The technology is ready.
Now the question is whether shoppers — and brands — will actually use it.
1. Virtual Try-On Is Getting Good. Shoppers Just Aren’t Using It Yet.
Generative AI has quietly solved most of the hard technical problems with virtual try-on.
Fabric drape, texture, how a cut sits on different body types —
the latest models handle all of it with photorealistic accuracy across a wide range of body sizes.
The rollout is broad.
Google now shows virtual try-on across billions of product listings,
letting shoppers upload a selfie to preview clothes on themselves.
Zalando is preparing a full customer rollout in 2026 after internal testing showed a 40% reduction in returns.
Nvidia and Antoine Arnault have partnered on RealFit,
a physics-based try-on platform targeting luxury retail.
Startups like Doji, Zelig, and Swap are each carving out niches in the same space.
The problem is adoption, not technology.
An October 2025 eMarketer survey of around 1,000 adults found only 1.4% regularly use virtual try-on.
Most shoppers don’t know it exists,
and many who do don’t trust it enough to replace actually holding or wearing a garment.
“Generative AI reduced a lot of the barriers for businesses to innovate,” said Reza Shirvany, Zalando’s Director of Applied Science.
The harder barrier now is consumer behavior.
The takeaway: If you sell online,
virtual try-on is no longer a niche experiment — it is becoming table stakes.
But adoption won’t happen passively.
Brands that explain the feature, make it visible in the shopping flow,
and pair it with strong product imagery are the ones likely to see actual usage lift.
2. ChatGPT Tried to Be a Store. Now It Just Wants to Be Your Shopping Assistant.
In September 2025, OpenAI launched “Instant Checkout” —
a feature that let users buy products and book hotels directly inside ChatGPT.
By March 2026, they quietly shut it down.
The reason isn’t hard to guess.
Running e-commerce means handling fraud, processing refunds,
managing shipping complaints, and fielding customer service.
None of that is what an AI lab wants to be doing.
So OpenAI is pivoting: ChatGPT now acts as a discovery and recommendation engine,
and when a shopper is ready to buy,
it routes them to the merchant’s own app or website.
Partners include Shopify, Etsy, Stripe, and PayPal on the processing side.
The underlying numbers still make this a story worth watching.
According to HubSpot research,
55% of consumers have already used ChatGPT for product recommendations.
Among those who tried it, 70% said they preferred ChatGPT’s suggestions over traditional search results.
Shopping-related searches on generative AI platforms grew 4,700% between 2024 and 2025.
The catch: a LocalFalcon analysis found roughly 83% of businesses lack any visibility on ChatGPT, compared to 14% on Google Search.
If you are not showing up in AI search results,
you are essentially invisible to a growing segment of discovery traffic.
The takeaway: ChatGPT is positioning itself as the new top of the shopping funnel — the place where consumers decide what to consider, before they go anywhere to buy.
Whether or not your products appear in those recommendations depends on how well your product descriptions, metadata, and imagery communicate what you sell.
Treating AI search visibility the same way you treat Google SEO is quickly becoming a real business priority.
3. Korean Fashion Brands Are Building AI Models — Not Just Using AI Tools
The conversation about AI in Korean fashion shifted noticeably this week,
from tool adoption to brand identity.
Shinsegae Tomboy’s contemporary brand VOV (VOICE OF VOICES) unveiled “Vittoria,” an AI-realized photorealistic model created through a collaboration with Hong Kong-born illustrator Stella Luna.
The character — drawn first by hand and then brought to life through AI —
represents the bold, individualistic woman the brand wants to embody post-rebranding.
Vittoria will appear in campaigns, animated content,
and a dedicated product collection,
with pop-up stores planned across major department stores starting next month.
VOV is not alone.
GS Shop deployed AI models for its fashion brand Buntroye,
using them to quickly generate diverse poses and looks for campaign imagery.
Musinsa is testing AI image generation for product detail pages,
generating size-specific try-on images and styling variations automatically.
LF Mall introduced a generative AI system that auto-generates product descriptions (reportedly 10x faster than manual MD work) and a styling recommendation engine trained on 230,000 curated outfit combinations.
Wellmade, part of the Sejung Group,
is rolling out AI-generated photography across its menswear lines on a monthly basis.
An industry spokesperson summed up the direction: “AI is no longer a simple work assistance tool.
It has become core infrastructure connecting design planning through logistics and customer service.”
The takeaway: Korean brands are moving faster than many international counterparts on one specific thing — committing AI-generated imagery to actual campaigns and storefronts, not just testing in the background.
The brands making this work share a common denominator: strong base product photography to feed into the AI systems.
What This Means for Your Brand
This week’s news points toward one consistent theme: AI is reshaping the gap between content production and customer trust.
-
Virtual try-on is a visibility feature,
not just a utility. Shoppers who see it, understand what it does,
and trust it enough to try it are more likely to convert and return less.
Getting it in front of people is now the hard part. -
AI search is becoming a real discovery channel. If ChatGPT is where 55% of consumers are now looking for product recommendations, being absent from that surface has real traffic consequences.
Clear product descriptions, good imagery,
and context-rich metadata are the foundation. -
AI-generated content is becoming campaign-ready. Korean brands this week demonstrated that AI imagery is moving out of the testing phase and into live campaigns, storefronts, and brand identity work.
The quality threshold has shifted.
For fashion brands producing content at scale — seasonal lookbooks,
product detail images, styling variations —
AI generation tools like LaonGEN can help close the gap between what your visual team can produce and what your product catalog actually needs.
Sources: Generative AI Is Revolutionising Virtual Try-On.
Will Shoppers Use It?
— BoF, OpenAI shifts ChatGPT shopping strategy toward recommendations — Communicate Online, 패션가는 지금 ‘AI 모델’ 전성시대 — 뉴데일리, AI 모델 내세우고 쇼핑 목록 추천도…패션업계 ‘AI 경영’ 속도 — 뉴시스