Since January 1, 2026, California law (AB 723) has required all real estate agents to disclose when the photos used in their listings have been digitally altered, and to provide access to the original photo, in the interest of more transparency about what is actually being offered. Meanwhile, New York is already proposing and evaluating similar laws for rental listings. But what is all this about? Why do authorities need to step in to protect consumers?
It's not hard to notice how fast technology is advancing, and with artificial intelligence gaining strength and becoming an increasingly universal tool, we need to step back and ask ourselves how far we are willing to go. With social media full of altered images and videos selling lifestyles, bodies, and faces that don't really exist, it's clear things are getting out of hand. But this problem is repeating itself everywhere (music, art, literature, and of course photography), something that affects everyone who works in this field, and the consumers who rely on images to, for example, buy or rent a property. So we need to accept that governments are starting to enforce laws, but from our side we also have to adopt healthy practices so we don't abuse these tools and become part of the problem.
We had already started to see this: people who take a photo with a phone and ask an AI to "improve" it, to add or remove things, change it, copy an editing style. At ConeCat, however, we stay true to the manual editing process, making decisions from a human perspective: every photo we deliver goes through manual HDR exposure fusion and brush-layer editing, done by hand by an artist-editor, not an automated "enhance" filter. Of course it's slower, but it's also the only way to get light falloff, shadow depth, and color temperature to read as real, because they are real. A trained eye still catches what an AI model only guesses at.
This doesn't mean we're against AI. We're honest that it plays a small role that doesn't alter the reality of what's being offered: occasionally we use generative object removal to clean up something distracting in the frame: a stray cable, a car in the wrong spot. We keep this to a minimum on purpose, because generative fill has a real cost to image quality: texture, sharpness, and consistency with the rest of the photo all suffer when a tool guesses what should be there instead of a camera having actually captured it.
So the line for us isn't "AI, never." It's quality first. We use it where it doesn't compromise the result, and avoid it where it does.
That distinction matters to us, and increasingly it matters to the people we photograph, beyond buyers, guests, and architects, the three largest audiences behind everything we do. Nobody wants to fall in love with a property that doesn't exist just because a retouched photo shows it.