BEAUTY HAS NO COPYRIGHT

Who owns a beautiful face? AI Is About to Rewrite the Rules of Beauty

Fashion has always behaved as though it discovered beauty.

It didn't. It selected it, photographed it, named it, circulated it, priced it and eventually taught millions of people to recognize it.

The model was central to that machinery. A casting director saw something in a face. A photographer interpreted it. A magazine repeated it. A fashion house attached its clothes to it. Eventually, in exceptional cases, a face stopped being merely a face and became Linda, Naomi, Kate, Gisele.

Now AI is pulling that machinery apart, and the most interesting question may not be whether AI will replace models. That is the obvious question, which usually means it isn't the best one.

The better question is: what exactly is a model?

The Face Becomes Property

Fashion is already experimenting with an apparently sensible solution to one of AI's more obvious problems: give models ownership of their digital likeness.

H&M, for example, began creating digital twins of real models with the stated intention that the models themselves would retain rights to those twins and could license them elsewhere. The Business of Fashion has since reported the emergence of businesses specifically built around managing AI versions of human models. The model, in other words, is becoming both person and intellectual property portfolio.

This makes perfect sense. If a company scans a model today and then sends her digital twin to Paris, Tokyo and Mars for the next twenty years without the human being ever leaving her sofa, that model has created an asset. She should participate in its value.

But this apparently straightforward principle immediately opens a much stranger question.

What exactly does she own?

Her identity? Certainly. Her exact likeness? Probably. But then we enter more peculiar territory. Does she own her eyes? Her cheekbones? The distance between those eyes? The shape of her jaw? Her freckles? Or does ownership somehow emerge only when all those ingredients are assembled in precisely the right proportions?

Suddenly, copyright law has wandered into the casting session and started asking who owns the face.

Why Copy a Model When You Can Invent One?

The fascination of generative AI isn't creating a fake Bella Hadid. That's arguably one of the least imaginative things you can do with it. We have spent decades producing Kate Moss photographs. Producing an infinite number of synthetic ones hardly constitutes a creative revolution.

The far more interesting possibility is to take the qualities that make a face compelling, break them into language and begin again: eye spacing, orbital shape, cheekbone projection, jaw width, lip geometry, asymmetry, skin, age, expression, attitude.

Change one element and the face becomes ordinary. Change another and suddenly something happens.

Anyone who has spent enough time casting or photographing models knows this sensation. You look through hundreds of faces and one stops you. Often you cannot explain why. As a fashion photographer, I spent years looking for that face without completely understanding what I was looking for.

AI changes the relationship. Taste becomes experimental.

Generate. Alter. Reject. Return. Exaggerate. Reduce. Combine.

Why did that work? Why didn't this? What exactly did I see?

AI isn't merely generating a model. It allows us to interrogate why we thought someone looked like a model in the first place. The mysterious eye of the casting director suddenly has something resembling a laboratory.

That could become considerably more interesting than replacing a day rate.

Cindy, Kaia and the Problem With Owning Beauty

Consider Cindy Crawford and Kaia Gerber. Kaia is unmistakably related to Cindy. Nature has taken an extraordinarily recognizable face, shuffled its variables and produced another extraordinarily recognizable face.

Nobody would suggest that Kaia Gerber infringes Cindy Crawford's intellectual property.

Biology does variations. Culture does too.

There are surely numerous women walking around the world who occupy nearby coordinates in the enormous landscape of possible human faces. They may share the cheekbones, the eyes, the proportions, perhaps even the mole. None of them needed a license from an agency before being born.

This is where the idea of owning a face becomes slippery.

Perhaps what deserves protection isn't the face itself. It is the face once it becomes a sign.

Cindy Crawford isn't merely an arrangement of anatomical features. Neither is Bella Hadid. “Bella Hadid” contains decades of accumulated cultural information: CHANEL, Steven Meisel, Paris, magazines, campaigns, controversy, famous relationships, public breakups, paparazzi, Instagram, celebrity and fashion history.

