A version of this commentary first appeared in Cosmetics Business on 27 July 2026 as part of Alessandro Carrara’s article.
Lisa Balm, Executive Creative Director
Every week another beauty company launches an AI companion or advisor. L’Oréal has teamed up with OpenAI’s ChatGPT. Estée Lauder’s Jo Malone London offers a personalised fragrance wardrobe on Pinterest. Fenty Beauty has partnered with WhatsApp. Douglas Group has its own AI Beauty Advisor chatbot.
As more beauty companies jump on the AI bandwagon, most of the conversation around it has settled into a debate: will AI kill organic discovery, and will it put in-store experts out of a job?
Both are worth asking. But underneath them sits a question that matters just as much: who ends up owning the relationship once the advice moves onto someone else’s platform.
Where the Trust Actually Moved
Customers were never asking for chatbots. What changed is where trust gets built. The question that used to go to a search bar, or to the woman behind the counter, now goes to ChatGPT, to Perplexity, or to a stranger shade matching on TikTok in her bathroom mirror. Beauty has always run on borrowed trust: a friend’s word, a reviewer’s honesty, a consultant’s expertise. Brands are now spending heavily to buy their way back into that conversation, because the conversation itself moved before they did.
Two Futures for the In-Store Expert
By 2026, most customers will arrive at a counter already knowing roughly what they want. That leaves in-store experts with two possible futures. In one, the advisor becomes purely transactional: locating the product, processing the sale, repeating information the AI already gave the customer on the way in. In the other, AI absorbs the functional work, analysing skin type, checking ingredient compatibility, narrowing the field, and leaves people to do the part technology genuinely cannot: reassurance, judgment, the read of a room a chatbot doesn’t have.
Which future a brand ends up in isn’t decided by the technology. It’s decided by what the brand chooses to protect.
What Brands Give Away One Question at a Time
If the transactional version wins, brands and retailers hand over something they spent years building: fragrance memory, shade history, the skin diary a customer built one question at a time across dozens of visits. That knowledge doesn’t disappear, it just moves house, into a platform the brand doesn’t control and didn’t build. The platform owns that now, and it can change the rules at any time.
That’s a commercial risk long before it’s a philosophical one. A brand that outsources its customer relationship to a third-party AI has also outsourced the thing that made it defensible in the first place: the accumulated, particular knowledge of what a specific person actually likes.
The Real Risk Isn’t the Robot
None of this is an argument against AI beauty advisors. Used well, they save people time, widen what gets discovered, and take the friction out of decisions that used to require a shop visit and a bit of nerve. The argument is against treating them as a plug-in that can be bolted onto an existing customer relationship at no cost.
The beauty companies that come out ahead won’t be the ones with the most sophisticated chatbot. They’ll be the ones who decided, deliberately, which parts of the relationship they were willing to hand to a platform and which parts they built their own AI to protect instead. The question isn’t whether AI will influence how people discover beauty products. It already does. The question is which brands still own that relationship once it does.
