Women are paying hundreds of dollars for AI face-scoring reports. Experts say the science doesn't hold up.

Key points

• Danielle, 37, paid $330 to a company called Qoves for a 30-page AI-generated facial analysis report after seeing it advertised on TikTok.
• Experts say facial analysis services of this kind are not backed by peer-reviewed science.
• The trend is linked to 'looksmaxxing' culture, an online movement focused on maximising physical attractiveness, which has grown rapidly on short-video platforms.
• Researchers warn the reports can worsen body image problems and nudge users towards unnecessary cosmetic procedures.

Danielle had been a model. She knew what it felt like to walk into a room and have people notice her. Then two years of unemployment drained her savings and, quietly, her confidence. "It was easy to wonder," she said, "if I was prettier, if some of my features were different, would that help?"

That doubt was exactly what a TikTok video seemed to answer.

The video advertised a company called Qoves, which promised "science-backed" advice for a so-called glow-up, meaning a noticeable improvement in appearance. Danielle paid $330, uploaded her selfies, and waited. Weeks later, a personalised 30-page report arrived, scoring and critiquing her face.

What is 'looksmaxxing' and why is it spreading?

Looksmaxxing is an online movement built around the idea that every physical feature can be optimised, scored, and improved. It started in niche internet forums but has reached millions of people through TikTok and Instagram, where creators rank faces and recommend everything from skin routines to surgical procedures.

AI facial analysis services, software that scans a photo and assigns numerical scores to features like jaw width, eye spacing, or nose symmetry, have grown alongside this culture. Companies typically frame their tools as objective and data-driven. The pitch is simple: science tells you where you fall short; we tell you how to fix it.

Is there real science behind these scores?

No credible peer-reviewed evidence supports the idea that an algorithm can reliably measure human attractiveness, according to experts quoted in the original Guardian investigation. Attractiveness is culturally variable, changes across time, and is influenced by movement, expression, and social context, none of which a still photo captures.

Dermatologists and psychologists consulted by journalists also flagged a second problem: even if the scoring were accurate, delivering a detailed written critique of someone's face carries serious psychological risk.

For Danielle, that risk was not hypothetical. After reading her report, she said it "confirmed her flaws" rather than helping her. That phrasing matters. She arrived already doubting herself. The report gave those doubts a number and a name.

What should users know before buying one of these services?

There are four practical things to understand before spending money on a facial analysis report.

First, a company calling something "science-backed" is not the same as that claim surviving peer review. Ask which published studies underpin the scoring model.

Second, the target audience for these services is often people already struggling with body image, meaning the customers most likely to buy are also the most likely to be harmed.

Third, reports that suggest cosmetic procedures as a remedy carry a financial conflict of interest worth considering: the company profits whether or not surgery was necessary.

Fourth, if you are finding it hard to look in the mirror without criticising what you see, a facial score will not fix that. A GP (family doctor) or a therapist who works with body image is a more useful starting point.

Qoves did not respond to a request for comment at the time of original reporting.

Common questions

Are these services regulated anywhere?

In most countries, AI beauty-scoring tools fall outside the medical device regulations that govern clinical diagnostics, so companies face few legal obligations around accuracy or safety claims.

Can a photo-based score accurately measure attractiveness?

Experts say no. A static selfie cannot capture the expression, movement, or social context that shape how other people actually perceive a face, and no large peer-reviewed study has validated these commercial scoring models.