"Isn't that just catfishing with extra steps?" It's the reflexive objection to AI dating photos, and it deserves a real answer instead of a defensive one — because there is a line, people do cross it, and the interesting thing is where the line actually sits. Spoiler: it has almost nothing to do with whether AI was involved.
Catfishing is an identity crime, not a production method
Catfishing means presenting yourself as someone you're not: another person's photos, a fabricated identity, a materially false self. The harm is specific — the person who shows up isn't the person advertised, and the deceived party's time, trust, and sometimes safety paid for it.
Now notice what dating culture has always accepted without calling it catfishing: professional photography, flattering angles, makeup, good lighting, the photo from your best day rather than your average one. Every profile is curated; the social contract was never "unfiltered documentary truth." The contract is narrower and sharper: the person in the photos must be the person who arrives.
That's the line. It's a resemblance standard, not a method standard — and it cuts straight through the AI question.
Applying the standard to AI photos
Clearly fine: photos generated from your own selfies that render your face in better light and better scenes — the enhancement category. The person who shows up is the person in the photos; what was synthetic was the café, not you. Morally identical to hiring a photographer. (Notably, this is the design constraint legitimate services build around — "looks like you, only better" isn't just marketing, it's the product's whole defensibility.)
Clearly catfishing: synthetic photos of someone who isn't you — a generated stranger, a "better-looking version" that's drifted past recognition, someone else's AI twin. The viewer is making decisions about a person who doesn't exist. Method irrelevant: this was catfishing when it was done with a stranger's Facebook photos in 2012.
The instructive middle: heavily filtered real photos — your actual photo, beauty-app'd past recognition. Real source, deceptive output. This is the case that proves the line isn't about AI at all: a heavily filtered genuine photo is more deceptive than a well-made AI photo of your true features. The moral variable is the gap between image and arrival, full stop.
The gray zones, honestly handled
- Era drift: your real photos from five years and one hairstyle ago. Technically you; functionally a first-date gap. The standard is current resemblance.
- The scene fiction: you've never been to that generated rooftop. Is implying a lifestyle deceptive? The honest answer: scenes are understood as staging — same as a photographer's location shoot — as long as the photo isn't presented as biography ("me in Positano last summer" as a caption crosses into a lie; an uncaptioned golden-hour photo doesn't). Keep captions honest and the scene fiction stays in the photography-license zone.
- Best-case-you: the genuinely flattering generated shot that's at the top of your real range. Fine — that's what "your best photo" has always meant. The test is whether a friend says "great photo of you" or "who's that?"
The practical enforcement (why honesty wins anyway)
The line isn't just ethical — it's load-bearing in every practical direction: app verification matches photos against your live face; the report button after a bad-faith first date is the deepest moderation layer apps have; and the entire point of dating photos — unlike any other photo genre — is an in-person meeting where misrepresentation gets audited within seconds. Catfishing with AI isn't a clever exploit; it's an expensive way to schedule your own exposure.
The one-sentence version
If the person who walks in is the person in the photos, it's photography; if not, it's catfishing — and the tools involved were never the question. Use the technology to close the gap between your photos and your best honest self, not to open one.