Walmart Hit $1 Trillion Without Owning the AI Stack

Walmart just became the first traditional retailer to cross the $1tn market-capitalisation threshold, joining a group largely dominated by technology companies such as Apple and Microsoft. The milestone came during John Furner’s first week as chief executive, following a year in which Walmart reported 27% growth in e-commerce sales and a 53% increase in advertising revenue, according to its most recent quarterly results.

Investors and analysts pointed to automation, advertising, and artificial intelligence as core drivers of that re-rating.

But the market response is notable more for how it chose to adopt AI.

Alongside Sparky and other internal agents, Walmart has developed retail-specific models, but has relied on partners for large-scale generative AI rather than building a proprietary stack.

Also read: Building an AI Operating System for Retail, with Walmart VP Anupriya Sharma

The Case for Not Owning the Complete AI Stack

That posture became explicit in October 2025, when Walmart announced a partnership with OpenAI to allow customers to shop, plan meals, and restock essentials through conversational interfaces powered by ChatGPT.

In the announcement, then chief executive Doug McMillon said that traditional e-commerce had been “a search bar and a long list of item responses” and that Walmart was “running toward a more enjoyable and convenient future” through partnerships.

OpenAI chief executive Sam Altman echoed that, describing the collaboration as a way to “make everyday purchases a little simpler,” without referencing exclusivity or co-ownership of underlying models. The language on both sides emphasised integration and access.

In January 2026, Walmart announced a separate integration with Google’s Gemini models to surface Walmart’s catalogue across AI-native discovery environments, again positioning intelligence as something layered onto its commerce infrastructure rather than owned outright.

Hari Vasudev, Walmart’s U.S. chief technology officer, said in an NRF panel that “open partnerships” are central to the company’s AI strategy, adding that Walmart intends to remain “very open to partnering” rather than committing to a single stack.

E-commerce analyst Juozas Kaziukėnas described to Modern Retail the strategy as “the opposite” of rivals that pursue closed systems, noting that Walmart has been willing to experiment across OpenAI, Google, and other ecosystems.

Walmart’s most recent annual filing emphasises capital expenditure on supply-chain automation, fulfilment centres, and logistics modernisation, with no disclosure of large-scale spending specifically tied to training or operating proprietary AI models.

By contrast, Amazon reports tens of billions of dollars annually in R&D and infrastructure spending tied to AWS and AI development, explicitly internalising the cost and risk of model ownership.

Why Investors Didn’t Penalize Dependence

A natural investor concern with this model is dependency.

Relying on external AI providers could, in theory, leave Walmart exposed if foundational models become scarce, more expensive, or strategically differentiated. Control over intelligence, the argument goes, may ultimately prove more valuable than optionality.

That concern has not been reflected in market behaviour. Following the OpenAI partnership announcement in October 2025, Walmart’s shares rose sharply, with analysts characterising the deal as a signal of technological credibility rather than a strategic concession.

By early February 2026, Walmart’s stock was up roughly 28% year-over-year, outpacing the S&P 500’s gains over the same period, according to The New York Times.

Part of the explanation lies in retail economics. Walmart’s competitive advantage is anchored in scale: more than 4,700 U.S. stores, dense fulfilment networks, and continuous real-time demand data across income brackets. AI systems improve the efficiency of those assets, but they do not replace them.

Reuters reporting on Walmart’s internally developed “super agents” describes them as systems designed to coordinate multiple tools and workflows rather than foundational models in their own right.

Regulatory exposure also cuts against the ownership argument. As scrutiny around AI hallucinations, data usage, and consumer protection increases, model owners face growing governance and compliance burdens. The Financial Times has reported on rising uncertainty over liability when AI systems generate incorrect or misleading outputs in consumer contexts.

By partnering rather than attempting to own and operate a vertically integrated, foundation-model stack, Walmart can adopt new capabilities while limiting direct exposure to model maintenance and regulatory risk.

The timing of the $1tn milestone follows that logic. Walmart’s valuation accelerated after a sequence of measured steps: AI partnerships in 2025, sustained growth in e-commerce and advertising revenue, a move to list its stock on the Nasdaq, and a leadership transition positioned as an acceleration rather than a reset.

Investors appear to have waited for evidence that AI could lift core metrics without distorting Walmart’s cost structure.

Walmart reached a trillion-dollar valuation without claiming to be an AI company. Instead, markets rewarded a retailer that treated large-scale generative intelligence as interchangeable infrastructure, capturing its gains while declining to absorb its volatility.