How AI is redefining beauty standards (and devastating people’s self-esteem)

There is a precise moment when you scroll through a photo on social media, look at it just a second too long and feel a micro-pang in your stomach. It happens while you’re on the couch, on your lunch break, in line at the post office. Those images have athletic, shiny, young, perfectly defined bodies. They are everywhere. More and more often they don’t come from a gym or a race, but from an algorithm.

Artificial intelligence is building a body image that slowly enters the head and sits there, like background music that in the meantime rewrites what it means to have the “right” body.

How AI narrates the athletic body and why this narrative becomes familiar

A study published in Psychology of Popular Media and created by a research group from the University of Toronto starts from a banal request: generate images of athletes and ordinary people. From there, a sample of 300 images created with platforms such as DALL·E, Midjourney and Stable Diffusion takes shape.

The images are observed in detail. Apparent age, build, muscles, clothes, body exposure, facial features. The picture that emerges has very precise outlines. The athletes generated by AI are almost always young, thin, muscular. Female bodies appear slim and smooth. Compact and defined male ones. The clothing tightens, reveals, showcases.

Even when the request remains generic, the images become largely male. Non-athletic bodies show a little more variation, while remaining far from true variety. In all this something is missing that exists in life: signs of aging, large bodies, visible disabilities. Absence becomes a constant presence.

When images become a distorting mirror

Billions of images pass through social media every day and an increasingly large portion arises from artificial intelligence. This continuous flow creates familiarity. That familiarity turns into reference. The reference becomes comparison.

The comparison works in silence. It leads to looking at yourself in the mirror with a harder look. It leads to feeling the body like an unfinished project. It leads to moving more as a punishment or stopping moving due to emotional exhaustion. It leads to eating with guilt or avoiding certain places because the body seems out of place there.

In the world of sport, where the body is already observed, judged and commented on, this pressure weighs even more. The idea of ​​performance is intertwined with that of appearance. The head fills with noise. The pleasure of movement loses space. Motivation wanes.

The absence of bodies with disabilities or with obvious signs of age tells another story. It tells who is seen and who is left out of the frame. It tells which bodies deserve attention and which become invisible. This mechanism works even when we don’t realize it.

AI as an amplified reflection of our prejudices

Artificial intelligence does not imagine from nothing. Absorbs. Repeat. Amplify. It takes the cultural material it finds online and returns it with a coherence that reassures the algorithm and impoverishes reality. This process includes gender stereotypes, rigid aesthetic models, silent exclusions. The result is a visual vortex that makes the space of normality increasingly narrower. Diversity is diminishing. Uniqueness loses volume.

This dynamic also affects how we feel in relation to others. A body perceived as inadequate tends to close itself off. The distance grows. Loneliness takes shape. All this happens without proclamations, without alarmist tones, in the simplest daily life.

The artificial images will continue to flow. The algorithms will continue to produce. We remain in the middle, with bodies that change, that get tired, that bear signs. We need a softer look towards ourselves, reminding us that those images tell of an aesthetic choice, not a measure of value. We need space to stay even outside the perfect frame.

Every body carries a history that no algorithm can copy: there is movement, memory, imperfection. There is real life. And that’s where it’s worth starting again.

You might also be interested in: