How artificial intelligence is shaping retail

The retail industry is full of AI adopters, and many are exploring agentic tools even as reliability concerns mount.

More than 3 in 5 leaders in the industry worry about the consistency and quality of generated outputs, according to a Monday.com survey of more than 1,800 retail leaders worldwide. Many retailers are also facing adoption roadblocks, such as privacy and data concerns, integration complexity, risk of alienating customers, high implementation costs and employee pushback. 

Despite the obstacles, retail leaders are confident in the technology's long-term potential. Nearly two-thirds of those surveyed said businesses that fail to adopt AI agents in the next two years risk falling behind. More than half of retailers expect most customer interactions will be handled by AI within five years.

The findings come as the industry has upped its investment in AI this year as economic pressures push business leaders to optimize costs and streamline operations. Retail companies are bracing for slower growth rates for the rest of the year, and many see AI as a lever to mitigate headwinds and improve experiences. 

Under Armour touted the benefits of its AI efforts during the retailer’s Q1 2026 earnings call. 

“AI is becoming more integrated across the business, enhancing design, planning and forecasting,” said Kevin Plank, founder, president and CEO at Under Armour.

To improve operations, the company is moving from siloed functions to collaborative teams with shared KPIs and real-time data, according to Plank. 

“After two years of data and platform development, we have over 80 automations that are streamlining workflows, reducing time to market and improving execution from predictive pricing to real-time inventory management,” Plank said. 

Under Armour is in the midst of a restructuring in response to significant revenue declines. The company reported a 10% revenue decline and a $305 million net loss. 

Advance Auto Parts is also turning to technology to address lackluster performance. While the company lowered its outlook for the year and reported mixed results during a Q2 2025 earnings call, executives highlighted technology’s role in advancing goals around the store experience, such as in its deployment of a new assortment management framework. 

“We have been able to accelerate this rollout by harnessing advanced technological tools, including the use of AI,”  Shane O’Kelly, president, CEO and director at the company, said during the call. “These innovative tools are enabling us to introduce greater intelligence in the assortment planning process, which has traditionally been done manually.”

Advance Auto Parts has implemented the framework in its top 30 designated market areas, and the company is planning to complete the process across its top 50 designated market areas ahead of schedule, which will represent around 70% of its sales by the end of Q3, according to O’Kelly.

Many consumers are interested in AI shopping tools ahead of the holiday season.

About 7 in 10 shoppers say they are open to using generative AI guidance, according to a survey released by Coveo. More than one-third say they have already used it. Respondents’ top use cases for generative AI tools are comparing products, getting budget-based recommendations, and learning more about what they’re buying, according to the survey of 6,000 people across the the U.S., the U.K., Australia and New Zealand.

When customers start shopping for a product or gift, nearly one-third start with a search engine and nearly one-quarter look to Amazon or another marketplace, according to the Coveo study. In comparison, only 3% start with an AI assistant-powered search like ChatGPT.

While consumers express curiosity about generative AI shopping tools, traditional product discovery methods are still the norm.

People are still turning to human-curated sources for recommendations, particularly when it comes to choosing gifts, according to Terra Higginson, a principal research director at Info-Tech Research Group.

“We see that people still rely on curated lists (think Wirecutter) and micro-influencers when it comes to gift-buying decisions,” Higginson said in an email. “The ability to create a connected, valuable, curated and easy journey is still lacking across AI.”

Younger consumers are more likely to be early adopters of AI. Among Gen Z and millennials, 8 in 10 say they’re at least curious about trying a generative AI shopping assistant. Additionally, about 9 in 10 consumers who have used generative AI for product guidance while shopping say they’re at least curious about using it again.

Most AI assistants currently deliver an underwhelming experience in terms of value, according to Higginson. However, she is looking forward to companies putting together AI workflows with the right tech stack to meet consumer expectations.

“If these AI tools want to become more than just shiny widgets, they need to go beyond being tools for content digestion and generation,” Higginson said. “In order to deliver consumers a valuable experience, they must understand previous purchases, preferences, trends and shopping habits to help them move confidently toward a decision.”

