The world’s largest retailer seeks to build on its AI-enabled culture and business results by embracing ‘a unified, agentic AI framework that reimagines how our business operates,’ its US CTO says.
Walmart doesn’t intend to lose its crown as the world’s largest retailer anytime soon. And, according to US EVP and CTO Hari Vasudev, the $815B company’s artificial intelligence strategy will play a key role in preventing that from happening.
The retail giant already leverages AI in multiple areas of its business, from its supply chains to its storefronts — use cases that have delivered “real results” for the business, Vasudev tells CIO.com. Now the company is looking to “build the future of retail” with a recently announced strategy that puts agentic AI at the heart of its business.
With 2.1 million employees and more than 10,750 stores globally, implementing AI hasn’t been a simple task for Vasudev and his team within Walmart’s IT organization. It’s required years of hard work and perseverance to get to the point where AI is an intrinsic part of the company’s day-to-day operations.
[ Related: Agentic AI – Ongoing news and insights ]
Vasudev says the team’s work has enabled the retailer to “remove days, weeks, and even months from traditional processes,” transforming the organization substantially as a result. Following is a look at Walmart’s AI present and where Walmart IT hopes to take the company’s AI transformation in the future.
Embracing AI end-to-end
Walmart has infused AI into operational processes across much of its business — with distribution being a key use case.
AI and automation systems have “nearly doubled” capacity in the company’s distribution centers, according to Vasudev. Since 2017, Walmart has been working with tech firm Symbotic to automate its supply chain network, utilizing AI-fueled robots for sorting, picking, and packing orders. These robots are also capable of stacking products on the most suitable pallets to avoid damage and allow easy removal by warehouse staff.
Earlier this year, Walmart announced plans to expand its Symbotic partnership by agreeing to develop and roll out new automation systems within 400 Accelerated Pickup and Delivery (APD) centers in the coming years. The deal has also seen Symbotic take control over Walmart’s advanced systems and robotics business.
Elsewhere in its supply chain, Walmart is using AI for “demand forecasting, route planning, and last-mile logistics,” Vasudev says, noting that these uses are essential in helping Walmart provide “95% of US households with same-day delivery by year-end.”
Walmart stores also rely heavily on AI systems. Digital twins, which are digital representations of physical environments, help maintenance staff keep stores running smoothly by predicting potential issues to address before they cause real harm. Vasudev says these systems have “reduced emergency refrigeration alerts by 30%” and “cut maintenance costs by 19% through predictive insights,” for example.
A people-centric approach
Vasudev attributes the success of Walmart’s AI tooling to the fact that it’s designed to accommodate the needs of employees and customers first and foremost. As a “people-first business,” Walmart views humans as its “spark for innovation,” he says.
Walmart IT has rolled out to the company’s stores an AI-powered system that automatically allocates tasks, enabling in-store associates to work independently and “focus on higher-impact work,” Vasudev says. Originally designed for late-night stocking, the tool is now being trialed for other employee responsibilities and has helped reduce the time it takes managers to plan shifts “from 90 minutes to 30,” he adds.
Conversational AI, such as chatbots, also forms an important part of the modern Walmart employee’s daily toolkit. Five years ago, Walmart began offering a conversational AI feature for answering simple queries such as where a certain product is located in the store or what hours an employee will be working any given week.
But Walmart has since expanded the tool to turn complicated process literature into simple instructions that can be easily and quickly digested, such as processing customer returns when they don’t have a receipt. More than 900,000 Walmart employees now use these tools to perform over 3 million queries daily.
Another recent addition to Walmart’s workforce AI tooling is a smart translation tool that allows retail workers to communicate with customers in 44 languages. On the corporate side, Walmart employees can use the MyAssisant chatbot for uses ranging from summarizing documents to providing creative advice on how best to start a new task.
Entering the agentic AI era
In a significant expansion of its AI strategy, Walmart recently announced a “super agents” strategy that will see it leverage agentic AI throughout its operations. Agentic AI refers to automated and decisive systems that perform tasks with little to no human input, and Vasudev sees them as the key to unlocking “seamless engagement” for the retailer and its various stakeholders.
Walmart’s agentic strategy comprises four super agents for associates, developers, customers, and suppliers. Vasudev describes it as a push “toward an agentic future where AI acts as a true partner across every part of the business.” On the employee side, Associate Agent does things like answer HR questions, while the Developer Agent helps the company’s software developers write, test, and roll out new applications quickly.
Along with AI workforce tools, Walmart has also launched recent AI products for customers, suppliers, and advertisers. The customer-facing chatbot, called Sparky, provides product review summaries, product suggestions, and product locations. In the foreseeable future, it’ll be even more advanced — automatically ordering items when customers run out of them, creating event plans with the required products, understanding video and pictures rather than just text, and much more.
For suppliers and advertisers that work with Walmart, they’ll soon be able to use a dedicated AI chatbot called Marty. It will assist with tasks such as onboarding, setting up advertising campaigns, and managing orders.
To avoid fragmentation, Walmart’s new super agents run on “a unified, agentic AI framework that reimagines how our business operates,” Vasudev tells CIO.com. “This isn’t a layer on top of legacy systems — it’s the result of years of foundational investment coming together to help Walmart move faster, reduce complexity, and deliver smarter, more personalized experiences at scale.”
Major lessons along the way
Although Walmart’s AI adoption has been mostly successful, Vasudev admits that implementing the technology across such a “vast footprint” has presented valuable lessons for the retailer.
In particular, he says Walmart has realized that a tool that may work in “one store format or region” may not be suitable in another area of the business. With this in mind, Vasudev and his teams avoid taking a “one-size-fits-all approach” and rather embrace systems that are “flexible, adaptable, and designed to scale responsibly,” he says.
“That means constant experimentation and iteration — whether that’s intelligent backroom systems, AI-powered equipment, or next-gen digital twins that simulate real-world complexity,” he explains.
Given the well-documented risks presented by AI systems, such as the threat of data leaks and unconscious bias, Vasudev says that scaling AI “while prioritizing strong data governance, transparency, and responsible use” is key to its success.
Industry outlook
Walmart isn’t the only retailer embracing AI technology. According to Amperity’s 2025 State of AI in Retail report, 45% of US retailers are using AI every week or for longer periods of time.
While AI use is rapidly growing in retail, the report also found that only 11% of retailers believe they’re prepared to use it in all areas of their business — a stark contrast to Walmart, which has found uses for AI in virtually all its departments.
Sandeep Unni, senior analyst in the retail industry research practice at Gartner, attributes this to retailers feeling “overwhelmed with all the hype around agentic AI” and because “most of the industry lacks Walmart’s maturity level or skillset to undertake in-house development.”
As such, the vast majority of retailers will likely need to outsource their AI projects to third-party AI vendors, he says, to tap their ability to “articulate use cases and business outcomes” and, most importantly, “aid with implementation.”
He concludes: “They must also navigate challenges, including data privacy, ethical AI use, integration complexities, and customer trust to scale its use.”
