TL;DR
- The thesis: the houses that are truly advancing on AI are not talking about it. They are building the infrastructure and letting the results speak.
- Brunello Cucinelli credits its AI e-commerce platform Callimacus with a significant contribution to its Q1 growth (+14% CER).
- LVMH formalizes its Digital Product Passport as the data backbone of the product lifecycle. The European DPP registry takes effect on July 19, 2026.
- Hermes opens its 25th leather goods workshop, Cartier launches a watchmaking fellowship. Artisanal infrastructure as counter-investment.
- Dolce & Gabbana appoints Stefano Cantino as Co-CEO to lead its transformation.
PARADOX OF THE WEEK
Houses are investing massively this week. In product data. In AI-driven e-commerce. In training artisans. Three directions, one common thread: zero mentions of agentic AI in any of the press releases. AI is building the foundations in silence. Craftsmanship occupies the storefront.
The Excellence invariant is at stake, but not where you would usually look for it. Excellence is traditionally measured by the finished product: the leather, the stitching, the watch movement. What this week reveals is that the excellence of 2030 will also be determined by what remains invisible: the data layer that proves provenance, the algorithm that adapts the welcome, the program that transmits the craft. The surface stays artisanal. The foundations are going digital. And no one is talking about it.
Sources: Brunello Cucinelli IR Q1 2026 ; LVMH DPP ; Jing Daily — Hermes ; The King's Foundation — Cartier
WHAT'S MOVING
LVMH / Louis Vuitton / Prada / Gucci: the DPP, pre-agentic infrastructure
On April 7, LVMH updates its Digital Product Passport (DPP: a digital identity card tracing the complete life of a physical product, from raw material to recycling) page. The system, deployed across several pilot Houses and backed by Aura Blockchain Consortium (a blockchain traceability consortium founded by LVMH, Prada Group, Richemont/Cartier, and OTB), documents every stage of the lifecycle: origin, production, use, repair, end of life. On April 8, Louis Vuitton publishes "Regeneration 2030": 98% of materials certified or recycled in 2025 (up from 52% in 2020), with a target of 100% of strategic supply chains under dedicated traceability by 2030. Meanwhile, Wanted in Rome notes that the European DPP registry enters into force on July 19, 2026. Prada and Gucci are positioning the DPP combined with Made in Italy as a strategic differentiator.
Sources: LVMH DPP ; FashionUnited — LV Regeneration 2030 ; Wanted in Rome
Hermes and Cartier: artisanal infrastructure as positioning
On April 10, Axel Dumas, executive chairman of Hermes, inaugurates the house's 25th leather goods workshop in Loupes (Gironde): 260 artisans, producing the Kelly, Constance, and Bride de Jour bags. All Hermes leather goods workshops remain in France. No offshoring, no automation. Three days earlier, Cartier (Richemont) and The King's Foundation announce a three-year fellowship dedicated to decorative watchmaking techniques: champleve enamel, grisaille, marquetry. A 7-month program, split between Scotland and the Cartier Maison des Metiers d'Art in La Chaux-de-Fonds. The announcement was released one week before Watches & Wonders. The message: watchmaking excellence cannot be automated.
Sources: Jing Daily — Hermes ; The King's Foundation — Cartier ; FashionUnited — Cartier
DEEP DIVE
Infrastructure decides everything
The AI battle in luxury is not playing out in visible agents. Not this week, anyway. It is playing out in the foundations each house is laying, or failing to lay, right now. I am convinced that the real dividing line between houses, three years from now, will not be "who has the best chatbot" but "who has the best infrastructure."
Three types of infrastructure coexist in luxury. None mentions agentic AI. All make it possible. But to understand why this infrastructure question has become existential, we first need to look at what is happening outside of luxury.
The tech signal of the week: Block buries hierarchy
On March 31, Jack Dorsey (co-founder of Twitter, CEO of Block, the fintech group that operates Square, Cash App, Afterpay, and TIDAL) publishes with Roelof Botha (managing partner of Sequoia Capital, the fund that backed Apple, Google, Airbnb) a 2,500-word text titled "From Hierarchy to Intelligence." The thesis: the hierarchical organization, invented by the Roman army 2,000 years ago and imported into business by American railroads in the 19th century, can be replaced by a coordination AI.
Not "augmented." Replaced. Dorsey is not talking about giving each manager an AI copilot. He is talking about eliminating the middle management layer whose function is to route information upward and decisions downward, and replacing it with two AI models: a Company World Model (everything the company does, builds, decides, blocks, in real time) and a Customer Signal Model (everything the customer does, not what they say). "Money is the most honest signal in the world. People lie on surveys. They ignore ads. They abandon carts. But when they spend, that's the truth."
The resulting architecture is radical. No more product managers hypothesizing the next features. An Intelligence Layer that composes solutions in real time from base building blocks (payment, lending, cards, payroll). When this layer attempts to compose a solution and fails because the block does not exist, the failure itself becomes the roadmap. No more managers routing information between floors: a "world model" that gives every person at "the edge" (the point of contact with reality) the context needed to act without waiting.
