The moment of choice is intrinsically human. That is where art begins.
For more than eight years, the perfume industry has been quietly rewriting the relationship between human intuition and machine intelligence, using algorithms not to replace the perfumer's hand but to extend its reach across molecules, memories, and markets. From Narciso Rodriguez to Prada to Rabanne, fragrances shaped by AI now occupy the shelves of department stores worldwide, compressing development timelines from months to weeks while raising a deeper question: whether speed and optimization can coexist with the irreducible human act of choosing which story a scent should tell. The great fragrance houses—Givaudan, IFF, DSM-Firmenich, Symrise—have each built their own algorithmic tools, yet their master perfumers speak with one voice on the matter that counts most: the moment of creative choice remains stubbornly, essentially human. What the industry is navigating now is not a takeover but a transformation, and its outcome will depend less on the power of the tools than on the wisdom of those who wield them.
- Development cycles that once stretched six to eight months have collapsed to three weeks, sending a shockwave of acceleration through every corner of the fragrance supply chain.
- AI systems like Philyra, Carto, and Myrissi can now analyze millions of formulas, predict consumer emotional responses, and track regulatory changes across global markets faster than any human team—raising the stakes for houses that fall behind.
- Francis Kurkdjian and others warn that when every player has access to the same algorithmic instruments, the mid-market risks drowning in mathematically correct but culturally hollow fragrances, indistinguishable from one another.
- The industry's answer to homogenization is not to resist the technology but to insist that the decisive act—choosing which creative path is worth following—belongs to the perfumer alone, and no training data has yet captured that judgment.
- With the global fragrance market projected to reach 106 billion dollars by 2028, the competitive advantage is shifting toward those who can use AI as a creative amplifier while preserving the cultural memory and artistic differentiation that algorithms cannot manufacture.
The perfume industry has spent more than eight years experimenting with artificial intelligence, and the results are now visible on department store shelves. Sonia Constant used AI to deepen the musky heart of Pure Musc Blanc for Narciso Rodriguez. Three perfumers working with algorithms conjured a new jasmine accord for Prada's Paradoxe Virtual Flower. Rabanne's Phantom combined AI with neuroscience to identify the molecular combinations that activate the brain regions associated with seduction and energy. These are not prototypes. They are fragrances you can buy.
The technology arrived gradually, building on decades of computational chemistry. The turning point came in 2017, when Symrise partnered with IBM to build Philyra, an assistant trained on millions of formulas. Two years later, David Apel used it to create Egeo On Me, the first fragrance openly marketed as AI-developed. Today every major fragrance house—Givaudan, IFF, DSM-Firmenich, Symrise—operates its own AI systems. Givaudan launched Carto in 2019 to suggest accords during creation, and Myrissi to predict the emotional responses a fragrance will evoke in consumers.
Master perfumer Olivier Cresp describes AI as simply another tool in a long history of technological advancement, one that analyzes molecules and generates new structural ideas while leaving intuition and inspiration to the human. Calice Becker, director of Givaudan's perfumery school, frames it as a tool for memory above all: a perfumer carries thousands of materials and formulas accumulated over a lifetime, and AI ensures nothing is lost to forgetting.
The speed is transformative. What once took six to eight months now takes three weeks. Beyond creation, AI is reshaping reformulation—adapting existing compositions to shifting regulations across markets at a pace no human team could sustain, while simultaneously mapping the environmental footprint of each ingredient and identifying sustainable alternatives.
Not everyone is celebrating. Francis Kurkdjian, one of the most respected voices in fragrance, warns that AI will eliminate non-creative perfume work entirely—analyzing existing fragrances, extracting formulas, and producing mathematically aligned results at the push of a button. He has smelled fragrances made this way, and none have impressed him. The concern is not hypothetical: when everyone plays the same instrument, what matters is the musician.
Yet the perfumers themselves are clear on what cannot be automated. Becker says the moment of choosing which creative path is worth following is intrinsically human—that is where art begins. The fragrance market is projected to reach 106 billion dollars by 2028, and the question is no longer whether AI will change perfumery. It will. The question is whether those who use it best will remember that a scent is not just a formula—it is a story, a memory, a moment in time that no algorithm alone can create.
The perfume industry has spent more than eight years experimenting with artificial intelligence, and the results are now visible on department store shelves. A perfumer named Sonia Constant used AI to amplify the musky heart of Pure Musc Blanc for Narciso Rodriguez. Prada's Paradoxe Virtual Flower emerged from three noses—Antoine Maisondieu, Nadège Le Garlantezec, and Shyamala Maisondieu—working with algorithms to imagine a new jasmine accord that captured the flower's most delicate aroma. Rabanne's Phantom combined AI with neuroscience to identify the exact molecular combinations that would activate the brain regions associated with seduction, alertness, and energy. These are not theoretical exercises. They are fragrances you can buy.