Sponsored content

By Mirakl

The holiday shopping landscape is undergoing its most significant transformation since the advent of eCommerce. As consumers gear up for another record-breaking season, a new player is quietly reshaping the entire experience: AI agents. From ChatGPT to Perplexity, these intelligent assistants are becoming the new Santa's helpers, guiding shoppers through increasingly complex purchase decisions with personalized recommendations and instant answers.

This shift represents more than just a technology trend—it signals the dawn of Commerce 3.0, or what we call the agentic commerce era. Just as mobile commerce revolutionized shopping in the 2010s, agentic commerce is poised to fundamentally alter how consumers interact with brands during the most critical shopping period of the year.

The New Holiday Shopping Assistant

Instead of spending hours scrolling through endless product listings, a shopper can now simply ask an AI platform, like ChatGPT or Perplexity, "Help me find the perfect gift for my tech-savvy teenager who loves photography, budget around $300." And in this new reality, ideally the AI agent provides curated recommendations, compares prices across multiple retailers, checks real-time inventory and some can even complete the purchase—all while learning preferences for future interactions.

This isn't science fiction. Recent data shows that 60% of shoppers already use AI in their shopping journey, and this percentage is expected to spike dramatically during the holiday season when decision fatigue peaks, the volume of gift purchasing increases and time pressures mount. The complexity of holiday gift-giving makes AI agents particularly valuable during this period.

Beyond Simple Search: The Intelligence Revolution

What sets agentic commerce apart is the shift from keyword-based searches to intent-driven conversations. Instead of typing "running shoes" and sorting through hundreds of options, consumers can engage in natural dialogue: "I need running shoes for someone training for their first marathon, prefer sustainable brands, size 9, under $150, available for pickup before December 20th."This shift shows up in how people search: Perplexity CEO Aravind Srinivas has noted that the median query on their answer engine is 10–11 words, versus 2–3 on Google.

This conversational approach mirrors how shoppers have always preferred to interact with knowledgeable sales associates. The difference is that AI agents have access to vast product databases, real-time pricing, inventory levels and customer reviews across multiple retailers simultaneously—capabilities no one individual could match.

In the new agentic commerce era, we are heading toward a reality where during the holidays, AI agents could manage complex gift lists, track delivery times across multiple purchases, suggest complementary items and help coordinate surprise deliveries. They will soon become the ultimate holiday shopping concierge.

The Speed of Change

The adoption rate is unprecedented. ChatGPT reached 5.8 billion visits with over 1 billion searches per week, and when it launched search capabilities, it captured 4% of the search market in just one month—matching Bing's decade-long market share overnight.

Major platforms are racing to capture this holiday opportunity. Perplexity introduced direct purchasing capabilities in late 2024, Amazon launched "Buy for Me" agents, and payment providers like Visa, PayPal and Stripe are developing solutions to enable seamless transactions within AI platforms.

What This Means for Holiday Retail

For retailers, this transformation presents both opportunities and challenges. AI agents can drive highly qualified traffic—shoppers who arrive with specific intent and are ready to purchase. Research shows that while conversion rates from AI search are currently slightly lower than traditional search, visitors spend more time on sites, view more pages and have lower bounce rates.

However, success in agentic commerce requires rich, detailed product content that can be easily parsed by AI systems. A simple product title and basic description won't suffice. AI agents favor products with comprehensive information, high-quality images, detailed specifications and positive customer reviews.

During the holiday season, when inventory turns rapidly and promotions change frequently, maintaining this level of data quality becomes even more critical. AI agents need real-time access to accurate pricing, availability and delivery information to provide reliable recommendations to time-pressed holiday shoppers.

The Trust Factor

Agentic commerce is built on trust—trust that the AI agent will find the best options, secure the best prices and ensure timely delivery. This is particularly crucial during the holidays when purchases often involve significant emotional investment and tight deadlines.

The retailers that succeed will be those providing AI agents with reliable, comprehensive data and seamless fulfillment capabilities. It's not enough to be discoverable; businesses must be dependable throughout the entire customer journey.

Looking Ahead

As we move deeper into the holiday season, we're likely to see even more sophisticated AI capabilities emerge. Agent-to-agent transactions, where a shopper's personal AI assistant negotiates directly with a retailer's AI system, are already on the horizon.