And Dorsey's closing question, the one that resonates loudest for luxury: "What does your company understand that is genuinely hard to understand, and is that understanding getting deeper every day?" If the answer is "nothing," AI is just a cost-cutting story. If the answer is "something profound," AI does not complete the company. It reveals what the company truly is.
What this changes for luxury houses
Let us transpose. What does a luxury house understand that is "genuinely hard to understand"? The craftsmanship of its artisans. The unwritten codes of its artistic direction. The intimate relationship with its VICs (Very Important Clients). The tension between scarcity and desire. The memory of 150 years of aesthetic decisions. These are deep, tacit assets, accumulated over decades. Exactly the type of "company world model" Dorsey describes.
The problem: in most houses, this understanding lives in people's heads. In the artisan's notebooks. In the retail director's memory. In the customer service team's Excel files. It does not circulate. It does not compose. It does not aggregate into operational intelligence. This is precisely what the three types of infrastructure observed this week are beginning to solve.
Product-data infrastructure (see WHAT'S MOVING) is the first step toward a luxury "company world model." The DPP documents what the house knows about each product. Agentic commerce protocols (Google Shopping Graph, the product knowledge graph that powers Google's shopping agents, or the OpenUSD standards, the universal 3D scene description format adopted by NVIDIA and Apple) are designed to ingest this structured data. The DPP was not designed for AI agents. But it is becoming their fuel.
AI e-commerce infrastructure (see THE STORY) is a "customer signal model" in action. Callimacus does not track pages: it tracks intent. Every interaction enriches the customer understanding model. Dorsey would say: "the richer the signal, the better the model." The structural question remains: how many houses have the resources to build their own AI e-commerce infrastructure rather than using Shopify or Salesforce Commerce Cloud? My estimate: about ten. The rest will have to choose their dependency.
Artisanal infrastructure (see WHAT'S MOVING) is what Dorsey calls "the edge": the point of contact with reality that AI cannot reach alone. "People sense things models cannot perceive: intuition, opinionated direction, cultural context, trust dynamics, the feeling in a room." A Hermes leather craftsman, a Cartier enameler, a Cucinelli sales advisor: these are luxury's "edge workers." Investing in their training means investing in the asset that AI can only complement, never replace.
The building blocks arrive: Claude Managed Agents
On April 8, Anthropic (the company behind Claude, valued at $61 billion, founded by Dario and Daniela Amodei, former OpenAI executives) launches Claude Managed Agents in beta: a cloud infrastructure to deploy autonomous AI agents at scale, with secure sandboxing (an isolated environment that prevents the agent from accessing unauthorized data or systems), credential management, granular permissions, and full traceability of every action. Notion, Rakuten, and Asana are already clients.
The signal for luxury: the building blocks for deploying secure autonomous agents now exist as a managed service. A house no longer needs to build the entire agentic stack in-house. It can define an agent in natural language or via a configuration file, set its guardrails (action limits, accessible data, scope of intervention), and deploy it on Anthropic's infrastructure. The agent can run for hours autonomously, with full traceability of every decision.
For a luxury house managing sensitive VIC data (purchase histories of high-net-worth individuals, life events, personal preferences), the promise of granular permissions and sandboxing is relevant. The EU AI Act Article 50 (deadline August 2, 2026) requires transparency for public-facing AI systems: the native traceability of Managed Agents could simplify compliance.
Jean-Noel Kapferer, in The Luxury Strategy (2009), stated a principle I like to recall: in luxury, a brand's value depends more on the quality of its image than on its recognition. Transposed to the agentic era: when an agent recommends a brand, it does not recommend an "image" in the marketing sense. It recommends a product whose attributes it can verify. Image is built on proof, not on promise. And it is infrastructure that provides the proof.
Block is reinventing the organization. Anthropic is providing the building blocks. Luxury houses are sitting on exactly the type of deep understanding Dorsey describes. The question is no longer "should we adopt agentic AI?" It is: "is your infrastructure ready to receive it?" This week suggests that the largest houses are laying the foundations. Silence is not inaction. It is construction.
Sources: Block — From Hierarchy to Intelligence ; Anthropic — Claude Managed Agents ; SiliconANGLE ; Google Shopping Graph ; OpenUSD ; Jean-Noel Kapferer, The Luxury Strategy (2009)
THE STORY
Brunello Cucinelli x Callimacus: when AI is the site
The problem was as old as luxury e-commerce itself. How do you welcome a customer online with the same attention as in the Solomeo boutique? Pages, dropdown menus, category filters -- all of it reproduces the architecture of a supermarket, not of a house that spends an hour serving an espresso before talking about a sweater.
Brunello Cucinelli, an Italian house founded in 1978 and based in Solomeo (Umbria), chose the most radical path: eliminate pages. Callimacus, developed by Solomei AI (an internal R&D center created to "explore the languages of AI and their creative, scientific, and technological applications"), is an adaptive system that interprets each visitor's intent in real time. No pre-built navigation. The visitor can ask "What can I wear to a dinner in New York?" and the system responds with suggestions, products, content -- an experience that builds around desire, not around the catalog.
