Kan AI generere en duftprofil for en ny parfume, der appellerer til en bestemt målgruppe ?

Fragansindustrien har begyndt at udnytte AI til at forudsige sensoriske præferencer ved at analysere kulturel, biologisk og markedsdata. Disse systemer kan modellere, hvordan forskellige forbindelser interagerer med menneskelig lugtesans og følelsesmæssige reaktioner. AI-genererede duftprofiler er allerede blevet brugt i kommerciel produktudvikling. Den endelige menneskelige test – at bære parfumen – forbliver dog afgørende.

Background

The fragrance industry has begun leveraging AI to predict sensory preferences by analyzing cultural, biological, and market data, modeling how different compounds interact with human olfaction and emotional responses. These AI-generated scent profiles have already been used in commercial product development, though the final human test—wearing the perfume—remains critical.

AI systems can now generate preliminary scent profiles by combining large fragrance-ingredient databases with consumer preference models and demographic data, yet they cannot physically compound, test, or bottle a finished perfume because scent is a chemical rather than digital phenomenon. Current tools rely on olfactory databases and machine learning to map odor descriptors to ingredients and suggest accords that historically resonate with target age, gender, or cultural groups, but these proposals still require human perfumers and analytical instruments for validation and scale-up. Early-stage prototypes have guided perfumers toward novel accords, while end-to-end autonomous perfume creation without human oversight remains beyond present capability.

— Enriched May 13, 2026 · Source: IFRA