For retailers looking to capitalize on this shift, the message is clear: the time to act is now. Success requires investment in rich product data, reliable fulfillment infrastructure and seamless integration capabilities. The retailers that move quickly to optimize for AI discovery will have a significant advantage in capturing holiday sales.

This holiday season marks the beginning of a new chapter in retail history. Those who embrace the change will find themselves at the forefront of the next great commerce revolution.

As one of the most promising, and fear-inducing, technologies of the day, artificial intelligence was a frequent topic of conversation across panels at Shoptalk’s spring conference in Las Vegas. Much like at the National Retail Federation’s Big Show in January, retailers considered the possibilities of the tech, how they go about implementing it and the ways retail remains very human.

Perhaps one of the most succinct statements about retail’s use of AI came from Mandeep Bhatia, senior vice president of global digital product and omnichannel innovation at Tapestry, who was in turn quoting someone else. 

“Humans won’t be replaced by AI. Humans with AI will replace humans without AI,” Bhatia said, a sentiment that was echoed across many panels as retailers considered how the tech could be useful and where human involvement is still very much needed.

Across the industry, retailers are turning to AI to help with customer experience initiatives and to boost loss prevention, while shoppers are turning to the technology to help them with purchases as well. Opinions are thus far mixed, but there’s no doubt that AI is one of the hottest testing grounds in retail right now.

Here’s a sampling of what companies had to say about the technology at Shoptalk Spring — and how they’re using it now.

Toys R Us

At Toys R Us, the brand jumped headfirst into AI with OpenAI’s Sora tool. According to Kim Miller, global chief marketing officer of toys at Toys R Us owner WHP Global, the retailer was the first to make a branded video through Sora. The retailer chose to make an homage to Charles Lazarus, who founded Toys R Us.

Although video storytelling with AI “has a long way to go,” Miller said it’s been improving quickly — and the effort at Toys R Us paid off.

“The video went viral,” Miller said. “A lot of people were asking, ‘Why did you do it?’ And I said, ‘The train has left the station and we wanted to be the first one on it. And we wanted to embrace it.’ We want to always be embracing new ways to tell stories and to reach people so that they are interested in not only the story you're telling, but how you're telling it as well.”

Meta

“All the small guys, startups, they're going — they're on the train first. They're experimenting. They're doing it. Where a lot of bigger brands go, ‘I'm going to wait. Maybe it won't fit my brand DNA. It's scary.’ But it is the ones who take the risk in this moment of gaining competitive advantage [that] will win,” Karin Tracy, group lead of retail and e-commerce at Meta, said. “I feel like so many have gotten over the hump of automation … the next hump is Gen AI for creative.” 

Meta already offers GenAI tools for ad creation, including background and text generation, according to Clara Shih, vice president of business AI at Meta. In one example, Shih said Living Proof saw a 13% higher click-through rate using its tools and 18% higher total purchase volume. But when it comes to creative, Tracy noted that human creativity is still a key part of that process.

“It's for the brands who figure out how to supplement the efforts of their creative team, of all the mundane stuff that's taking their time — think resizing images, writing different versions of copy, even making a still image move — these little moments that will have incremental points in performance and lead to this diversification that we know needs to happen,” Tracy said. “There's really low-risk ways you can get on the train. Doesn't have to be a full video yet, but there's ways that you can absolutely start using this and giving your creative teams time to be creative.”

Meta is also rolling out new ways for brands to use AI, including ads that customers can have a conversation with. Shih outlined a test case of this functionality with Kitsch, in which a shopper sees a video ad for a shampoo bar and has the ability to then ask the ad whether it will work for their hair type and order three different bars through the conversation. Voice agents like this also have the potential to break down silos between customer support and purchase-focused marketing chatbots, according to Shih.

“That voice agent might have a conversation with a consumer about pre-sales questions about the product — just asking if there's different allergens, how quickly can it be shipped to a certain location — but there also might be customer support questions,” Shih said. “In the past, we've deployed chatbots that were just for marketing or just for support. And so I think it's just time to make it a more cohesive customer experience and AI to lead the way and enable that going forward.”

The social media giant also just released omnichannel ads, which use AI models to connect shoppers with products that are in stores in addition to online. Shih proposed the example of a customer in Las Vegas shopping for makeup for the night, who needs to buy a product in person.

