Bain and Comité Colbert reported substantial AI use among surveyed luxury customers, especially top-tier spenders. “New baseline” is an editorial interpretation, not a census of all luxury shoppers.12

Bain reported that 82 percent of top-tier luxury customers in its study used AI during their most recent purchase. Segment definitions, sample, countries, and survey methods govern the estimate; it should not be generalized to all high spenders worldwide. Country, reuse-intent, reasons-for-use, and pre-store-use percentages are also stated survey behavior and intention, not independently observed transactions or future use.12

The reasons shoppers give are practical: 68% say AI speeds up their decisions, 55% say it gives them more confidence in quality and product details, and 52% say it surfaces brands or pieces they would not have found otherwise. Nearly half of in-store shoppers, 47%, used AI before ever walking into a boutique. The AI conversation is happening before the sales-floor conversation starts.

The industry is still moving in the back office

Bain reported rising strategic priority and deployment across surveyed luxury houses. Customer-facing, back-office, and at-scale categories depend on the report’s definitions and respondent sample, not the entire industry.12

Translation: the industry is using AI to run supply chains, forecast demand, and streamline logistics; all real, all valuable, none of it visible to the client at checkout. Meanwhile that client has already used AI to research the piece, compare it, and half-decide on it from three cities away.

The LVMH maison and salesperson counts that circulate in trade coverage do not trace to any current primary deployment source.

Where the gap actually costs money

The 90-percent citation-share, Salesforce personalization gap, McKinsey expectation, and 40-percent revenue claims used different datasets and were not verified here as one luxury-specific causal chain. Stacked together they look like a luxury-specific causal chain. They are not one.

Personalization expectations back this up from the demand side: Salesforce research puts the share of customers expecting brands to treat them as individuals at 73%, while only a third of brands currently meet that bar. McKinsey's figure is 71% of consumers expecting personalized interactions, and companies that deliver it generate 40% more revenue than peers that don't.

Forty percent more revenue is not a rounding error in a category where margin is the business model. Luxury has always sold itself on the promise of being known, the tailor who remembers your inseam, the sommelier who remembers your last vintage. AI is simply the modern infrastructure for keeping that promise at scale, and clients have already decided they want it whether or not the house has built it yet.

What lagging actually risks

None of this argues for chatbots replacing the concierge, or for rushing a half-built assistant into a flagship. "Quiet tech": LVMH's own phrase for AI that stays invisible while doing the work, is the right instinct. But invisible cannot mean absent. A shopper who has already used AI to research, compare, and shortlist before arriving at the counter does not want to re-explain her preferences to a salesperson working from paper notes and memory. She wants the store to already know what she told the assistant an hour earlier.

  • The defensible conclusion is that Bain’s surveyed top-tier customers reported faster AI adoption than many surveyed houses reported customer-facing deployment. That gap is a research finding within the report, not proof of lost revenue for every brand.12
  • 90% of AI-cited luxury shopping links point away from the brand's own site.
  • Only 9% of houses have advisor-grade AI live at scale with measurable results.

The brands closing that gap in the next eighteen months will not be the ones with the flashiest AI Factory press release. They will be the ones whose clients cannot tell where the concierge's memory ends and the algorithm's begins, and never have to ask.