Fashion took a human face and turned it into intellectual and commercial capital. That sign has value, and protecting it makes sense.

But protecting the ingredients from which it was constructed is another matter entirely.

Beauty has no copyright.

Fashion Never Invented From Nothing

The same confusion now surrounds AI and originality.

AI image-making is frequently criticized because models have learned from existing imagery. There are serious questions here about copyright, compensation, unauthorized reproduction and the commercial exploitation of artists. Those questions shouldn't be waved away merely because the technology is exciting.

But fashion should be particularly careful about pretending that creativity has historically occurred in some immaculate white room untouched by other people's work.

Fashion is an enormous reference machine. Designers reference designers, photographers devour photographers, stylists quote decades and creative directors raid archives with impressive confidence. Models arrive carrying echoes of previous models. Entire seasons are built around the rediscovery of something fashion had supposedly forgotten twelve years earlier, at which point everybody congratulates themselves on having discovered it.

Art isn't much cleaner.

Picasso absorbed other painters, movements, cultures and contemporaries relentlessly. Artists have always studied other artists partly by copying them. Photographers learned by looking at photographs. Designers learned by taking clothes apart. Musicians learned other people's songs. Writers read.

The interesting question was never whether something entered your visual vocabulary. Of course it did.

The question was what happened to it after it got there.

Perhaps the creative progression has always been:

Imitation → absorption → transformation → authorship.

AI hasn't invented that process.

It has made the process embarrassingly visible.

The Machine Can See Our References

This may be one reason AI makes the creative industries so uncomfortable. Human influence is conveniently invisible.

Nobody asks a photographer to disclose every Irving Penn photograph, Guy Bourdin advertisement, Renaissance painting, movie, album cover, stranger, lover, magazine and childhood memory that entered their brain before they pressed the shutter. We accept that this vast cultural compost has somehow been absorbed into something we dignify with the word taste.

A machine's influences, however, can theoretically be examined. Suddenly we can see the digestive system.

And because we can see it, we have begun questioning a process that culture previously romanticized when humans performed it.

That doesn't make every use of AI legitimate. Directly asking a machine to reproduce a living artist's recognizable work so that you don't have to employ that artist is fundamentally different from absorbing references and transforming them into something new. The technology doesn't abolish authorship simply because authorship was never as pure as we liked to imagine.

But the useful boundary isn't between influenced and uninfluenced.

Nothing interesting is uninfluenced.

The boundary is between imitation and transformation.

And fashion, of all industries, should understand the difference.

Photography Is Dead. Images Won.

Photography had already begun losing its monopoly long before generative AI arrived. Instagram merely accelerated the collapse.

The photographer, model, magazine, editor, publisher and audience once occupied clearly defined positions. Fashion imagery passed through an elaborate hierarchy before reaching the public. The photographer photographed. The model modeled. The magazine published. The audience looked.

That arrangement now seems almost quaint.

The model publishes. The audience photographs. The customer art-directs. The algorithm edits. The brand becomes a media company. The influencer becomes a retailer. The retailer becomes a studio. And increasingly, the image itself doesn't require a camera.

Photography didn't disappear.

Its borders did.

This is why saying Photography Is Dead isn't particularly pessimistic. Images won.

There are more images, made by more people, moving through more channels and performing more functions than at any point in human history. Photography was once the dominant industrial process for manufacturing them. It is now one process among many.

AI is simply removing another boundary: the requirement that an image must first have existed in front of a lens.

For photographers, that can sound existential. For images, it is liberation.

The Model After the Model

Fashion's first instinct will naturally be to recreate its existing world with the new technology. New tools usually spend their infancy impersonating old ones.

Digital models will wear digital clothes. Virtual try-on will place real clothes on synthetic representations of consumers. Real models will license digital twins. Brands will produce campaigns without flying thirty people, forty suitcases and one exceptionally nervous garment bag to Ibiza.

