Inside Walmart’s AI strategy

The comeback is always greater than the setback, even for a retail giant like Walmart, which was once late to the world of e-commerce—and is now making strides in the retail industry with AI.

Gone are the days when Walmart was only known for its low-price inventory and vast selection of products under one roof. Today, Walmart has been driving conversations around AI and how it can be deployed within the retail industry. The scale of Walmart’s tech ambition is significant, with Walmart Global Chief Technology Officer and Chief Development Officer Suresh Kumar saying this is “just the beginning” for the legacy retailer.

Walmart earned an advantage by being early to the AI race, Scot Wingo, CEO of ReFiBuy, an AI startup focused on solving problems for retailers using advanced AI frameworks, told Retail Brew.

“They’re one of the first people to do something, and they’re not just a fast follower of Amazon,” Wingo added. “It’s a really interesting signal to me that they’re charting their own course.”

Several experts told IT and Retail Brew that Walmart, after a long game of catch-up in the e-commerce space, has found its bearings with its organization-wide AI deployment. But what does its AI strategy actually look like?

You’re probably wondering how we got here. Walmart is no stranger to leveraging technology and bolstering its IT infrastructure to enhance its operations. As previously reported, the supercenter made a set of strategic acquisitions in the late 2010s, including the purchases of online marketplace Jet.com and social media company Kosmix, which helped expand its tech talent and strengthen its e-commerce business. Retail experts say this was a steep learning curve for the company.

“For a very long time, Walmart was lost in the world of e-commerce,” Wingo said. “They didn’t take it seriously at first, and then, when they did, they had several failed attempts at the dot-com [era] and it was very hard to get the online piece lined up with the stores.”

The retailer eventually found its footing. In Q1 of fiscal year 2021, Walmart’s e-commerce sales increased 74% YoY, thanks in part to the Covid-19 pandemic. By 2023, Walmart’s website ranked as the third popular e-commerce website in the US, per Similarweb data.

More recently, the conversation has shifted from which retailers do e-commerce to which retailers best inject AI into the shopping experience.

“E-commerce was a big thing,” Leigh Helsel, partner at Centric Consulting, said. “Now, it’s all about conversational commerce and enabling that, most notably through AI.”

Go fetch! As part of its AI agent strategy, Walmart has four of what it calls “super agents,” which are agents composed of multiple, task-focused agents working together. One of these is Sparky, a virtual AI assistant available on the Walmart app. Sparky is meant to help shoppers find what they need, reorder items, and organize events, the company said.

Behind the scenes, new AI agents for Walmart associates, partners, and developers are unifying workflows, streamlining operations, and accelerating innovation across the company.

Sparky isn’t the only shopping assistant out on the market. Shoppers may be familiar with Amazon’s Rufus AI assistant, launched in 2024, that can similarly answer questions and offer recommendations to consumers. In theory, Rufus and Sparky offer a smoother, more effective shopping experience compared to traditional search, helping users get to the product they’re looking for faster, Wingo said.

“This improves user experience by shortening decision cycles for end customers,” Amine Allouah, partner and co-founder of My Custom AI, which helps businesses craft their own AI models across retail and e-commerce, told Retail Brew.

Amazon recently reported Rufus has been used by more than 250 million customers to date, and the chatbot’s users are 60% more likely to convert to a sale. The dollar value of Rufus’s impact was highlighted during Amazon’s latest earnings call, with a predicted $10 billion in incremental annualized sales.

Behind-the-scenes agents and AI. Along with Sparky and Walmart’s associate agents, Walmart also debuted super agents for its developers and partners. Wibey, Walmart’s developer-focused agent, is a tool that unifies the retail giant’s 200 agents for the company’s engineers, who are tasked with building and maintaining the tech that supports Walmart and Sam’s Club and provides shoppers with an AI-enhanced shopping experience.

In a September interview with IT Brew, Sravana Kumar Karnati, EVP of Walmart Global Tech, said “early renditions” of Wibey had been around for more than a year, and it’s now available for use across multiple channels to assist developers, acting as a “middle layer” in Walmart’s agentic ecosystem.

“We’re at the point where we are developing Wibey using Wibey,” Karnati said. “The next-generation capabilities are developed using Wibey [command line interface], and it’s a bootstrapping technique. So, the adoption of Wibey and GenAI is really good at Walmart.”

