How AI Is Rewiring Fashion’s Art And Science—Beyond The Hype
- Angela Chan
- Mar 24
- 5 min read
From Hype to Infrastructure: How AI is Scaling Intuition and Streamlining the Fashion Value Chain
AI has moved beyond hype and into infrastructure, rewriting how fashion operates across design, merchandising, and the supply chain. What was once experimental is now becoming essential, shaping both creative processes and commercial strategy.
While AI accelerates decision-making and surfaces insights at unprecedented speed, it is not replacing creativity. Across C-suites, a clear consensus is emerging: AI is a multiplier, not a substitute. The companies that will win are those that bridge data-driven precision with human instinct.
The Left Brain of Merchandising: Art Meets Analytics
Today’s merchants are seeing AI reshape the balance between art and science. Masako Konishi, Executive Vice President of Merchandising & Buying at ALO Yoga, sees AI as a tool that frees creative leaders from the burden of manual analysis.
“One of the biggest opportunities with AI is shifting merchants’ focus to AI-driven insights, moving away from time-intensive data work that can now be delivered in seconds,” Konishi says.
In her view, AI strengthens the analytical “left brain,” allowing merchants to focus on product sensibility and storytelling. As predictive analytics improve, intuition becomes even more critical. “A ‘gut feeling’ is really pattern recognition, a rapid synthesis of experience,” she notes, emphasizing the importance of experiencing retail “IRL” (in real life). Touch, context, and in-person cultural observation keep intuition grounded and remain irreplaceable inputs.
In this framework, AI enhances intuition, but it does not replace it.
Designing for Heritage: The Human-in-the-Loop
In design, the role of AI is measured against craftsmanship. At Levi’s, assistant designer Kewei (Kyra) Yu notes that heritage categories like denim still depend on physical fittings and wash experimentation.
“AI tools act as analytical partners,” Yu says. “They help us understand data and summarize performance, but creative teams still direct the design.”
AI can identify patterns, but it cannot replicate emotional and cultural nuance. The product’s “soul” still requires human interpretation.
Richard Zhao, a senior menswear executive, describes AI as a “capacity multiplier.” Tasks like creating tech packs, previously a tedious multi-day process, can now be completed within minutes. Designers are shifting from drafting to editing. Zhao notes that while AI can assist with fit and quality for basics, tailored items still require human expertise due to their complexity.
The Next Frontier: From AI Tools to AI Agents
Despite rapid adoption, AI has yet to fundamentally transform how consumers shop. According to Cate Khan, CEO of Trendalytics, most current applications, from virtual try-on to personalized recommendations, are improving convenience, but not changing behavior at a structural level.
“AI is already embedded across the retail experience, but its impact today is incremental rather than transformational,” Khan explains.
She describes the true change as the rise of agentic commerce, where AI shifts from merely aiding decisions to actually making them. Instead of just browsing and filtering, consumers will increasingly depend on systems that understand their intent, curate choices, and handle purchases automatically.
But technology alone is not the differentiator. The real competitive advantage lies in data quality and depth.
“AI is only as good as the data behind it—garbage in, garbage out,” Khan notes. Platforms like Trendalytics differentiate through high-frequency, proprietary data that tracks not just what is trending, but how trends evolve and convert over time.
Not all signals carry equal weight. Khan highlights three critical metrics: volume (trend size), velocity (growth rate), and time to purchase (proximity to transaction). While tools like virtual try-on generate valuable near-term signals, they are more effective for short-term optimization than long-term forecasting.
Looking ahead, this data will play an increasingly important role in shaping not just merchandising, but product creation itself. Consumer interaction data from search to styling will begin to inform design decisions in real time.
Yet even as AI closes the loop between demand and creation, Khan emphasizes that the brands that succeed will be those that balance data with creativity. “Humans feel, laugh, and love, and these emotions foster authentic connections. Since technology cannot emote, it is constrained in its ability to create a brand people truly love,” said Khan.
The VTO Explosion: Turning a Novelty into a Strategic Necessity
Virtual Try-On (VTO) has shifted from a novelty to a fundamental part of retail, tackling one of the industry’s highest costs: purchase uncertainty.
By enabling consumers to visualize products before buying, VTO reduces hesitation and returns. The U.S. VTO market, valued at $4.2 billion in 2024, is projected to reach $24.39 billion by 2034, growing at a 23.4% CAGR. At the same time, McKinsey reports that 44% of companies are already seeing cost and revenue benefits from AI adoption.
AI as the New Entry Point to Commerce
Consumer behavior is changing quickly. According to McKinsey’s The State of Fashion 2026, nearly 25% of shoppers start with generative AI before using a traditional search bar. E-Commerce is transitioning from search-based to intent-based models.
Here are some examples from a new wave of platforms driving this change:
Google Shopping (AI Mode): Realistic try-ons across diverse body types, embedded directly into search
Daydream: Conversational search powered by natural language prompts
ZARA: In-app 3D avatar try-ons with 360-degree visualization
DRESSX AGENT: “AI twins” enabling virtual try-on across thousands of SKUs
Conversion, Engagement, and Returns
The effect is real and has a financial impact.
Higher Conversion: Removing “Will this work for me?” increases purchase confidence and basket size. Macy’s use of AR and VR for furniture visualization, for example, led to larger baskets and fewer returns in pilot programs.
Deeper Engagement: Interactive experiences replace passive browsing. Sephora’s Virtual Artist recorded over 8.5 million try-ons in its first year.
Fewer Returns: By addressing fit uncertainty upfront, Shopify reports that 3D and AR tools can reduce return rates by up to 40%.
According to Adobe Analytics, 85% of U.S. consumers who have used generative AI for shopping say it improved their experience.
The CIO Perspective: Reshaping the Supply Chain Feedback Cycle
AI outfitting technology adds real value by enabling products to be styled and sold as complete looks rather than as separate items. Sahal Laher, Executive Vice President and CIO of Centric Brands, emphasizes that the significant benefits arise when AI combines purchase behavior, inventory management, and margin considerations, beyond just aesthetics. “When those variables come together, you stop leaving attachment opportunities on the floor. The most interesting shift I’m seeing is brands beginning to treat outfitting not just as a merchandising feature, but as a demand signal, and that reframing is where the real commercial value starts to emerge.” Laher says.
AI-cultured organisations, like REVOLVE, are already implementing “Build a Look” features that let customers mix and match items in real time. But the impact extends far beyond merchandising; it fundamentally changes how data is captured and monetized. These tools not only drive higher conversion and increase basket size but also deepen customer engagement and retention by turning shopping into a more interactive, personalized experience.
At the same time, outfitting reduces one of retail’s biggest pain points: returns. By helping customers visualize complete looks and make more confident purchase decisions, brands can minimize mismatched purchases and post-sale friction. More importantly, this shift introduces a new class of forward-looking data. Instead of relying solely on historical sales, retailers can now understand what items are styled together, how consumers build outfits, and why certain combinations resonate. This unlocks more precise merchandising, smarter inventory decisions, and a more predictive approach to demand.
“When a retail partner sells through a complete outfit rather than individual units, the replenishment pattern changes,” Laher notes. This creates a new feedback loop between consumer behavior and product planning, enabling earlier, more confident inventory decisions.
Disclosure: The author serves on the advisory boards of several AI fashion technology companies, including Zelig, which is referenced in this article. The views expressed are independent and reflect the broader industry landscape.
Comments