For the better part of three years, retail AI has been primarily a backend and digital phenomenon: demand forecasting, product page optimization, chatbot customer service, personalized email sequences. The consumer shopping experience inside a physical store — the browse, the question, the try-on, the checkout — remained largely untouched.

That's starting to change, according to a new analysis from Modern Retail tracking how specialty retailers are deploying AI inside locations. The implementations range from kiosks to QR codes to smart fitting room screens — and they reveal something important: retailers are experimenting broadly with in-store AI, but there's no consensus yet on what consumers actually want from it.

What's Being Deployed

The Vitamin Shoppe launched a "Shoppe Advisor" touchscreen in its Upper East Side New York location. The screen gives shoppers access to product information, wellness articles, and real-time data on in-store and online inventory. It's positioned as a supplement to associate knowledge rather than a replacement — useful for shoppers who want to do their own research in the moment without hunting for someone to help them.

Guitar Center took a different approach with Rig Advisor, an AI shopping assistant that customers access by scanning a QR code on their phone. The tool delivers product recommendations filtered by what's actually available in that specific store's inventory — solving a frustration that's endemic to large-format specialty retail, where the product you want is often available "at another location."

Victoria's Secret, Under Armour, and Foot Locker are piloting smart fitting room screens built on Crave Retail's platform, running on Zebra Technologies hardware. The screens let shoppers request different sizes or styling options without leaving the fitting room — reducing the friction that kills conversion when a customer can't find what they need and gives up.

Walmart is experimenting differently, adding AI-generated audio summaries to product pages in its mobile app for more than 1,000 beauty products. The feature delivers "short, conversational soundbites" that help customers compare items — essentially replicating what a knowledgeable associate might say, but at scale, in the digital-physical hybrid space shoppers increasingly inhabit even while standing in a store aisle.

The Consumer Acceptance Problem

Adoption of AI assistance in shopping remains significantly below what the technology's advocates assumed. According to Gartner research, approximately 44 percent of consumers say they accept AI assistance for tasks like product research and reordering — which means 56 percent don't.

That's the core challenge. Retailers are deploying AI in stores at a moment when consumer trust in AI tools is unevenly distributed, with acceptance skewed toward digitally native younger shoppers and highly engaged loyalty members. The shopper who most needs an AI advisor — someone unfamiliar with the category, overwhelmed by options, hesitant to ask an associate — may be the least likely to engage with a touchscreen kiosk.

The Verizon/Incisiv 2026 Connected Retail Experience Study found that while 83 percent of retailers say AI is a competitive necessity, only 6 percent rate their AI capabilities as "mature." The gap between ambition and execution is wide, and in-store is where that gap becomes most visible — because unlike a product recommendation algorithm that the customer never sees, an empty kiosk or a confusing QR code flow makes the failure visible and immediate.

What's Actually Working

The strongest signal from the current wave of in-store AI deployments is specificity. The Guitar Center use case — AI recommendations filtered to local inventory — solves a problem that shoppers have actually felt. The Vitamin Shoppe advisor succeeds in a context where shoppers already expect to be overwhelmed with supplement choices and want permission to do their own research without sales pressure.

The broader lesson is that general-purpose AI assistants placed in a retail context without a clear pain point are struggling to drive engagement. The winning deployments are solving discrete problems: "I need a different size and I don't want to get dressed again." "I want to know if this guitar is actually in stock here before I ask someone." "I want to compare these two protein powders without a five-minute conversation."

CNBC's analysis of AI startups targeting virtual try-on notes that the generative AI improvements of the past 18 months have made these applications "good enough to meaningfully impact retailers' bottom lines" in a way that earlier attempts — clunky AR overlays, poorly-lit product visualizations — never were. The fitting room screen that lets you request a size without leaving is a version of that same idea applied to in-store logistics.

The Associate Equation

One tension worth naming: retailers deploying AI associates often face pushback from store teams who view the technology as a signal about their future. The data so far suggests the best outcomes come from AI that makes associates more effective rather than routes around them — giving them answers faster, helping them manage fitting room queues, flagging reorder needs before a shelf gaps.

Retailers that frame in-store AI as "associate augmentation" rather than "labor reduction" are seeing better internal adoption and, by extension, better implementation quality. That framing matters: a kiosk that sits unused because associates don't support it delivers no ROI.

The physical store remains the majority of retail in 2026 — by share of revenue, by consumer preference, and by the depth of relationship it can build with a shopper. The question isn't whether AI belongs there. It's whether retailers can find implementations specific enough, and trained staff confident enough, to make the technology work in the messy reality of a busy store floor.