“Our local ads can now show which store nearby, across the street in the Las Vegas mall, has this particular product in stock and target us specifically because of a variety of factors that AI picks up,” Shih said. 

Overall, Shih likened the state of AI now to the early days of the internet.

“Roll up your sleeves and start trying. I think that AI is one of those technologies that you really learn best by doing,” Shih said. “Those who are able to lean in, just like those who were able to embrace e-commerce early 20 years ago, it's a moment to create sustained competitive advantage.”

Foot Locker

“I think it can be really tempting with retail tech trends to think of it as an ‘or,’ right? It’s framed as stores or digital, human-centric experiences or AI-powered experiences,” Kim Waldmann, global chief customer officer for Foot Locker, said. “It's easy to get to the ‘or,’ but really it's a false dichotomy. I think what we really believe at Foot Locker is tech enables more human powered-experiences … tech and AI can remove some of the more mundane, transactional, manual work from our store employees and free them up to do what they do best, which is be creative, storytell.”

As one example, Waldmann highlighted that the retailer is using AI to sift through customer experience data and help customer service employees resolve issues faster. That has in turn improved net promoter scores. The retailer also uses AI to create product details that are relevant for different markets — for example, translated into a different language or using different measurements — from the same product photo.

Improving search results and product recommendations on the retailer’s website is another “key focus,” according to Waldmann. But many of the applications are on the back-end. Foot Locker has invested in business reporting and intelligence functionalities that use AI, for example.

“Right now, you can make a vocal command or vocal request of the system and get, for instance, what were my gross margins and sales in Spain yesterday?” Frank Bracken, who was named president of Foot Locker last week, said. “And a report is generated automatically.”

An AI-driven workforce management system the retailer recently implemented takes into account data like sales forecasts and promotional calendars to help land the right number of employees in stores.

“The more productive we can be as an organization, the more we can flow into consumer-facing investments, whether it's brand building, whether it's storytelling or more refreshes and reimagined Foot Lockers,” Bracken said. “So it becomes a very perpetuating cycle in a great way.”

Google

“We can't say it enough that AI is a significant opportunity, not just to run better ads, but to actually run a better business,” Sean Scott, vice president and general manager of consumer shopping at Google, said. “And I really do think the brands that are the most willing to adapt and the most quickest to adapt will be the ones that get ahead.” 

For Google (and sister company YouTube), AI is behind features like Google Lens (which allows shoppers to search for items based on images), AR product viewing and local shopping. Much like Meta’s omnichannel ads, Google’s local shopping lets customers locate an item at a store nearby versus just online. As an example, Scott noted he’d used Google to locate a part he needed to fix a leaky faucet and picked it up that day at a nearby store. But it only works if brands participate.

“What that means is you have to upload not just your online inventory or not just one store’s inventory, but all of your stores’ inventory,” Scott said. “Because, again, shoppers are everywhere and if you're going to reach them wherever they're at, you need to make sure that that inventory is available for them so dads like me can fix leaky faucets.”

When it comes to features like Google Lens, Scott said the company sees about 20 billion visual searches per month and 25% of them have commercial intent. The executive also noted that AI is helping make very specific searches on Google yield more helpful results. A search for “men’s jeans for a Shoptalk keynote” brought up advice for dark washes and straight cuts, according to Scott, whereas a search for “skis for the Pacific Northwest” recommended wider skis to cope with the heavier snow in the region.

“What's really, I think, unique and amazing about all this is that we're not a retailer. We don't have any inventory. We don't ship anything,” Scott said. “And yet, over a billion times a day, people are shopping across Google.”

Tapestry

Summarizing associate feedback and generating product copy are two ways Tapestry is already using AI — and the luxury retailer is also considering how it can be used to generate code and experiences faster. But even as more uses become obvious, it’s important to start simple.

“Let's be real, right? There is fear in peoples’ mind, there is hesitation in peoples’ mind,” Bhatia said. “And how is it going to impact my life? How is it going to impact my job? And those things are very real and you need to acknowledge that — you need to have empathy with that.”

Bhatia stressed the need to train employees and bring them along on the journey across all levels of the organization.

“Have a lot of empathy with people, try different things and then create communities where it's safe to ask questions, where it's safe to learn, and that's how this change and training will happen,” Bhatia said.