Much of that transition is already underway.

But that's merely using tomorrow's technology to reproduce yesterday's fashion system.

The more radical possibility is that AI changes the very idea of the model.

Why should there be one permanent face? Why should a campaign require the same human identity across every image? Why couldn't a face evolve with the story? Why couldn't age change? Why couldn't gender become ambiguous, ethnicity become deliberately difficult to place, proportions drift, beauty become stranger from one frame to the next?

Why couldn't beauty itself become a creative material?

For more than a century, fashion photographers manipulated light, pose, makeup, hair, film, lenses, printing, retouching and eventually Photoshop to construct an image of beauty while maintaining the useful fiction that the camera had simply found it.

The model was “naturally” beautiful after three hours of hair and makeup, carefully selected lighting, a $100,000 production, 600 exposures, one chosen frame and an enthusiastic encounter with retouching.

AI destroys that fiction.

There is no discovered beauty anymore. There is constructed beauty.

Perhaps there always was.

Looking Inside the Beauty Machine

That may be the genuinely revolutionary thing about AI fashion imagery. Not that it can make beautiful people who don't exist. Fashion has effectively been doing that forever.

The revolutionary part is that we can now manipulate the components of beauty with extraordinary precision and watch our own reactions.

Make the eyes slightly wider. Nothing. Alter the jaw. Worse. Introduce asymmetry. Interesting. Change the age.

Suddenly extraordinary.

The machine isn't merely generating faces. It is reflecting our preferences back at us.

And that could eventually become uncomfortable, because fashion may discover that the supposedly mysterious instincts behind casting, beauty and desirability aren't quite as mysterious as we imagined. Taste may contain patterns. Those patterns may contain prejudices. And once they can be generated, varied and measured at enormous scale, some of fashion's most carefully protected intuitions may begin looking suspiciously like data.

AI could expose biases alongside preferences. It could reveal which characteristics we repeatedly reward, which we reject, which combinations we describe as expensive, elegant, sexy, cool, strange or wrong, and how much supposedly individual taste has been culturally programmed.

That raises a much more provocative possibility.

For generations, fashion has insisted that beauty is subjective while operating enormous industries built on remarkably consistent judgments about it. Agencies cast. Magazines select. Brands approve. Algorithms now measure the response. Everyone insists there are no rules, yet somehow the same kinds of faces keep making it through the door.

AI may finally call fashion's bluff.

If millions of variations can be generated and our reactions to them observed, the machine may begin mapping something the industry has preferred to keep mysterious: the grammar of beauty itself.

Make Beauty Stranger

That doesn't mean there is one universal formula for a beautiful face. Quite the opposite. The interesting discoveries may occur where formulas fail, where an apparently “wrong” proportion suddenly produces fascination, where age improves rather than diminishes a face, where asymmetry creates character, where a feature conventional casting might reject becomes the entire reason an image works.

Beauty has always depended partly on deviation.

The perfect face can be remarkably boring.

And this is where AI becomes more interesting than the endless production of synthetic perfection currently flooding our screens. If we use it merely to manufacture smoother skin, larger eyes, smaller noses and infinitely compliant twenty-two-year-olds, we will have built one of the most powerful visual instruments in history and used it to automate a beauty filter.

Fashion can do better than that.

It can use the machine to make beauty stranger.

It can test the edges of attraction, elegance, age, identity and imperfection. It can construct faces that no casting director would have known to look for because no such person had yet walked into the room.

And perhaps, in doing so, it can learn something uncomfortable about the faces it has been selecting all along.

For an industry that has spent generations manufacturing beauty while claiming merely to recognize it, that is a far more interesting disruption than whether the next campaign was photographed with a Hasselblad or generated on a computer.

AI isn't just giving fashion another way to make images.

It is giving fashion a way to examine its own gaze.

And once we can see how the beauty machine works, we may never look at a beautiful face quite the same way again.