Fifteen dimensions of agent-readiness across the sector, read once a week.
The fashion sector's AI posture this week reads as a widening split between what agents can already do to fashion and what fashion has built to meet them. On the demand side, agentic commerce is here: Retail Dive reports 42% of American millennials would let an AI agent buy on their behalf, and 59% of U.S. consumers credit AI platforms with surfacing brands they had not known (source). Discovery is moving to a machine layer the houses do not own. On the supply side, the sector's own agent-readiness is still thin, which is why Agent-Ready Infrastructure and Content AI-Readiness each rose only modestly off weak bases.
The structural problem is data. Fashion catalogs are visual and attribute-dense, and Pixyle.ai diagnoses them as routinely missing the silhouette, fit, and pattern fields that agents need to reason over a product (source). The Salesforce reference model for agentic commerce runs entirely on schema.org and GS1 structured data (source). Where inventory and pricing are not machine-readable, agents skip the store. That is the fault line the benchmark is watching, and it sits squarely on the Decision Surfaces frontier: the point where the buying decision passes from a human to an agent is being built by platforms, and fashion is a passenger.
At the top of the cohort, the picture is stronger and better documented. LVMH runs MaIA across 40,000+ users at more than two million requests a month, on a Google Cloud platform spanning 75+ Maisons (Vistoya). Kering runs conversational clienteling through KNXT. Richemont is credited with a depth-over-breadth strategy in hard luxury (Klover.ai). Market perception holds at the sector's high mark of 50. The caution is that perception runs ahead of scaled production, and the privately held houses, Chanel and Hermès, disclose almost nothing, so the sector figure rests on high-public-signal groups and should be read as such.
Two warnings sharpened this week. Talent is fragile: Lululemon lost its chief AI officer inside a year, with analysts flagging turnover (Retail Dive). And governance is the sector's weakest link, held at 38. Foley & Lardner lays out the NIL and right-of-publicity exposure created when brands digitally replicate model likenesses without documented consent (source), and Cimplifi notes regulators now want demonstrable controls, not ethics statements. This is where the Identity Control Surface matters: consent tracking and governance for synthetic likenesses are legal obligations the houses have not yet operationalized.
The value case is genuine where it is measured. Global Fashion Group tied AI integration to its first profitable H1, and DressX data shows virtual try-on lifting luxury conversion up to tenfold. The houses that treat AI-readiness as infrastructure, starting with governed, machine-readable data, will compound the advantage as agents take over discovery. The rest will be invisible to the agent doing the shopping.
Lululemon's chief AI and technology officer departed after less than a year, with analysts on record as increasingly concerned about turnover, per Retail Dive. AI leadership benches in apparel are being stood up and losing hold at once, pulling the dimension down one point to 44.
Named adoption of MCP and ACP discovery protocols in fashion commerce, reported by Vistoya, plus a clarified schema.org and GS1 baseline from Salesforce, lift a weak base two points to 36. Agents now mediate discovery, and protocol readiness is becoming concrete.
42% of American millennials would allow AI agents to buy items on their behalf
Consumer willingness to delegate purchasing to agents is now measurable, pulling the interaction layer toward the agentic end of the spectrum.
Where the sector stands on the fifteen-dimension framework. Reference, not the headline. The movement above is the story.
The Agentic Fashion Report scores the fashion vertical on the Applied AI Index 15-dimension enterprise AI-readiness framework (0-100 across Organization, Brand, and Product, five dimensions each, equally weighted within category). Scores derive from public signals: analyst and trade coverage, house communications and filings, and machine-readable commerce evidence. A sector cut of the Applied AI Index methodology.
Produced by Applied Identities using specialized AI analysis. All scores based on publicly available data. Internal deployment data not captured. This analysis is independent and does not represent the views of any company named herein.