Amazon is expanding its suite of generative artificial intelligence products targeted at marketers with a new conversational assistant, according to an announcement shared with sister publication Marketing Dive.  

The chat agent, positioned as a “creative partner,” draws on Amazon’s troves of retail data and cloud-computing muscle to assist with tasks like product and audience research, brainstorming and storyboarding, and producing display and video assets that can be deployed across Amazon Ads. 

The tool is launching in beta on Amazon’s Creative Studio portal at no additional cost to advertisers. Like many AI products, it promises to improve speed and efficiency for marketers that are contending with tighter resources but higher demands for efficiency and performance. 

Amazon’s new chat assistant joins a growing fleet of AI agents, or agentic AI, that are designed to enhance productivity while cutting down on cost. The feature — described in a blog post as a digital “creative partner and strategist” — works with marketers across the campaign-planning journey, analyzing shopper data, product pages and other sources of information to help with everything from preliminary research to finalizing creative assets. Ads can be run across Amazon DSP, Sponsored Display, Sponsored Brands and Sponsored Brand Video. 

As with its other generative AI bets, Amazon is positioning the assistant as a democratizing feature for small- and mid-sized brands that may not have the budgets to develop or manage high-gloss ad campaigns on their own. The announcement emphasized that marketers do not need any specialized AI or design training to make use of the solution, and that “granular” controls are in place for giving feedback each step of the way.   

“This is about more than speed — it’s about giving every advertiser and agency access to the kind of strategic, high-quality creative support that once only large brands could afford,” wrote Jay Richman, vice president of creative products and technology at Amazon Ads, in the blog post.

That said, large marketers like Nestle Health Science were among some of the chat’s beta testers, praising its impact on workflows and ideating in press materials. The chat consistently turned up insights that Nestle Health Science hadn’t previously considered based on its product set, said Dayexi Tomko, brand manager, in a statement.

In a use case for the agent illustrated in Amazon’s blog post, a fictional outdoor gear retailer works through a product launch, brainstorming taglines with different tones, lifestyle imagery, background music and finally multiscene ads that play as a couple watches TV in their living room. The agentic capabilities work off of Amazon Bedrock foundation models, including Amazon Nova and Anthropic Claude.

The assistant helps round out Amazon’s generative AI suite for marketers as the e-commerce giant continues to build a significant advertising services business. Previously, the company introduced text, image and video generators, with the latter getting a major upgrade and wider availability earlier this summer.

At the National Retail Federation’s Big Show at the Javits Center this year, the topic of AI was inescapable.

Nvidia’s Azita Martin, who is the company’s vice president and general manager for retail and CPG, helped set the tone as she discussed “game-changing times for the retail industry” with Walmart U.S. chief John Furner. She also spoke with Lowe’s Chief Digital and Information Officer Seemantini Godbole and Randy Lack of Dell Technologies about how the home improvement retailer “leverages AI to become an agile digital retailer.”

“AI is real, and I encourage all of you to get started,” Martin told the NRF audience, after Furner noted that AI can be intimidating. “I would first of all really recommend that the AI initiatives be top-driven. In other words, you need executive sponsorship. You need to have the top executives in your company believing. The second thing I would look at is, what are some of your biggest business challenges?”

Because of the massive number of AI-related sessions, speakers and vendors at the 2025 Big Show, there is no way to capture all that was said on this vast topic, but here are four themes that emerged.

AI could revolutionize supply chains

Retail supply chain executives were late to the party, but have recently embraced the potential of AI to streamline and improve several processes, from forecasting and procurement to fulfillment.

The pandemic roiled supply chains, which led to inventory woes that some retailers are still grappling with. That has revealed the need for updated technologies. Nvidia itself recently unveiled Mega, a new AI product that employs a digital twin to develop, test and optimize systems before they’re actually deployed at facilities, while Lowe’s has employed digital twins at 1,700 of its stores that provide daily operational and inventory updates.

“Supply chain, more than anywhere in retail in my opinion, is going to benefit the most from AI,” Martin told Furner. 

AI pumps up personalization, but privacy remains a concern

E-commerce players have long boasted that they have access to customer data that stores can’t collect, and AI is set to deploy that in fresh ways, several speakers noted during the conference.

