Fashion bosses sell AI to investors, but few will put a number on it

In the third quarter of 2025, the term “AI” was cited on 306 earnings calls of S&P 500 companies, according to FactSet data reported by Fortune. That was the highest count for any quarter in ten years and more than double the five-year average of 136. Fashion has joined the chorus. When Zalando presented its 2025 results on 12 March 2026, its headline promised “further acceleration in 2026 through scaling AI innovations”, and its shares jumped 7% in early trading, Reuters reported.

The question for investors, competitors and the industry's own staff is what lies behind the words. We went through results releases, earnings call transcripts, annual reports and capital markets presentations from Zalando, Inditex, H&M, Kering, LVMH, Ralph Lauren, Levi Strauss, ASOS and Nike covering 2025 and 2026. The answer is consistent: executives describe customer-facing tools and productivity programmes in growing detail, a few report usage figures, and only one, Zalando, attaches operating metrics to its AI work. Not one of them breaks out what it spends on AI, what AI has saved in currency, or how many jobs it has changed.

That gap matters now because AI is moving from the back office into how customers find products. Reuters noted that analysts have grown concerned that shoppers could use large language models such as ChatGPT to research and buy products, bypassing Zalando's platform. When a technology threatens the channel itself, investors need more than adjectives.

Which fashion company gives investors the most concrete AI numbers?

Zalando, by a wide margin. Its full-year release of 12 March 2026 reads in places like an AI operating report. The company said it scaled AI-generated product content from near zero to 90% in one year, cut campaign creation time from six weeks to days and raised content output by 70%. AI models improved the precision of real-time delivery promises by 22 percentage points. About 3,000 tech specialists use AI coding tools, which the company said increased code changes by more than 20%.

On the customer side, Zalando said its matchmaking, powered by its own foundation models, drove a 13% increase in items added to bags, and its Size and Fit AI, which draws on body measurements from more than one million customers, cut size-related returns by more than 8%. The Zalando Assistant reached six million users, four times the level a year earlier, and an AI-powered discovery feed reaches more than nine million weekly users. Zalando also said it is one of two European launch partners for Google's Universal Commerce Protocol, which lets customers buy through AI chatbots such as Gemini.

Even here, the numbers have limits. They are company claims, not audited figures, and they describe operating effects rather than money. The release gives 2025 revenue of EUR 12.3 billion, adjusted EBIT of EUR 591 million and 2026 EBIT guidance of EUR 660 to 740 million, but it does not say how much of that improvement AI contributed, what the AI programme cost, or whether the content shift changed staffing in creative production. Reuters found no headcount or AI cost figures in the announcement.

What did Inditex and H&M tell investors about AI?

The two largest fast-fashion groups took different routes. At its 2025 results presentation in March 2026, Inditex chief executive Óscar García Maceiras framed AI as part of innovation, saying it was key “to continuing to use the latest technologies, which includes artificial intelligence” (quoted in translation by FashionUnited). The concrete example was Zara's “Try-on” function, launched in mid-December 2025, which lets customers create virtual avatars to try on products. By the close of the financial year on 31 January 2026, it was active in 43 markets and had generated more than 7 million sessions, and Inditex said it was being deployed in the group's other formats.

That is an adoption metric. Inditex gave no conversion, return-rate or sales effect for the tool. Its investment guidance was equally broad: ordinary investments of around EUR 2.3 billion in 2026, focused on the quality of its commercial presence in stores and online, with technology one of several destinations rather than a separate line.

H&M was more explicit about intent and more modest about results. Chief executive Daniel Ervér said the group wants to “leverage artificial intelligence” to forecast demand much more accurately, down to each stock node, and that AI tools for design teams “can really enhance their creativity on a large scale”, Modaes reported from the 2025 results presentation. H&M guided capex of SEK 9 to 10 billion for its 2026 financial year, mainly for the store portfolio and technology infrastructure. Asked in April 2026 about AI shopping assistants such as the one launched by Gap, Ervér told the Swedish news agency TT that only a very small part of H&M's traffic currently arrives that way, and added: “The honest answer is that I don't know, but it is something we are exploring” (our translation, via EFN).

How are Ralph Lauren, Levi's and ASOS using AI with customers?

The American brands have put conversational shopping tools at the centre of their AI story. Ralph Lauren launched Ask Ralph, an AI styling assistant built with Microsoft on Azure OpenAI, in its US app in September 2025. Its November 2025 results release listed the launch among key achievements, and on the February 2026 call chief executive Patrice Louvet said the assistant “is providing us with powerful insights”. He told analysts that customers are “moving beyond traditional search toward rich, natural-language product conversations”, with styling and outfit discovery accounting for more than 50% of engagement, and called the tool an important source of first-party data, according to Digital Commerce 360.

Ralph Lauren is also one of the few groups to link AI to an operating process with a measure attached. On the May 2025 call, Louvet said the company had integrated predictive buying across 25% of its international direct-to-consumer businesses, and chief financial officer Justin Picicci said the company was “increasingly leveraging AI and analytics to enable even more efficient inventory planning”. Picicci put capital expenditure, including technology and AI, at about 4% to 5% of sales. AI's share of that figure was not given.

Levi Strauss followed a similar script. On its fourth-quarter call in January 2026, chief executive Michelle Gass said the company would launch “a new consumer-facing AI stylist chatbot” and was testing an internal “super agent” built with Microsoft, which it planned to roll out to employees in 2026, Digital Commerce 360 reported. An Outfitting feature in the Levi's app, built by an internal team, went live in November 2025. Gass described the aim of the internal tools as giving teams faster access to information and insights, without quantifying the benefit.