The technology arrived quietly, like a scent itself. For decades, the industry relied on statistical models and computational chemistry—software that could predict how atoms and molecules would behave without requiring physical testing. A perfumer knew that to recreate the intoxicating smell of tiare flower, you mixed ylang-ylang, salicylates, and coconut. Then, in 2017, Symrise partnered with IBM to build Philyra, an assistant trained on millions of formulas. Two years later, a perfumer named David Apel used it to create Egeo On Me, the first fragrance openly marketed as developed with AI. Today, every major fragrance house—Givaudan, IFF, DSM-Firmenich, Symrise—has its own AI systems or collaborations with tech companies. The tools are everywhere now.
What these systems actually do is expand the perfumer's reach. Olivier Cresp, a master perfumer at DSM-Firmenich who created Light Blue for Dolce & Gabbana, Aire for Loewe, and Noa for Cacharel, describes AI as simply another tool in a long history of technological advancement. The algorithm analyzes existing molecules, ingredient combinations, and chemical reactions to generate new ideas for molecular structures. Combined with a perfumer's knowledge and intuition, it opens new creative possibilities. Calice Becker, a master perfumer at Givaudan and director of the company's perfumery school, frames it differently. She says AI arrived silently, like an aroma filling a room before you can identify it. The real creative shift came when Givaudan understood how to develop AI for memory—not for composition, but for preservation. A perfumer carries thousands of materials, emotions, and formulas accumulated over a lifetime. AI ensures nothing is lost to forgetting. The company launched Carto in 2019, a tool that suggests accords during the creation process. It also built Myrissi, which can predict the emotional response a fragrance will evoke in consumers.
The speed is transformative. What once took six to eight months now takes three weeks. A perfumer can request a summer fragrance with certain characteristics, the technology generates five or six proposals, they are tested and refined, and the work is done. This acceleration is reshaping the entire industry—not just creation but also the tedious work of reformulation, adapting existing compositions to changing regulations across different markets. AI can track regulatory evolution at a speed no human team could match. It maps the environmental footprint of each material and identifies sustainable alternatives without betraying the intended scent. It integrates what once seemed contradictory: creative ambition and regulatory compliance, speed and depth, innovation and responsibility.
But the industry is not unified in celebration. Francis Kurkdjian, a perfumer for Dior and one of the most respected voices in fragrance, anticipates disruption in the mid-market segment. He argues that AI will eliminate non-creative perfume work—it will analyze all existing fragrances, extract their formulas through chromatography, and let you press a button. The results will be mathematically aligned. Is that creativity? He is not sure. The algorithm cannot understand cultural context. He has smelled fragrances made this way, and none have impressed him. The concern is real: if everyone has access to the same tools, the advantage belongs to whoever can stand apart. When everyone plays the same instrument, what matters is the musician.
The perfumers themselves are clear on one point. Becker says the perfumer will not be replaced. A person deciding which path is worth following—that moment of choice is intrinsically human. That is where art begins, and no algorithm, however powerful, has learned to master it. Cresp agrees. As perfumists face changes in their palette of ingredients, the new models provide assistance, freeing them to spend time on inspiration and creativity. The industry is not disappearing into automation. It is being reshaped by it. The fragrance market is expected to reach 106 billion dollars by 2028. The question is not whether AI will change perfumery. It will. The question is whether the perfumers who use it best will be those who remember that a scent is not just a formula—it is a story, a memory, a moment in time that no algorithm alone can create.
Citações Notáveis
The perfumer will not be replaced. A person deciding which path is worth following—that moment of choice is intrinsically human. That is where art begins, and no algorithm has learned to master it.— Calice Becker, master perfumer at Givaudan
AI will eliminate non-creative perfume work. It will analyze fragrances, extract formulas, and let you press a button. Is that creativity? I'm not sure. It cannot understand cultural context.— Francis Kurkdjian, perfumer for Dior
A Conversa do Hearth Outra perspectiva sobre a história
So the AI doesn't actually smell anything. It's working entirely with data about molecules and formulas.
Exactly. It has no nose. It can't experience what a fragrance actually feels like when you wear it. That's the hard limit.
Then what's the real value? If it can't smell, how does it help a perfumer create something that smells good?
It's a memory and suggestion engine. A perfumer might carry thousands of formulas in their head, but they can't hold all of them at once. AI can hold all of them and say, here are the combinations that worked before, here's what happens when you mix these molecules. It's like having a library that talks back.
But couldn't that make everything the same? If everyone's drawing from the same data?
That's the fear. But the perfumers say no—because the data is different for each company, and more importantly, the person choosing what to do with the suggestions is still human. The algorithm generates five options. The perfumer picks one and modifies it. That choice, that sensibility, that's where the art lives.
Francis Kurkdjian seems skeptical though.
He is. He's worried about the middle market especially—that AI will make those fragrances uniform and forgettable. He's seen AI-made scents and they haven't moved him. But he's also not saying AI should disappear. He's saying it will change the game, and the winners will be the ones who understand that a fragrance needs a soul, not just a formula.
So the real question is whether the perfumer stays in control.
Yes. And whether they remember why they became a perfumer in the first place.