This means a new era of digital shopping assistants and other methods of communication. Tapestry, which runs Coach, Kate Spade and Stuart Weitzman, is piloting AI-generated imagery that the company’s Vice President of Omni, Trang To, said “has the potential to unlock hyper-personalization at scale.”

“Cross-domain information feeding our AI models, to be able to deliver more insightful outputs on what the customer wants to see next, is critical,” she said. “And the second piece, historically, that has been the biggest hurdle for us in personalization is the speed at which we could create creative content to address all of these individuals versus segments. And that's something that I'm really excited about.”

But privacy remains a concern for many consumers, making it a concern for retailers as well.

“A decade ago [there was] a tug of war between privacy and personalization, and the tug of war is still going on,” Macy’s, Inc. CEO Tony Spring said during another session. “The reality is, we own the responsibility with the relationship ... We have to be better caretakers of information. We have to be smarter with the information we already have, and earn your trust over time with what we do with it.”

The use-case isn’t always clear

It was difficult to find a retailer at the Big Show that wasn’t at least piloting some AI-driven technology, in some capacity, for some area of its business. But many, if not the majority, seem to be in test mode.

“It's always going to be a learning experience, right?” Karen Etzkorn, chief information officer at Qurate Retail Group, said. “I think we find with Gen AI, every use-case we try, some of them have gotten significant value. Others we put to the side.”

At H&M, 2024 was a year of exploring AI, in the spirit of what Chief Digital Information Officer Ellen Svanström called an entreprenuerial mindset that is part of company culture.

“The interest and the exploration from our colleagues has been big in this area,” Svanström said. “However, we've yet to see a very real ... use-case that we can truly see specific and good value from, but we see enormous potential. In parallel, we have invested in and focused on and been very disciplined to build out our new data platform to ensure that we can truly scale in a secure way going forward. So, going forward, being open to what the real use-cases will be, to deliver value, we are ready to really focus and nail down on that. So I think in 2025, focus will be key for us.”

At least for now, there is no retail without humans

AI may have dominated the Big Show, but the 40,000 humans who attended it represent the billions of customers and employees who dominate the world of retail.

At American Eagle, Chief Marketing Officer Craig Brommers said the apparel retailer is approaching the use of AI with caution, in an effort to avoid the kind of high-profile AI-fueled blunders made by a few brands last year and preserve the authenticity and diversity its customers value.

In fact, AI is emerging as a tool to assist humans, rather than take over from them.

“Even in an environment where we know technology is going to play a bigger and bigger role in retail, listening to consumers and human interaction still matters,” Target CEO Brian Cornell told an NRF audience. “I think it's that balance between technology and that human interaction in retail that's going to pave the way for the future. So I think you’d better spend some time making sure you build trust, have those honest conversations, explain the ‘why.’ In a changing environment, we know it's going to be the combination, it's not an either-or. It's going to be that combination of technology that's going to fuel growth, and human interaction in retail, I believe, is going to play a role for years to come.”

Though Gen Z is leading the charge on AI shopping assistant usage, purveyors of AI technology face skepticism from consumers. 

Forty-three percent of Americans are aware of AI shopping assistants, but only 14% have used one, according to a survey from YouGov. Such tools are most popular among Gen Z users (24%), followed by millennials (17%), Gen Xers (12%) and baby boomers (7%). 

For consumers who are open to AI shopping assistants but haven’t used them, about two-thirds (67%) said they would use AI to find the best prices and 56% would like help comparing products. Nearly half (46%) said they’d want to use them to receive product recommendations, the survey found. 

Trust remains a barrier to adoption, with 41% of survey respondents stating they have no trust in AI shopping assistants. Indeed, they are one of the least trusted sources of shopping information, according to YouGov.

“In open-ended responses, some consumers expressed outright hostility with comments such as they ‘don’t trust AI,’ ‘hate AI,’ or that it’s ‘destroying our planet,’” YouGov said in its report. “Others framed AI as a threat to democracy, humanity, and the environment. These sentiments highlight a deeper resistance, particularly among more tech-wary consumers.”

