AI shopping is moving fast, but brands still do not control how their products are interpreted

For years, digital commerce was built around a simple assumption: brands controlled their product pages, retailers controlled the storefront, and search engines helped shoppers find both.

That assumption is breaking.

AI shopping is moving quickly from experimentation into infrastructure. Target is already building integrations so shoppers can buy products through ChatGPT and Google’s Gemini, while major retailers are putting real money behind AI-enabled shopping and merchandising workflows. Zalando has publicly credited AI with helping drive profitability through things like AI-generated product imagery and virtual try-ons. In other words, this is no longer theoretical. AI is becoming part of how products are discovered, evaluated, and purchased.

But there is a major problem hiding underneath all the excitement:

Brands still do not control how their products are interpreted by AI.

And in fashion and beauty, that problem is especially dangerous.

The new bottleneck in commerce is not access. It is interpretation.

A lot of the current conversation around agentic commerce is focused on access and integrations.

Can an AI agent browse a retailer’s site?
Can it add to cart?
Can it complete checkout?
Can it act on behalf of a user?

Those questions matter. But the more important question for brands is this:

When an AI system encounters your products, does it actually understand them the way you want them to be understood?

That is where things get messy.

A dress is not just a dress. It may be occasionwear, travel-friendly, flattering for a certain silhouette, aligned to a specific aesthetic, appropriate for warm weather, and better suited for one customer’s body type than another’s.

A foundation is not just a foundation. It may be dewy, medium-coverage, olive-undertone-friendly, sensitive-skin-safe, wedding-appropriate, flashback-resistant, and compatible with certain skincare routines.

Humans understand this kind of nuance through context. Most product catalogs still do not express it well enough for AI systems to do the same.

Amazon vs. Perplexity exposed the real fight

The recent Amazon-Perplexity legal battle made headlines because it looked like a dispute about whether AI shopping agents should be allowed to act on a shopper’s behalf. A federal judge granted Amazon a temporary injunction blocking Perplexity’s AI shopping tool from using Amazon, after Amazon argued the tool unlawfully accessed customer accounts without permission. Perplexity framed the dispute as one about user choice and whether shoppers should be free to use their preferred AI tools.

On the surface, this is a story about platform control, security, and gatekeeping.

But underneath it is a deeper signal for every brand:

The next commerce battle is not just over who gets to access the shopper. It is over who gets to shape the shopper’s decision.

If AI becomes the interface between consumers and product catalogs, then product interpretation becomes a source of power.

And right now, many brands are underprepared for that shift.

In beauty and fashion, AI interpretation shapes the outcome

This issue matters in every retail category, but fashion and beauty are uniquely exposed because these categories rely on language that is subjective, situational, and highly contextual.

Beauty shoppers ask questions like:

• What foundation works for dry skin with olive undertones?
• What skincare routine should I use if I have acne and sensitivity?
• What lipstick gives a soft bridal look without feeling drying?

Fashion shoppers ask questions like:

• What should I wear to a summer wedding in Spain?
• Which jeans are best for petites with curves?
• What black-tie-optional dress feels modern but not too revealing?

These are not keyword searches. They are intent-rich, context-heavy decision prompts.

And increasingly, people are asking them directly to AI systems.

Business of Fashion recently reported that influencers have been asking ChatGPT to create skincare routines and share the results on TikTok, and that some beauty brands are already showing up more often than others in these AI-generated recommendations. The article noted that keywords alone are not enough; brands also benefit from stronger informational footprints and trusted external mentions.

That should be a wake-up call.

Because if one beauty brand is repeatedly recommended and another is not, the difference may have less to do with product quality than with how legible that brand is to AI.

The risk is not just invisibility. It is misinterpretation.

Most brands still think of the AI-discovery problem as a visibility problem.

How do we show up in ChatGPT?
How do we get cited in Gemini?
How do we appear in AI answers?

