Should people never use chatbots for skin questions?

They can be useful for general orientation or preparing questions. The answer should not be treated as a diagnosis.

When is a doctor more important?

For new, painful, bleeding, fast-growing, infected or persistent skin changes, and for strong reactions to products.

What is a safer basic routine?

Many people need only gentle cleansing, moisturiser and daytime sunscreen. Strong active ingredients should be introduced slowly and matched to the skin.

AI skincare advice shows the risk of confident answers

What this is about

The Guardian reported on 7 July 2026 on an everyday trend: people upload skin photos, ask chatbots for a full skin analysis and then buy products or change their routines. The reason is understandable. Dermatology appointments are expensive, waiting lists are long, and a chatbot answers immediately.

That is exactly why the topic matters. This is not science-fiction medicine. It is about bathroom decisions: retinol, acids, vitamin C, sunscreen, allergens and whether a rash is harmless or needs a clinician.

What AI skincare advice actually does

A general chatbot does not make a diagnosis. It processes text, images or product lists and generates a likely answer. That can be useful for simple orientation, for example preparing questions for a doctor. But it can also recommend a concrete routine without knowing skin type, medical history, medication, allergies or examination results.

In the Guardian report, experts cite examples: duplicated active ingredients, wrong product order, recommendations for products that do not exist or inaccurate claims about allergens. One dermatologist notes that there are more than 3,000 skin conditions and that chatbots are currently not good at separating them reliably.

Why it matters

Skincare feels harmless until a wrong routine damages the skin barrier. Too many active ingredients can trigger irritant contact dermatitis or worsen rosacea. The risk is higher when a user treats a skin change as a cosmetic issue even though it needs medical assessment.

The problem is not that AI can never help. Research into skin image analysis is real, and specialised systems can perform well in narrow tasks. The problem is the jump from controlled diagnostics to open-ended chatbot advice without sources, examination or responsibility.

In plain language

Imagine sending a friend a photo of your kitchen cabinet and asking which medicine to take. They may recognise labels and sound confident, but they do not know your blood tests, allergies or diagnosis. In skincare, the chatbot is often that friendly, confident adviser without an examination.

A practical example

A 29-year-old has dry, red patches on the face and asks a chatbot for a routine. The answer recommends vitamin C in the morning, retinol at night, exfoliating acid twice a week and a heavy cream. After 14 days the patches are worse because the real issue was rosacea and a damaged skin barrier. The person has spent 120 euros on products and still needs an appointment.

Scope and limits

First, a chatbot can provide general information, but it cannot replace skin swabs, biopsies or dermoscopy. Second, many training and teaching images have historically been uneven; skin changes on darker skin may be less well represented. Third, product recommendation is not neutral medical advice if sources, ingredients or commercial influence are unclear.

A sensible use is to record symptoms, collect questions and read basic information from reliable sources. A poor use is to accept a diagnosis, stack strong active ingredients and ignore warning signs.

SEO & GEO keywords

AI skincare, dermatology AI, ChatGPT health advice, skin diagnosis, skin of color, retinol irritation, rosacea, DermNet, medical AI, consumer health