AI photos in real estate and rentals

Three kinds of AI listing photo

AI turns up in property listings in three very different ways, and only one of them is a scam on its own:

• Virtual staging: a real photo of an empty room with furniture, rugs and plants added by software. It is common and often legitimate, as long as it is disclosed and the room itself is shown as it is.
• Enhancement: a real photo with a bluer sky, greener grass, a tidier street or a brighter room. It becomes misleading when it hides something a buyer or renter would care about, such as a power line, a damp patch or a building site next door.
• Generated homes: rooms, views or whole houses made from a prompt. In a listing, this is the warning sign of a scam: the home may not exist, or may not be available.

The first two start from a real home and change part of it. The third starts from nothing. That difference decides what you can check, and what a detector can tell you.

Virtual staging: ask, then compare

Staged photos are usually easy to spot once you look for them: furniture placed a little too perfectly, rugs without wear, nothing personal anywhere, rooms that feel larger than the floor plan says. The simplest check is to ask.

• Ask the agent or landlord which photos are staged, and ask for the unstaged versions.
• Look for a note on the listing: in many places, agents and portals are expected to say when photos are virtually staged.
• Compare the staged room with the floor plan and the measurements. Staging can make a small room look bigger by choosing small furniture.
• Check that the parts that matter are untouched: walls, windows, floors, damp, cracks and the view.

Rental scams with generated photos

Rental scams are old. What has changed is that a scammer no longer needs to copy photos from a real listing, which a reverse image search could find: they can generate a home that exists nowhere. The pattern around the photos stays the same:

• The rent is far below similar homes in the area.
• The landlord is abroad, busy or unreachable, and can't show the home in person.
• You are asked for a deposit or the first month's rent before you see the property or sign anything.
• Payment is by bank transfer to a personal account, crypto or gift cards, rather than a traceable, protected method.
• There is pressure to decide today, because "many people are interested".

How to check a listing before you pay

1. 1

   Compare the rooms with each other

   Windows, floors, doors, light switches and ceiling heights should match from photo to photo. A kitchen with a different floor in every shot, or a window that moves between photos, suggests rooms made one by one.

2. 2

   Match the outside

   Find the address on a map and look at the street view. The building, the entrance, the neighbours and the view from the windows should match the photos.

3. 3

   Check that the address exists and fits

   The flat number, the floor and the number of bedrooms should fit the building. Search the address to see whether it appears in other listings, with other prices or other agents.

4. 4

   Search the photos

   A reverse image search finds photos copied from older listings, other cities or hotel websites.

5. 5

   Ask for a live video tour

   Ask the agent to walk from room to room on a live call, open a window, show the view and the building's entrance. A pre-recorded video can be copied or generated; a live walk-through is much harder to fake.

6. 6

   Verify the landlord or agent

   Check the agency's registration where your country has one, and confirm who owns the property before you sign or pay.

Then pay through a traceable channel with protection, such as a regulated deposit scheme or the portal's own payment where one exists. Never pay in cash, crypto or gift cards to someone you haven't met.

What the photos can still give away

Property photos are often wide-angle, bright and heavily processed even when they are real, so the usual signs need care:

None of these settles the question alone. The visit, the live tour and the paperwork do.

How Perspicax helps

Perspicax combines proven computer vision techniques with TrueEye 2.3, our state-of-the-art detector. It examines every region of a photo, down to the pixels, and answers in one of three bands:

It is built and measured on fully generated, photo-like images, the kind a scam listing invents. In testing, it wrongly flagged fewer than 1 in 2,000 original camera photos, and it caught 98.6% of images from the newest OpenAI, Google and ByteDance models on average.

Renters and buyers can save a listing photo and drop it into the check on this page. Portals and agencies can screen every upload through the Perspicax API.

For portals and agencies

A portal that lets generated homes through loses renters' trust quickly. A few rules keep staging honest and scams out:

• Screen new listing photos through the API, and send generated-looking photos to review rather than removing them automatically.
• Ask agents to mark virtually staged photos, and show that mark on the listing.
• Ask for at least one unedited photo of every staged room.
• Check new advertisers' identity and right to let the property before their listings go live.
• Watch for the same photos across several listings, addresses or accounts.

To talk through a set-up for a portal or an agency, contact us. For listings of goods rather than homes, see how to tell if product photos are AI.