Technology, Society & Housing

NYC Wants You Told When an Apartment Photo Is AI-Made

A new New York City proposal would force landlords and listing sites to label AI-generated or digitally altered rental photos. The target is a problem generative AI has made effortless.

• Mayor Zohran Mamdani's 68-page "Rental Ripoff Report" bundles the AI-disclosure rule with 22 other tenant protections, including fixes for mold, fire hazards and broken elevators.
• The rule would apply to brokers, landlords and major listing platforms such as StreetEasy and Zillow, which would have to carry "clear and conspicuous" disclosures.
• The enforcement challenge is deeper than fraud: generative AI has made it cheap to build a perfect listing that never existed.

In a tight rental market, the first photo of a listing does most of the convincing. A bright, airy room, a sunny window, a floor that stretches further than it actually does. For years that meant a skilled photographer, a wide lens and careful staging — effortful but traceable. Then generative image models arrived, and the cost of a flawless listing dropped to nearly zero.

That shift is what New York City is now naming and regulating. The practice of luring renters with a listing that looks far better than the actual unit — historically using a photo of a nearby building, a staged model apartment, or an empty space that isn't part of the unit — is already called "housefishing." AI just made it trivially scalable. A broker no longer needs an empty apartment or a borrowed building; a few prompts produce a clean, photoreal unit that matches nothing on paper.

The disclosure rule answers the problem at the level of transparency rather than truth. It does not try to certify whether a photo is genuine. Instead it forces the broker, the landlord and the listing platform to disclose when an image has been digitally altered or AI-generated. The city's Department of Consumer and Worker Protection would coordinate enforcement with the platforms that host millions of listings, including StreetEasy and Zillow.

The rule sits inside a broader 68-page dossier of hearings and findings. Alongside AI-disclosure, the report targets black mold, pest infestations, broken elevators, fire-safety gaps and general landlord accountability. The common thread is information asymmetry: a renter signs a lease based on what they are shown, and the person showing it has every incentive to make the unit look better than it is.

That incentive is exactly why disclosure is hard to enforce in practice. If a broker removes a stain with basic photo editing, is that "digitally altered"? If they generate a new view from an existing photo, at what point does an edit become a fabrication? Most photo tools already brighten rooms and straighten walls. The rule draws a line at AI-made and significantly altered content, but policing the line requires detecting what the images actually are.

Platforms are the likely choke point. A broker can edit on their phone, but every image eventually lands on a feed. If StreetEasy or Zillow is required to tag or remove undisclosed AI listings, the market incentive tilts back toward honesty — because the listing that never existed can be caught at the moment it is published.

The proposal is not yet law, and real-estate groups have argued that new disclosure requirements add cost and friction without fixing the underlying shortage of housing. But it is an early and concrete attempt to apply ordinary consumer-protection logic to a new kind of deception — one where the lie is not in a word but in a picture.

Knowledge takeaway: NYC's proposed rule requires disclosure of AI-generated or digitally altered rental photos; the practice of luring renters with fake listings is called housefishing, and generative AI makes it far cheaper; enforcement depends on platforms like StreetEasy and Zillow tagging the images at the feed level.