AI-recommended skincare: what it can’t know about your skin

More and more people come to consultations with a routine already prepared. They didn't read it in a magazine, and an esthetician didn't give it to them: they asked a chatbot. They arrive with five products, an application order, and the conviction that the work is already done.

L'artificial intelligence often gives correct information: the right active ingredients, the right order, clear explanations. But a routine isn't built from information. It's built by looking at someone's skin, and a chatbot only works with what it's told.

Cover image generated with the support of artificial intelligence.

[INDICE:h2]

Why the answers seem so good

Because often it is, though not always: a model can be confidently wrong, confuse two ingredients, or cite studies that don't exist. But the most common problem is something else.

A language model has read an enormous amount of dermatological and popular science literature, and when it comes to theory, it's often more accurate than much of the content circulating on social media. If you ask what niacinamide does, the answer is usually correct. If you ask what order to apply serum and moisturizer in, the answer is usually correct.

This is exactly what makes the problem hard to spot: the error isn't in the individual claims, but in the leap from knowing what an active ingredient does to deciding that a particular person should use it.

The four points where the routine breaks down

1. Tolerance isn't a variable you can simply declare

Someone with reactive skin writes, "I have sensitive skin," and gets a routine tailored to that statement. But "sensitive" isn't a fact; it's a description. In consultations, I find very different situations behind the same word: a compromised skin barrier from using too many exfoliants, dehydrated skin that feels tight, a reaction to a single ingredient, and sometimes a problem that needs to be seen by a dermatologist. These situations call for different, sometimes opposite, responses.

2. It rarely starts with what you already have in your bathroom

The recommended routine almost always starts from scratch, as if the person had to buy everything. In reality, they usually already have several products open. Often, some overlap, and sometimes one of them is actually the cause of the problem.

The first question I ask in a consultation isn't "what's your skin type?" It's "what are you using right now?" Very often, the solution is to remove something, not add it. A model responding to a request for a routine tends instead to build a new one, because that's what it was asked to do.

3. The concentration and the formula aren't in the answer

This is the distinction I repeat most often: an active ingredient is not a finished product. Knowing that azelaic acid works on blemishes says nothing about which formulation actually delivers it, at what pH, with what level of tolerability, in what texture, or whether that texture suits skin that flakes in winter.

The advice generated stays at the level of the molecule. The difference between success and failure is almost always one level down, in the formulation.

“A correct answer given to the wrong person is still correct. But it's not the right answer for them.”

4. The recommended range may no longer exist

This is the most concrete point, and the least discussed. The products mentioned may have been discontinued, reformulated, or may not comply with regulations that have since come into force. This is the case with cosmetics containing high concentrations of retinol: Regulation (EU) 2024/996 set the limits at 0.3% for the face and 0.05% in body lotions, and products exceeding those limits have a date after which they can no longer be sold. Advice that cites them as a reference is already outdated by the time it's given.

Where it is a useful tool

I'm not saying not to use it.

It works really well for understanding: what an abbreviation in an INCI list means, the difference between two forms of vitamin C, why a preservative is a guarantee, not a flaw, and what a claim you read on a package means. On all of this, it's an excellent teacher, patient and available at any time.

It stops being useful when you ask it to decide. Understanding an ingredient and choosing what to put on your face are two different things, and only the first can be done with information.

How to use it well, in practice

1. Ask it to explain, not prescribe. "What does this ingredient do?" instead of "What should I buy?"
2. Show it the INCI list of what you already use and ask whether there are any overlaps. That's where it delivers the most value.
3. Introduce one new product at a time, a couple of weeks apart: if your skin reacts, you'll know what caused it.
4. Always check that the product is still available and compliant, especially retinoids and sunscreens.
5. If you have an ongoing skin condition, the right answer is a dermatologist. This isn't a limitation of AI: it's a limitation of cosmetics.

What remains at the counter

I find it interesting that the arrival of a tool that answers everyone has made the value of someone who looks at one person's skin more evident, not less.

A pharmacist who sees the person in front of them notices the redness on their cheeks while they talk about something else, sees that their hands are chapped, asks what they used last month, and hears the hesitation in their answer. These are details no one would type into a chat, because the person sharing them doesn't know they're relevant.

That's why we chose pharmacies and drugstores as the channel for our formulas, not out of nostalgia for brick-and-mortar stores: the hard part of giving advice isn't knowing things, it's understanding which of the things you know applies to this person here. If you already have a routine in hand, whether yours or one you've generated, take it to the counter at one of our partner pharmacies and drugstores and ask them what to remove.