AI Personalized Skincare

L’Oréal and Shiseido are named companies working on AI-enabled skin assessment and personalized care. L’Oréal applies computer vision to facial and skin analysis, while Shiseido uses skin-condition information to tailor recommendations and routines. Procter and Gamble, Unilever, Johnson & Johnson, Estee Lauder and Coty also appear among the organizations associated with recommendation, conversational-assistant or formulation approaches.

Nomige, Detail, Revieve Oy and CareCare are named startups in the available data, but the evidence does not support ranking them as the most promising. Nomige is associated with customized facial-care products informed by DNA and lifestyle data. Detail, Revieve Oy and CareCare appear in the organization list, with activity primarily recorded as news rather than patents or publications.

University of California at Los Angeles, Karpagam University and Karunya Institute of TechnologyandSciences are named research organizations. UCLA researchers evaluated a machine-learning skincare recommendation for mild-to-moderate acne. Karpagam University and Karunya Institute researchers are associated with image-based assessment of concerns such as acne and oiliness, linked to tailored guidance.

The data records 10 funding rounds totaling $2.38 billion and one grant worth $3,447,840 since 2020. These totals include entries such as infrastructure financing and institutional funding, so they should not be interpreted as investment dedicated to AI personalized skincare. The figures are best treated as broad funding signals associated with the topic, not a clean sector-specific total.

Sephora USA, Inc and ENTERTAKE CO., LTD. are the leading named patent holders, with two patents each. Other listed holders include Pbs Consulting, Cosnetix, Llc and SkinMinds Derma Private Limited, with one patent each. These counts identify organizations associated with the patents; they do not establish the commercial importance or technical quality of the inventions.

Patent activity declined 57.1% in the last 12 months compared with the previous 12 months, while 10 patents fall within the 24-month window. The overall direction is therefore down in the latest annual comparison. Recent work still includes new technical directions, such as ENTERTAKE CO., LTD.’s September 2026 patent for a cartridge-based skincare device linked to user skin-condition information.

AI personalized skincare has commercial applications, but its evidence base and technical maturity vary by approach. The data records 60 publications in 2025, with publication activity growing 76.5%, while patent activity declined 57.1% in the latest annual comparison. Image assessment and recommendation tools are established directions; longitudinal adaptation and multi-omics integration point to further development.

Recent developments include a September 2026 study of age-adjusted machine learning and facial skin microbes, an August 2026 review of multi-omics and AI, and a September 2026 cartridge-based device patent from ENTERTAKE CO., LTD. In March 2026, DreamSkin launched an AI skincare platform focused on melanin-rich skin. These examples span research, connected devices and consumer-facing services.

Applications include at-home skincare (536 documents), beauty e-commerce and retail (315), salons and spas (246), and dermatology and medical aesthetics (63). At-home tools can support personal routines, while retail systems help guide product selection. Salons and spas are another active application area; medical and aesthetic uses connect skin assessment with professional settings.

The five approaches are image-based skin assessment, profile and questionnaire recommenders, longitudinal adaptive personalization, conversational beauty assistants and customized product formulation. Profile and questionnaire systems have 1,707 associated documents, while image assessment has 232. They differ in their inputs and outputs: from interpreting a scan or user profile to updating routines over time, answering questions or tailoring formulations.

The United States leads the listed country activity with 6,025 documents, followed by the United Kingdom with 2,131, Germany with 2,014 and Japan with 1,729. Country tags can overlap, so these counts should not be added to estimate a global total. The figures describe recorded document activity, not necessarily where products are manufactured or sold.