The disclosure must identify the specific image, state that it was AI-created or AI-altered, and describe the nature of the alteration. The law focuses on rental housing, landlords, housing providers, prospective tenants, fees, and leases, so it is not a universal rule for every residential sale or every listing agent. It is also not a blanket ban on AI staging, and it does not automatically answer every question about MLS rules, seller consent, or buyer transactions.

The practical question for a city, agency, or platform is whether a photo change is merely cosmetic or whether it could affect a housing decision. A generated sofa in an empty rental is likely to be treated differently from a generated view, renovation, or unit feature. A disclosure should therefore be specific enough to explain the alteration without making the listing look deceptive. For a sale listing, the safest approach is to follow the applicable MLS policy, state the material edits in the remarks, and retain an internal record even when the platform does not require a formal label. If a rule is unclear, ask the broker, MLS administrator, attorney, or platform before publishing. A cautious disclosure costs little and reduces the chance that a buyer later says the image implied a feature that was never present.

Common mistakes

The most common mistake is treating all AI edits as if they were equally minor. Adding a plant is usually low risk, while adding a swimming pool, a second balcony, a finished basement, or a renovated kitchen can change the substance of the listing. Another mistake is placing the disclosure only in fine print or in the agent’s private notes. The buyer should see a clear statement near the image or in the same listing panel, and the wording should match the actual edit. A third mistake is using the same caption for every photo, which can hide important differences between a staged room and a repaired exterior. A fourth mistake is assuming that an AI tool’s “virtual staging” label proves legal compliance; the tool may generate the image, but the listing party remains responsible for accuracy. The checklist should catch these errors before the image goes live, not after a buyer questions it.

When to act

Act before the first upload, not after a buyer asks for the original. The checklist should be completed during listing preparation, before the seller approves the marketing package, and again before any major revision. If the property is being rented, check the local rule and the timing requirement immediately because the New York City rule applies before a fee or lease. If the property is being sold, check the MLS and brokerage instructions at the same time, especially for exterior renders, renovation previews, or images that show a future condition. Revisit the checklist whenever a photo is regenerated, a prompt changes, or the listing description adds a claim about a feature. This timing matters because a late correction can create confusion even when the original disclosure was accurate.

Cost and pricing

The software cost is usually separate from the disclosure work. Basic AI staging tools may be free for a small trial or billed per image, while subscription plans commonly range from roughly $20 to $100 per month for active users. A one-off edit can cost about $5 to $30 per image, and a larger package of 10 to 20 images may cost roughly $50 to $300, depending on the provider, resolution, and editing level. Professional retouching, 3D rendering, floor plans, and custom exterior visualization can cost more, often $50 to $250 or more per image. The checklist itself has no fixed price, but it saves time by reducing rework, rejected uploads, and buyer disputes. For a small landlord or agent, the most economical setup is a free or low-cost staging tool plus a written approval sheet; for a brokerage, the better investment is a consistent review process that travels with every listing.

A practical workflow

Use a simple four-stage workflow: prepare, edit, review, publish. During preparation, confirm that the seller has approved AI use and that the listing agent knows which claims may be made. During editing, save the original, record the software and date, and list every material change. During review, compare the final image with the source and decide whether the disclosure should say only “virtually staged” or describe a specific alteration. During publication, attach the approved text to the correct image, keep the unedited file and edit log, and archive the final version. If an image shows a future renovation, label it as a visualization and do not present it as a completed feature. If the edit changes a view, ceiling height, room size, or structural element, stop and obtain a second review rather than relying on a generic caption.

Colossis approach

Colossis can support this process as an AI virtual staging workflow, not as a legal advice service. The useful output is a repeatable set of generated images paired with an edit record, reviewer fields, and buyer-facing disclosure text. The platform should preserve the original property image, record the staging changes, and make it easy to produce a version that is clearly labeled as digitally added. That approach fits the practical need behind an AI staging disclosure checklist: help a listing communicate space and style without making an unverified feature look real. It also gives sellers and agents a defensible record if a platform asks for confirmation or a buyer requests context. The goal is not to make every image sound alarming; the goal is to make the difference between the physical listing and the digital presentation clear, proportionate, and easy to verify.