Those are the right questions, but they are incomplete.

The deeper risk is not only that your products fail to appear. It is that they appear for the wrong reasons, in the wrong contexts, or with the wrong framing.

A prestige serum may get summarized as “anti-aging skincare” and lose the nuance of texture, use case, layering, and customer fit.

A luxury dress may get reduced to “formal black dress” and lose everything that makes it right for a particular body shape, climate, styling need, or event.

A beauty product may be recommended because of broad popularity, while more suitable niche options never surface because the AI lacks enough structured, customer-relevant product context to distinguish them.

In other words, the problem is not just ranking.

It is representation.

The retailers moving fastest are proving the urgency

The market is not waiting for brands to catch up.

Target’s technology leadership has said the company is enabling shopping via ChatGPT and Google’s AI tools as part of its broader turnaround strategy, alongside AI investments in design, forecasting, and advertising. That is a major signal that large retailers increasingly expect AI-mediated shopping to become part of mainstream commerce behavior.

Zalando has gone even further in showing the commercial upside. Reuters reported that the company expects a strong jump in 2026 profit, driven in part by AI tools including AI-generated product imagery and virtual try-ons that reduce advertising costs and customer returns.

The takeaway is not just that AI is useful.

It is that the brands and retailers building AI-native product experiences now will have a head start in the next phase of discovery.

Why open standards alone will not solve this

There is a growing push toward more formal infrastructure for agentic commerce. That is important. Brands will need standard ways for agents to discover product data, understand merchant policies, and potentially transact.

But standards solve transport, not meaning.

An open protocol can help an AI system access your catalog. It does not guarantee that the catalog is rich enough, structured enough, or customer-aware enough to be interpreted correctly.

That is why so many brands are at risk of over-focusing on enablement while underinvesting in understanding.

It is possible to be technically available to AI and still commercially invisible.

It is possible to be connected to agents and still poorly represented by them.

It is possible to “show up” and still lose.

What brands need to control now

If fashion and beauty brands want more influence over how AI shopping evolves, they need to think beyond access and start investing in product interpretation.

That means building product data that goes beyond basic taxonomy and merchant-side attributes.

It means encoding the kinds of signals shoppers actually use when they ask AI for help:

• occasion
• body-context or fit context
• undertone and finish
• styling compatibility
• fabric behavior
• routine compatibility
• aesthetic language
• intent and use-case wording
• edge cases and exclusions
• comparison context

It also means paying attention to the trust layer around the catalog. The brands surfacing most effectively in AI recommendations are often not just the ones with the biggest media budgets, but the ones whose products are supported by better product detail, stronger educational content, and broader trusted references across the web.

The next winners will not just be searchable. They will be interpretable.

The old model of commerce optimization was about making products searchable.

The emerging model is about making them interpretable by machines that increasingly shape the path to purchase.

That is a fundamentally different challenge.

Search engines indexed pages.

AI systems synthesize meaning.

And when they synthesize meaning from weak, generic, incomplete, or unstructured product information, brands lose control over how they are represented.

That is the real risk in AI shopping.

Not that agents are coming.
Not that platforms are changing.
Not even that the rules are still unsettled.

The real risk is that brands are entering this new era without a strong enough product intelligence layer to influence how their products are understood in the first place.

Where Veristyle fits

At Veristyle, we believe the future of commerce will not be won by brands that simply plug into AI.

It will be won by brands whose products are most understandable to AI in the moments that shape customer decisions.

For fashion and beauty, that requires more than generic SEO, broad taxonomy, or standard feed cleanup.

It requires product enrichment built for nuanced, customer-centric interpretation across AI answers, shopping agents, and on-site discovery experiences.

Because in AI commerce, visibility matters.

But interpretation decides who wins.

AI shopping is moving fast, but brands still do not control how their products are interpreted was originally published in Veristyle on Medium, where people are continuing the conversation by highlighting and responding to this story.