“AI has emerged as a transformative tool in cosmetic dermatology, driving advances in image analysis, dermocosmetic development, and personalised treatment planning. Its application has enhanced diagnostic accuracy and optimised the identification of skin conditions and therapeutic strategies,” the researchers wrote. And AI applications have experienced “remarkable growth” across the category, they said.
Testing, devices, procedures and education
AI tech, for example, is being used to replace animal and human studies when assessing the sensitising potential of dermocosmetic substances. And whilst their equivalence “remains under debate” currently, many models are today making comparable predictions and uptake of computational methods will expand.
More visibly, AI technologies are being integrated into at-home and in-clinic dermatology devices to evaluate skin health and suggest suitable products, formulations and ingredients. US company Proven Skincare, for example, integrates genetic and environmental data to formulate customised skin care routines. Clinical and home-based devices like Visia and Opté are also improving skin assessment and treatment monitoring.
AI technology is also being used for cosmetic procedures, the researchers said, with AI models predicting patient response to laser treatments and outcomes for dermal fillers, for example. More broadly, industry is also leaning into AI as an educational tool to provide insights around product use, expected outcomes and post-treatment engagement.
However, use of AI in the cosmetic dermatology sector is not without its risks, the researchers warned: “Challenges persist, including data bias, the need for rigorous clinical validation and unequal access to these technologies across regions.”
Quality of care and data bias
According to the review, “one of the most concerning issues” is potential loss of clinical judgement as increasing automation of diagnosis and treatment can discourage continuous medical education and critical thinking amongst professionals. Inappropriate or unsupervised use of AI systems may also “jeopardise patient safety”, they explained, particularly on sensitive medical decisions.
AI models, they said, lack continuous learning mechanisms comparable to the cumulative experience of human clinicians. Looking ahead, therefore, the researchers said balance between AI and human expertise will be “fundamental”.
“AI is not intended to replace human expertise, but rather complement it. In cosmetic dermatology – where professional judgement is essential – AI functions as a supportive tool, enhancing efficiency without replacing the specialist’s role.” If industry can find this synergy, “significant advances” await, the researchers said.
“Another major risk” to overcome is in the subjectivity of aesthetic concepts like beauty and skin quality, which can lead to “inappropriate, unethical or culturally-biased recommendations”, they explained. Many algorithms, for example, are developed using small, low quality or insufficiently diverse datasets, also limiting the generalisability in clinical practice.
From a regulatory perspective, application of AI in dermatology also faces “a notable absence of specific legal and ethical frameworks” and lacks “robust international standards”, the researchers said.
From beauty to overall health
Speaking to Premium Beauty News, Robin Raskin, tech expert and founder and CEO of The Virtual Events Group, said: “In a nutshell, today’s products can recognise common skin problems, track changes over time, and make product recommendations pretty well. But the accuracy of the diagnosis depends on things that humans still do best: looking at a patient’s history, maybe the drugs they’re taking. Humans can also touch and palpitate the skin, which can often provide more information than a scan.”
Looking ahead, Raskin said the most exciting developments will be those that combine AI with other imaging techniques to establish “predictors of overall health”. Can a look at blood flow through the face give doctors some indication on heart health, for example? Can facial hair analysis yield insights into body health or possibly hormonal imbalances? “These things will happen,” she said.
“In 2027 and beyond, you’ll see AI combined with things like hyperspectral imaging to provide a better picture of what’s going on below the surface of the skin. You’ll also see more work being done to personalise cosmetics based on your AI skin analysis – so, you won’t get a product recommendation for a specific brand but you’ll get a formulated product that’s personalised for you; a product of one, if you will.”
Article initialement publié dans notre numéro spécial Ingrédients Cosmétiques, avril 2026, à lire ici. Sommaire
|
