The group chat has a new thing, and it wants your selfie.

For $150 to $330, an app will analyze your face and hand you a report: your ear shape, the "femininity" of your features, your skin health as a percentage, and a shopping list of procedures to fix what it finds. Qoves. Epica. Facemaxify. You upload two photos, check a box, and wait for the algorithm to tell you what's wrong with you.

The Guardian went inside this economy in August 2026, and the reporting should be required reading before anyone in your group chat uploads anything Established: The Guardian. Because here's the thing the apps don't put in the fine print: the science isn't there, the scores aren't stable, and the people whose research the apps cite are asking to be left out of it.

What actually happens when you pay

Danielle, 37, used to model. After two years of unemployment she saw a TikTok about a "science-backed" facial analysis and bought it on the spot. Three hundred and thirty dollars later, she had a 30-page report that scored everything from her ear shape to the femininity of her features, and recommended a rhinoplasty, a lower blepharoplasty, a canthopexy, chin implants, a malarplasty, microblading, lash enhancements, filler in her temples and cheeks, filler and Botox in her chin and jaw, and Botox around her mouth Established: The Guardian. She is now planning to travel to Mexico for procedures on her nose and eyelids once her credit card debt is paid off.

That's the funnel. Qoves charges $150 for "nonsurgical" recommendations and $180 more for the surgical ones, and has more than a million YouTube subscribers Established: The Guardian. The Guardian's own reporter received an 18-tab analysis of her features plus a 16-page "aesthetic protocol" complete with a mockup of her face with the recommended changes – a "healthy dose of lip filler" and makeup included Established: The Guardian. Cheaper tools like Facemaxify and Epica score your face against an opaque metric of "facial harmony," and the services layer on paid subscriptions. Epica's specialty is skin and makeup plans for women, and it directs users to affiliate links for cosmetics Established: The Guardian.

Diagnosis and treatment, sold on the same screen, by the same company. Notice a pattern yet?

The science is load-bearing. And it's not there.

The apps lean hard on "science." Qoves cites peer-reviewed research to lend itself credibility – the halo effect (documented, real, people do perceive attractive faces favorably) and studies linking attractiveness to pay. But here's the move: a documented phenomenon about how people perceive faces is not a prescription for your chin.

The researchers themselves are explicit. Dr. Justin Mogilski co-authored the 2018 study on facial feature perception that Qoves references – and his findings were about how a group of primarily college-aged Americans perceived certain features as masculine or feminine. He told the Guardian: "Would I make suggestions for facial enhancements based on our or anyone else's data? I would not" Established: The Guardian. Dr. Mitch Brown, who has studied the science of attraction for over a decade: "I don't think the science is anywhere clear enough to do that. I would caution anyone against that" Established: The Guardian.

And then there's Dr. Tracy Vaillancourt, whose work Qoves used in a video called "The Cost of Being Too Pretty." She doesn't study beauty – she researches bullying and mental health – and she is not a fan of the company keeping. On the checkbox users tick to confirm they don't have body dysmorphia before they get their list of flaws: "You can circumvent that so easily, it's just going to give them the appearance of due diligence... ultimately, they will be exploiting people who are vulnerable. These sorts of things are so destructive to people who have body dysmorphia, dysphoria, a history of eating disorders and people with low self-esteem" Established: The Guardian.

The detail that tells you everything

The Guardian reporter did the thing we'd all secretly do: she uploaded her own selfies.

Epica gave her a "skin health" score of 61%, flagged uneven texture as her main concern, and filled her screen with serums to fix it. Then she moved to her couch, retook the selfie in new lighting – and scored 81% Established: The Guardian.

Twenty points. Nothing changed but the lamp.

That is the whole industry in one number. The scan is not measuring your face. It is measuring your willingness to believe the score – and selling you the fix in the same breath.

The midlife catch: the list only gets longer

This business runs on aging anxiety, and aging anxiety has a target demographic: you.

At 28, the app finds a few things to note. At 48, the list gets longer – thinner skin, more lines, lost volume – because the "problems" it's scoring are just what getting older does. Skin changes, including dry and itchy skin, are a listed symptom of perimenopause on the NHS's own page Established: NHS. It's not a pathology. It's a season of life.

An app that profits when you find more problems will always find more problems. That is not a bug in the analysis. It is the business model. And it's the same machinery as the wider misinformation economy we've watched aim at this exact audience – the Guardian has already documented how perimenopause misinformation on social media is putting women at risk Established: The Guardian. The allergy-aisle cure. The face-scan verdict. Different products, same pitch: you are broken, and we can fix you – for a fee.

Red flags before you pay

If you're still tempted, here's what to look for before you hand over a single dollar:

  • The self-certification check. They ask you to tick a box confirming you don't have body dysmorphia. That's not a screen, it's a disclaimer – it protects them, not you Established: The Guardian.
  • The opaque methodology. Epica refused to disclose its algorithm. Qoves didn't respond to the Guardian at all Established: The Guardian. You can't audit a metric nobody will explain.
  • Scores that move with the lighting. A "skin health" number that jumps 20 points when you change rooms is not a measurement Established: The Guardian.
  • The fix sold on the same screen as the diagnosis. Serums, subscriptions, affiliate links to cosmetics, "nonsurgical" packages. If the diagnosis and the treatment come from the same checkout, you're not getting a second opinion. You're getting a sales pitch with charts.
  • No clinician anywhere in the loop. A 30-page report with zero human medical judgment behind it is a PDF, not a consultation.

What to do instead

If a scan result has you spiraling, close the tab. Your skin changing is normal, it's documented, and it's not a verdict on your worth Established: NHS. If your symptoms are bothering you, talk to your health care provider – that's the actual starting point, not an algorithm Established: OWH.

If you're genuinely considering a procedure, take the conversation to a real clinician: a dermatologist or plastic surgeon, in person, who examines your actual face and your actual health history – not a selfie you took in your bathroom. Ask them what they'd actually recommend, and get their honest answer on whether you need anything at all.

The bottom line

The AI face-scan economy isn't selling information. It's selling the feeling that you need fixing – and it manufactures the insecurity it profits from. The report will find flaws because finding flaws is the product. The experts it quotes want nothing to do with it. The scores change when the lighting changes. And the only thing the app is certain about is what you owe it.

You don't need an algorithm to tell you your face is yours. If a report suggests something worth doing, bring it to a human who can look you in the eye – and whose income doesn't depend on you finding more things to fix.