Wibey is joined by Element, the AI platform that Walmart’s developers use to build and test GenAI models. Karnati said the platform has been around for about four years, but has been enhanced to support developers in the agentic era, allowing them to monitor agent behavior.

“There needs to be a way of figuring out that the agent is actually doing what it’s doing, and nothing more,” Karnati said. “That evaluation framework is part of it, as well.”

Joseph Ours, director of AI solutions at Centric Consulting, told us any company that wants to remain “viable” needs to leverage agents in order to remain competitive, making it unsurprising that Walmart is doing the same and “leading” in the retail industry by doing so.

“Walmart learned early on that no matter what company you are, no matter what business you’re in, you’re a tech company, and you have to treat tech as an investment, not as a cost center,” Ours said.

Walmart’s super agent for partners, dubbed Marty, assists suppliers, sellers, and advertisers with orders, expanding the retail giant’s investments in AI to enhance its supply chain. The company has been a longtime user of digital twin technology to map out new distribution centers, and leverages AI forecasting tools to inform supply-chain decisions, including rerouting inventory to stores that are low on stock and selecting the best delivery routes for perishable foods. Last year, Walmart began to sell its AI-powered logistics technology to other businesses as a software-as-a-service offering.

Helsel said Walmart has always been strategic with its technology strategy since its early days, especially after understanding how it could aid it against arch-rival Amazon and benefit customers and associates.

“They work in an industry with a lot of razor-thin margins and…they believe in giving up the best price to the customers,” Helsel said. “They have to leverage technology [and] AI to enable workflows to keep supply chain costs down.”

Walmart’s true advantage when it comes to AI is the “tremendous databases” it possesses, according to Lucille DeHart, marketing and branding expert at Columbus Consulting. She said Walmart’s brick-and-mortar stores give them a significant leg up over Amazon because of the consumer behavior data points they can obtain.

“Walmart’s going to capture the eyes and the actual physical behaviors in their store environment. That will put them at a significant advantage, which is why Amazon keeps dabbling in grocery locations,” she added.

Walmart’s ChatGPT cart. Besides building superagents, Walmart recently said it’s joining forces with OpenAI to build out new AI capabilities, including the ability to shop directly within ChatGPT. The new AI-driven shopping experience is framed as an effort to meet shoppers where they are, especially given the rising referral traffic from generative AI sources like ChatGPT to retailer websites.

“Leaning in on ChatGPT is really clever, because what we’ve learned is a lot of these platforms, the early adopters get an unfair advantage long-term, because they expand to cover a lot of the surface area, and Walmart has a massive surface area with all their SKUs,” Wingo added.

If executed well, Wingo said, the ChatGPT shopping feature could become one of those turning points that eventually places Walmart ahead of Amazon.

“It’s another big bet. [Walmart is] saying, ‘This is our chance to really maybe outflank Amazon in [an] interesting way here,’” Wingo added.

While the rollout of the ChatGPT checkout on Walmart is not live yet, Wingo said, this could be the “second win” that Walmart gets in this arena.

But Walmart hypothetically risks losing control of some of its distribution to OpenAI, Allouah pointed out.

“They don’t control the traffic anymore,” Allouah said. “And also they are sharing that traffic with others, which means that ChatGPT now has more leverage on them.”

“Since OpenAI is not exclusive to Walmart, ChatGPT could divert traffic to competing retailers, later, once OpenAI gets the initial push from its Walmart partnership, increasing OpenAI’s leverage over Walmart’s distribution,” Allouah said.

Rewind. Despite its recent tech upgrades, Walmart is still catching up to Amazon. Walmart’s e-commerce business in the US turned profitable in Q1 of FY 2026, with e-commerce sales rising by 22% overall. Amazon’s sales from online stores around the same period grew 5%.

“It seemed like they started to get their act together around 2009, in my world, because they launched their marketplace,” Wingo said, adding that the e-commerce site still worked as a side appendage to Walmart’s store business.

It wasn’t until seven years later, during the Jet acquisition in 2016, that things started to shift for Walmart, Wingo said.

Ultimately, Wingo added, the perception everyone has is that e-commerce “seems easy,” but a lot has to happen on the backend to make it look seamless for customers to just go and buy a cool outfit online: “Everyone always underestimates how complicated it is.”