ASOS used AI in a turnaround context. With full-year results on 21 November 2025 showing sales down 12% and a narrower pre-tax loss of GBP 281.6 million, the company highlighted a trial called “Styled for You”, an AI model trained on more than 100,000 curated studio outfits that recommends looks to shoppers. Shares fell 8% that afternoon, PA Media reported, a reminder that AI features do not offset weak trading in investors' eyes.

What do luxury groups say about AI to their shareholders?

Luxury has traditionally said little. A Business of Fashion analysis of call transcripts from more than 30 fashion and retail companies, published in 2023, found no mentions of AI or machine learning in LVMH's transcripts, even after the group's 2021 AI partnership with Google, and noted that Kering discussed AI for customer targeting and forecasting in 2019 and 2020 and then stopped. The same analysis showed how long some American groups have used the vocabulary: Tapestry cited AI and machine learning for inventory, merchandising and pricing as early as 2018, and Levi's had run an internal AI bootcamp for employees. That has changed in tone across the industry, if not in numbers.

At its capital markets day on 16 April 2026, Kering under chief executive Luca de Meo set out a “client hub” that consolidates proprietary and external data into a unified client base “powered by AI”, and a technology hub built on cloud-native systems, agentic AI and digital twins. The group, which reported 2025 revenue of EUR 14.7 billion and 44,000 employees, said it would reinvest 5% to 6% of revenue to support organic growth. It gave no quantified savings target linked to AI and no headcount plan.

LVMH addressed AI in its 2025 annual report letter rather than in a programme with targets. Group managing director Stéphane Bianchi wrote that new AI models will create opportunities for the maisons “provided we understand how AI fits into our customers lives”, and said the group wants to use AI to boost operational productivity, stimulate the creativity of its teams and rethink its processes. The letter contains no AI figures.

Which fashion groups are staying quiet about AI, and why?

Silence is also a signal. In the transcript of Nike's fiscal 2025 fourth-quarter call in June 2025, as published by MarketBeat, we found no mention of AI, artificial intelligence or machine learning by chief executive Elliott Hill or chief financial officer Matt Friend. Hill's references to technology concerned Nike Air and investment in Nike Direct. For a company in the middle of a turnaround focused on product, wholesale partners and inventory, that choice of emphasis is coherent.

Other groups talk about AI in general terms because the investment is hard to separate. When AI is a feature inside a planning system, a content workflow or an e-commerce platform, the spend sits in technology capex and operating costs that were already there. That explains why Ralph Lauren, Kering, H&M and Inditex all fold AI into broader investment envelopes. It does not explain why so few report the operating effects that Zalando shows are measurable.

Where are the numbers on AI savings, headcount and capex?

Set side by side, the disclosures fall into three tiers. Zalando reports operating KPIs. Inditex and Ralph Lauren report usage or process coverage. Everyone else reports intent. The table summarises what each company actually put on the record.

The last column is empty for every company. No group reviewed disclosed a separate AI budget, a saving in euros, pounds or dollars attributed to AI, or a number of roles created, cut or redeployed because of it. The closest proxies are capex envelopes such as Kering's 5% to 6% of revenue or Ralph Lauren's 4% to 5% of sales, which bundle AI with stores, systems and infrastructure.

What are the risks and limits of AI claims on earnings calls?

The first risk is comparability. Zara's 7 million try-on sessions, Zalando's six million assistant users and Ralph Lauren's 50% share of styling conversations are all real disclosures, but they measure different things and none of them maps directly to revenue or margin. A session count can rise while conversion falls.

The second is selection. Companies choose which metrics to publish, and they publish those that improve. Zalando's 8% fall in size-related returns is meaningful, but there is no external benchmark and no disclosure of where results disappointed. All figures in this article are company statements unless stated otherwise; none has been independently audited as an AI effect.

The third is that the narrative can run ahead of trading. ASOS presented an AI styling trial alongside a 12% sales decline, and its shares still fell 8%. Zalando's shares rose on guidance, not on AI metrics alone. Investors appear to reward AI talk only when it arrives with results.

There is also a fair counter-argument for discretion. AI capabilities in demand forecasting, pricing or design workflows can be a competitive advantage, and detailed disclosure helps rivals. Isolating the cost of AI from the platforms it runs on is genuinely difficult. And H&M's admission that it does not yet know how AI assistants will affect its traffic is arguably more useful to shareholders than a confident forecast.

What should fashion companies do now?

1. Pick three AI metrics and keep them Choose operating measures such as content cost per item, forecast error or size-related returns, and report them every quarter so investors can see the trend.
2. Separate adoption from impact Report usage figures like sessions or users, but always pair them with an outcome such as conversion, return rate or margin.
3. Disclose an AI spending range Even a range within the capex envelope helps investors judge whether claimed benefits are proportionate to the investment.
4. Be explicit about jobs State whether AI is changing roles in content, customer service or planning, and how many people have been retrained or redeployed.
5. Prepare for agentic commerce questions Analysts already ask whether chatbots will bypass retailers; have a view on traffic from AI channels and how it converts.
6. Align the narrative with trading Avoid presenting AI features as a substitute for sales performance; investors discount them when the core numbers disappoint.