Among those who avoid using AI shopping assistants, more than half (54%) of survey respondents said they don’t see the need for such tools in the first place and 45% said they prefer human assistance. Over a third (34%) of respondents were concerned about data privacy and security, and 30% suspect that AI shopping assistants would attempt to sell them things they don’t need, YouGov found.

YouGov’s survey also offers some insights into which products shoppers would consider using AI shopping assistants to purchase. While 43% of respondents said they wouldn’t use AI to purchase anything, about a fifth (21%) said they would use AI to buy consumer electronics and 19% would purchase groceries or household essentials and clothing, shoes and accessories. 

Despite consumer concerns, companies are diving headfirst into AI shopping tools. Retail and tech heavyweights like Google, Amazon, Walmart and eBay have released AI shopping tools this year to help consumers with everything from finding products based on their interests and virtually trying on items to summarizing reviews and offering shoppers advice and suggestions. 

Across the industry, retailers are pouring more resources into AI. A Capgemini report recently found that 56% of retail organizations have increased their investments in generative AI compared to last year.

Retail buyers are trying to get ahead of tariffs this holiday season by leveraging AI tools.

Nearly 80% of industry buyers say they have utilized AI-enabled tools to enhance buying activities, according to Deloitte’s 2025 Retail Holiday Buyer Survey. Nearly three-quarters said they are specifically using the technology to handle changes in trade policies.

In the survey of 50 retail industry buyers, respondents said they placed over half of holiday orders by the end of May — about two months sooner compared to the survey’s 2024 insights. 

Meanwhile, 78% of respondents said they’re worried about attaining proper inventory levels and 76% are concerned with supplier reliability due to geopolitical tensions.

The impact of shifting consumer sentiment and an evolving trade landscape is top of mind ahead of the holiday shopping season.

In Deloitte’s new holiday survey, 20% of retail buyers say they’ve paused shipments of some orders until the trade environment is more certain and 18% have even canceled orders they determined were unprofitable. The survey was conducted between May 20 and June 4, and U.S. tariff policies have continued to change weekly since then. 

Almost half of retail buyers noted they plan to source from new vendors, while a smaller amount aim to negotiate better terms with suppliers or increase domestic sourcing. In recent months, several retail companies (such as Target, Best Buy and Warby Parker) have expressed mitigation efforts related to tariffs that include sourcing from new countries, particularly outside China. 

Additionally, 76% of the surveyed retail buyers expect consumers to buy discretionary goods only around promotional periods. The new insight adds to a June report from Deloitte that found price-conscious consumers anticipate spending 40% to 50% less this year on discretionary goods. 

Data from Inmar Intelligence released in July found that consumers are certainly looking toward holiday promotions to fuel discretionary gifts — even at the expense of other purchases. In its 2025 Holiday Commerce Forecast, 82% of surveyed consumers said they’d plan to cut back on groceries and other essentials to cover holiday purchases.

It’s becoming the norm for consumers to use generative AI tools when deciding what to buy, a survey by Capgemini of 12,000 global shoppers found. 

Nearly three in five consumers have replaced traditional search engines with generative AI and over two-thirds of shoppers want the tech to aggregate search results from online search engines, social media platforms and retailers’ websites. Roughly two-thirds of millennials and Gen Z, closely followed by Gen X, have replaced traditional search engines with generative AI tools such as ChatGPT for product recommendations, the survey found. Just 44% of baby boomers use generative AI when shopping.

Consumers are also growing reliant on social media as a primary source across the entire shopping experience, from discovering products to making purchase decisions and engaging with retailers. 

Nearly one-third of shoppers used social media to buy products and identify new brands last year, up from one-quarter in 2023. That number is even higher for Gen Z shoppers, where more than two-thirds have discovered a new product or brand on social media.

“Social commerce is the retailer’s route to Gen Z,” Capgemini found. Over half of Gen Z have bought products through a social media platform.

Beyond recommendations, consumers are also looking to social media for answers to questions about products, for reviews and to reach the retailers they purchase from. Over half of consumers occasionally use social media to search for product information and reviews. Two in five shoppers use social media for customer service. 

Other reports have shown that Gen Z looks to social media the most for making purchase decisions. Over half of Gen Z have made an online purchase while browsing social media, a survey by Walmart and Morning Consult found. 

