AESTURA, Amorepacific’s dermocosmetic brand, is expanding its artificial intelligence (AI)-based skin analysis technology to cover skin diseases, moving beyond conventional assessments of visible skin conditions.
AESTURA said it presented its latest research on AI-powered skin diagnosis and its technology roadmap at KDD 2026, a leading global conference on knowledge discovery and data mining, where it participated as an official sponsor.
About 3,500 AI experts, researchers and developers from Korea and around the world attended Korea Day at the Jeju International Convention Center on Aug. 11.
During a special session, AESTURA unveiled research findings from a skin diagnosis engine designed to incorporate the clinical criteria used by dermatologists in actual medical consultations. The presentation was delivered by Kim Hye-sung, a dermatology professor at Catholic University of Korea’s Incheon St. Mary’s Hospital and an advisory professor to AESTURA’s research advisory board, and Ahn Jong-hoon, head of Amorepacific’s AX division.
The project combines medical expertise with skin data to enhance the accuracy and scope of AI-based diagnosis.
Researchers used approximately 750 patient datasets and around 1,000 datasets collected from the general population through Amorepacific’s clinical laboratory. Dermatologists participated directly in data labeling and de-identification to establish a clinical dataset for AI training.
The researchers then added data related to 29 skin diseases to an existing AI system that analyzed conditions such as pores, wrinkles, redness and pigmentation. The number of skin-related conditions that the system can analyze has consequently been expanded to 33.
AESTURA said the project is intended to bridge dermatology and everyday skincare through AI. Rather than limiting the technology to measuring visible skin characteristics, the company aims to translate professional dermatological knowledge into personalized skincare experiences.
The company also plans to expand the use of clinical data while strengthening privacy protections. One of the key technologies under consideration is federated learning, which enables AI models to learn from data held by different institutions without requiring the underlying personal data to be directly shared.
AESTURA will also continue improving its machine-learning models to enhance the accuracy and reliability of skin analysis.
The company’s longer-term goal is to allow consumers to receive more detailed information about their individual skin conditions in everyday settings and use those insights to develop personalized skincare routines.
Industry observers say the expansion of AI skin analysis from cosmetic characteristics to skin-disease-related indicators could broaden the convergence between dermocosmetics and digital healthcare. However, they also note that clear boundaries between medical diagnosis and consumer skincare services, as well as data reliability and regulatory compliance, will be important considerations as the technology moves toward commercialization.
Ahn Jong-hoon emphasized that dermatology and beauty share the common goal of maintaining healthy skin and presented AESTURA’s long-term AI roadmap for connecting professional skin diagnosis with everyday skincare.
An AESTURA representative said the project is significant because it establishes a foundation for an AI skin diagnosis service that incorporates dermatologists’ clinical expertise.
“We will continue combining AI technology with dermatological research to provide consumers with more reliable and personalized skin solutions,” the representative said.
