Six months into the AI prompt boom, profiles sound smoother—and more interchangeable. The honesty question has shifted from what we write to why.
Open any dating app in 2026 and you'll notice a sonic shift—the same way every coffee shop eventually adopts the same playlist. Profiles have become smoother. Prompts that used to read as awkwardly personal now arrive polished, witty, emotionally articulate. "Looking for someone who can match my energy but also knows when to just sit in comfortable silence." "I'm fluent in sarcasm, bad puns, and genuine apologies." The sentences scan beautifully. They also scan identically.
When we first examined AI-assisted dating prompts, the question was whether they constituted dishonesty. Part two of that investigation suggests a more complicated problem: AI isn't always making people lie. It's making them sound like everyone else who used the same tool—which may be worse for finding connection than a clumsy but authentic sentence ever was.
The Honesty of Intent Versus Output
Most users we surveyed don't consider AI prompts fraudulent. They describe them as editing help—like asking a friend to punch up a cover letter. The intent is honest: I do want someone who values communication; I am genuinely looking for a long-term relationship. The output, however, is templated. AI excels at plausible universals. It struggles with the specific, embarrassing, irreplaceable detail that makes a person recognizable.
Consider two answers to "A perfect Sunday looks like..."
AI-assisted: "Farmers market, a long walk, cooking something ambitious while listening to a playlist we built together."
Handwritten: "Eggs on toast, reorganizing one drawer I will never reorganize again, and maybe calling my sister if I'm feeling social."
The first performs attractiveness. The second performs a person. After six months of AI prompt adoption, the feed is filling with the first—and daters are reporting a new fatigue: profile blur, the sense that everyone has become a well-marketed version of the same candidate.
When Smoothness Signals Risk
Experienced daters are developing heuristics. Overly balanced phrasing—every prompt the same length, the same tone, the same arc of humour-then-vulnerability—triggers skepticism. Not because polish is bad, but because human writing is usually uneven. We rush one answer and labor over another. We joke in the wrong place. AI smooths those tells away.
Some users now treat AI-polished profiles as a yellow flag rather than a green one—not evidence of deception, but evidence that they'll need extra time to find the human beneath the copy. First dates increasingly include meta-conversations: "Did you write that yourself?" asked lightly, not as accusation but as calibration. The answer matters less than how someone answers—defensiveness, humour, honesty about the assist.
The Genericity Problem
Genericity is the hidden cost of AI honesty. If fifty thousand people ask an model to "sound warm but not desperate, funny but not try-hard," the model converges on a narrow band of acceptable personality. Dating apps were already suffering from homogenization—algorithmic sorting, photo trends, performative hobbies. AI prompts accelerate it.
We spoke with a product designer at a mid-size dating platform who admitted, off the record, that engagement metrics initially rose when users adopted AI writing tools—profiles looked better, matches increased. But message-to-meet conversion flatlined. People matched with profiles that sounded ideal and met humans who didn't quite fit the voice. The bottleneck moved from discovery to recognition.
The Counter-Move: Deliberate Roughness
A counter-trend is emerging among intentional daters: deliberate roughness. Shorter prompts. Specific references. One woman in Edinburgh wrote: "I will talk about urban planning for too long if you let me." A man in Cardiff answered a prompt with a single line: "Bad at apps, good at showing up." These profiles perform worse in aggregate metrics, according to the designers we interviewed, but better in the quality of conversations they generate.
The roughness isn't anti-AI purity. It's a recognition that connection requires differentiation. If your profile could belong to anyone, it will attract people who are interested in anyone—which is its own kind of waste.
What Honesty Looks Like Going Forward
The honesty question has evolved. It's no longer "Did you write this yourself?" but "Does this represent how you actually show up?" AI can help articulate values you hold but struggle to phrase. It becomes a problem when it substitutes for values you don't hold but wish you did.
Our practical advice: use AI to clarify, not to invent. Start with a bad first draft in your own voice. Ask the tool to tighten, not to generate. Keep one prompt entirely unassisted—the weird one, the overly specific one, the answer that would never perform well in a focus group. That prompt is often the one your person is looking for. In a sea of smooth, the rough edge isn't a liability. It's a signal.