Research from the ICSC found the vast majority, 85%, of Gen Z says social media impacts their purchase decisions, particularly TikTok and Instagram as the top platforms.

Generative AI has made similar strides, but across all generations of consumers. Nearly three in five respondents to an Adobe survey last year said generative AI has already improved their online shopping experience. Two-thirds of survey respondents want brands to analyze previous purchases and customer information to personalize the shopping experience.

Walmart in March introduced Wally, a generative AI assistant for its merchants.

Wally is intended to help merchants better understand product performance, including why some products are under or overperforming or which categories had the largest gross merchandise value increases. 

“By automating time-consuming tasks and providing actionable insights, Wally enables merchants to focus on strategic, creative and innovative activities that enhance customer experiences and meet evolving customer expectations,” the retailer said in a statement.

The tool can conduct data entry and analysis, diagnose product performance, answer operational questions and complete advanced calculations. The AI model used in Wally has an advanced algorithm enabling it to process larger volumes of product data. The company said it plans to continuously improve the tool with feedback from Walmart merchants.

Walmart was an early adopter of generative AI. The retailer has incorporated generative AI for its associates to help them answer customer product questions or to help a shopper find a specific item. It also deployed the tech to improve its search, summarize reviews, prioritize product features for customers, and as part of a spatial design tool. 

Behind consumer-facing use cases, Walmart also incorporated generative AI for coding assistance. The retailer reports the tech helped save about 4 million developer hours, streamlining deployments and delivering code faster with fewer bugs.

Many other retailers are leaning on generative AI as part of company operations. In a recent earnings call, Target said it used the tool with social media, tapping generative AI to catch online trends and boost relevant products faster on its storefront, curating its digital experience.

Luxury shoppers are increasingly embracing artificial intelligence.

Two-thirds of luxury consumers say they use AI features when they shop for fashion online, according to a survey of 1,688 people released by Saks Global. The number of shoppers using AI is up 2 percentage points compared to the same time in 2024.

The most common use of AI for online shopping was providing body measurements for size recommendations or product visualization. However, respondents said it is more unlikely they will upload personal pictures for virtual try-ons. The vast majority of luxury shoppers say they are likely to engage with personalization activities, according to the survey. The top actions were creating an account, generating a wishlist and answering questions about their shopping preferences.

Luxury shoppers have high expectations for the service they receive compared to non-luxury shoppers, particularly when it comes to convenience, according to Matt Moorut, senior director analyst at Gartner. However, not all forms of personalization are equal.

“Personalization needs to be in support of better customer experiences,” Moorut told CX Dive in email commentary. “It needs to help customers achieve their objectives, rather than just showing that you know customers. Focusing on the benefit is a way of avoiding ‘creepy’ personalization, which can deter customers from engaging with brands.”

Luxury consumers’ top personalization demand is content and recommendations that offer exclusive access to limited products, rewards and offers, according to the Saks study. They also seek reminders of these offers.

AI can be a boon for retailers aiming to win personalization-hungry luxury customers, and the right precautions can maximize that opportunity.

Luxury shoppers usually adopt new technologies faster than the general population, and AI can offer retailers a leg up with this demographic, according to Moorut. However, consumers in general have mixed feelings about generative AI and luxury shoppers are no exception.

“Data privacy is increasingly a public interest,” Moorut said. “If you consider that consumers are often hesitant to provide a cell phone number due to privacy concerns, you can imagine how they would perceive the risk of providing something like a full body image.”

High-profile data breaches are weighing on consumer trust as well, according to Julie Geller, principal research director at Info-Tech Research Group. This could contribute to them refraining from sharing images while engaging with less personal options.

“Luxury shoppers, in particular, are wary of potential data leaks and the misuse of personal images, fearing they could be exploited for scams or unauthorized marketing,” Geller told sister publication CX Dive in emailed commentary. “Additionally, many value their anonymity and prefer to keep their shopping habits private, further discouraging them from sharing personal photos.”

AI can still enhance the customer experience without needing customers to share individualized personal information.

Luxury retailers can use AI with their own customer data to help them predict consumer preferences and deliver the right personalization to meet sophisticated expectations, according to Geller. They can also forecast trends and demands with greater accuracy to ensure they have what customers want available for purchase.