SAN FRANCISCO — “Super agents” are coming to Walmart Inc. to super-charge revenue.
Specifically, a quartet of artificial intelligence (AI) agents crafted to help shoppers and employees save time and effort (don’t forget money) are coming to supercenters and aisles of the country’s largest retailer. Their less-than-superhero names are Marty (for sellers and suppliers), Sparky (consumers), Associate Agent (where associates can access agents built on the corporate back-end), and Developer Agent. Their mission: Simplify the experience at multiple points of interaction with the company, which employs more than 2 million.
“Having a plethora of different agents can very quickly become confusing,” Suresh Kumar, chief technology officer at Walmart Global, said at the company’s recent Retail.Rewired innovation event, explaining the high-end agents and their ability to organize operations worldwide.
Taking agentic AI to the physical world, Walmart also is upgrading its Element machine learning platform by adding a super agent for software development of its semi-autonomous software robots.
It doesn’t end there. Added to the mix are digital twins propelled by spatial AI so Walmart can “detect, diagnose and remediate issues up to two weeks in advance” at its nearly 11,000 locations, said Brandon Ballard, group director for real estate at Walmart US. Digital twins helped Walmart US slash its emergency alerts by 30% and reduce its maintenance spend in refrigeration by 19% last year, he said.
Machine learning, too, has aided Walmart’s back-end operations in better calculating the time it takes delivering an order to a customer’s doorstep.
The agents are designed to tackle long-standing headaches, such as helping shoppers navigate complex product discovery and the lack of personalized results, enabling merchandisers to manage fast-changing assortments, and giving merchants instant access to insights instead of static reports, says Zohar Gilad, CEO at Fast Simon.
“The fact that a retailer of Walmart’s scale is embracing AI agents underscores that these solutions are not experiments, they’re essential tools for delivering agility and operational efficiency in e-commerce,” he said in an email.
Adds Frank Lin, CEO of AI Operator provider Shipflow, “What will be crucial for these tools is to go beyond simple tasks like data aggregation; these AI agents must handle multi-step processes end-to-end while dynamically adapting to complex, real-world conditions. Agentic AI that leverages real data to manage entire workflows will be the next true wave of value for these companies.”
There is a method to the retailer’s agentic blitz — the latest snapshot of a retail industry rushing to adopt AI in an effort to offset any slowdown in consumer spending amid economic, geopolitical, and inflation worries. Walmart’s use of agentic AI is part of a broader trend in reshaping supply chains and other industries.
Consider: Generative AI use catapulted 3,300% during this year’s Prime Day event, according to Amazon.com Inc. Google Cloud, meanwhile, is partnering with body care retailer Lush to visually identify projects without packaging, thus slashing the cost of training new employees.
“AI is transformative in a lot of ways, and streamlined customer service and advanced search are obvious” in retail and other industries, Nicholas Holland, head of AI at HubSpot Inc., said in an interview Wednesday in San Francisco.
Research from Sinch published in July showed half of those polled said they trust AI as much as humans for product recommendations. Indeed, it might just prompt even more spending on goods; the survey found nearly 40% want Black Friday to be moved up to Oct. 28 from Nov. 28. “Consumers especially like what they see with AI for retail and finance. It is trusted, rich and valued,” Julia Fraser, executive vice president of Americas at Sinch, said in an interview.
To that end, Fayez Mohamood, CEO of Bluecore, said Walmart’s investment in super agents “validates what we’re seeing across the industry — agents are quickly becoming the new interface for retail, both in how shoppers discover and buy and in how associates work.”
Nearly 60% of Americans use GenAI for product research, with more than 18 million conversations with shoppers across retailers like Wayfair, DXL, and Brilliant Earth, he said.
“Agents that behave like knowledgeable associates drive twice conversion rates, millions in incremental revenue, and a 30% reduction in support tickets,” Mohamood said in an email. “One of our customers, evo, has seen 25% of purchasers interact with its agent, generating over $3 million in incremental revenue.”
It isn’t all peaches and cream for AI-assisted retail, however.
“Retailers are moving fast into agentic AI, but with that innovation comes exposure to expanded attack surfaces for retail organizations,” David Holmes, chief technology officer of app security at Imperva, said in an email. “AI agents are starting to blur the line between helpful automation and harmful bot activity. Even simple tasks like pricing updates can create ripple effects on a scale, making it harder to tell what’s truly good for businesses and customers.”
Last year, bots outpaced humans on the internet, with automated traffic comprising 51% of all activity, and nearly three-fourths of that came from malicious bots weaponizing AI, according to Imperva.