The Q1 2026 results (published April 9) give this approach a quantified foundation: revenue of 369.1 million euros, up 14% at constant exchange rates. Retail grew 20.1% (Americas +20.3%, Asia +17.8%). Online sessions are "significantly longer," which Riccardo Stefanelli, co-CEO, interprets as "greater opportunities for sales and communication, as well as a more elevated and distinctive image."
And the founder adds an unexpected piece of information: Callimacus has attracted the attention of "major tech companies from Silicon Valley." When a 3,000-person cashmere house in Umbria builds an e-commerce infrastructure that interests Silicon Valley, it says something about the state of the industry. AI innovation in luxury does not necessarily come from tech. It comes from the obsession with hospitality.
Three questions:
- Callimacus is proprietary. Will it remain a lasting competitive advantage, or will Cucinelli end up licensing it to other houses, becoming an unwitting tech vendor?
- The "Human Artificial Intelligence" approach holds because the founder embodies it. What becomes of it when Brunello Cucinelli is no longer at the helm?
- If Callimacus attracts Silicon Valley, who will end up selling what to whom?
Sources: Brunello Cucinelli IR ; MarketScreener ; Fibre2Fashion ; Digital Commerce 360 ; Il Sole 24 Ore
MY INDISCREET QUESTION
How many mid-sized luxury houses -- the ones that have neither LVMH's data teams nor Cucinelli's R&D budget -- will be ready when the European DPP registry takes effect this summer? And when purchasing agents start comparing products by verifiable traceability, what will they recommend to customers who ask for "a responsibly made bag": the house that can prove it, or the one that promises it?
ON MY READING LIST
- Cognizant — AI in luxury retail: 4 trends to watch in 2026 — Cognizant (a digital services company, 350,000 employees) structures four AI trends for luxury retail. The Swarovski example (2,300 stores orchestrated via AI CX) is the only quantified case. Worth reading for the taxonomy, not for the scoops.
- Wanted in Rome — How Made in Italy and Digital Passports Are Redefining Luxury — The most concrete article I read this week on the DPP x provenance pairing. For Italian houses, traceability is no longer optional: it is "market-decisive."
- Brunello Cucinelli IR — Q1 2026 Results — The primary source on Callimacus. What the founder tells analysts, not the press release, not the journalistic summary. The nuance is in the wording: "deeply human," "welcoming our online guests."
- Reuters — Luxury brands face profits squeeze as Iran conflict shrinks Dubai Mall sales — Reuters scoop from April 13. Luxury sales at the Mall of the Emirates: -30% to -50% in March. The geopolitical shock is real-time testing the pricing and stock allocation infrastructure of the groups.
- Kering — Capital Markets Day, Florence, April 16, 2026 — First public address by Pierre Houles, Chief Digital, AI & IT Officer (appointed March 17, formerly Renault Digital). What he says about the AI budget will determine whether Kering enters the infrastructure race or remains in the discourse.
- Block — From Hierarchy to Intelligence (Jack Dorsey & Roelof Botha) — The most ambitious text I have read this month on organizational transformation through AI. 2,500 words, a 2,000-year fresco of hierarchical organization, and a radical proposal: replace middle management with an AI "world model." Not a techno-utopian manifesto: an operational plan with a four-layer architecture. Essential reading, regardless of your industry.
- Anthropic — Claude Managed Agents (April 8, 2026) — The Anthropic announcement that makes autonomous AI agents deployable as a service. Sandboxing, granular permissions, full traceability, long-running sessions. Notion, Rakuten, Asana already on board. For luxury digital leaders who want to understand what "deploying an agent" concretely means in 2026, this is the technical reference.
- Hypebeast — LVMH Shares See Worst Q1 on Record (-28%) — The financial context in which LVMH's Q1 results (April 15) arrive. The group's worst quarter ever on the stock market. The question for the newsletter: will AI be presented as resilience infrastructure, or will it remain absent from the analyst discourse?
COMING NEXT
The LVMH stock has just closed its worst first quarter in its history: -28%, worse than 2008, worse than during COVID. Tuesday, Q1 results drop. Morgan Stanley anticipates -1.5% in Fashion & Leather Goods. The question I will ask when opening the transcript: will AI be mentioned to analysts as a resilience lever, as Cucinelli did with Callimacus? Or will it stay in the shadows while the discourse focuses on Donald Trump's tariffs and the war in the Gulf?
In parallel: Watches & Wonders Geneva is in full swing (April 14-20), and I will be watching for tech announcements from the 66 brands and the 15 startups in the LAB. Wednesday, Kering CMD in Florence, Pierre Houles' first public appearance. Thursday, Hermes publishes its Q1. Next week, I will break down what these events revealed about the real state of AI infrastructure in luxury. The announcement vs. the proof. Luxe oblige!