How important is retail as a proving ground for AI?

The AI hype train hit a prominent bump last month when Microsoft CEO Satya Nadella expressed concern that the sector could be dangerously overvalued if—and here’s the caveat—a wider set of companies don’t start benefitting from the technology soon.

“For this not to be a bubble by definition, it requires that the benefits of this are much more evenly spread,” Nadella said during the World Economic Forum. He added that non-tech firms specifically need to see the upside of AI investment.

OpenAI, which is partly owned by Microsoft, is seemingly harboring similar concerns. In a recent blog post, CFO Sarah Friar said “practical adoption” is the company’s main focus in 2026, and zeroed in on one use case in particular: helping people buy things.

“People come to ChatGPT not just to ask questions, but to decide what to do next,” she wrote. “What to buy. Where to go. Which option to choose.”

This heady combination of fear and opportunism from the highest levels of the tech industry highlights a question that could prove essential not just for AI firms but for the economy as a whole: If large retailers don’t get on board with artificial intelligence, and quickly, will fears of a bubble prove true?

“It's a great question,” Nikki Baird, retail analyst and VP of strategy and product at Aptos, told Retail Brew. “The overall retail industry is something like $20 trillion, and they're going to need to take a piece of that at some point in order to help justify the valuations.”

Big opportunity, big risks: As it stands, AI firms have made some progress in this area. Fintech firm Ramp’s monthly measure of AI adoption shows that 40% of retailers currently have paid subscriptions to AI models, platforms, and tools, placing it ahead of healthcare, construction, and food services but behind manufacturing, finance, and technology.

Ramp economist Ara Kharazian said retail lags slightly behind the broader estimate of 46% adoption across industries, but still believes retail is uniquely positioned to “prove out whether or not AI can be implemented at scale by large corporations.”

This is what accounts for retail’s slight lag in adoption, he added. Because many retailers operate at such a large scale, companies such as Walmart and Costco need to make sure the technology is working properly before a full rollout. Additionally, the consumer-facing nature of the business means any mistake could have serious downstream effects.

“If you're working with consumers,” he said, “you really want to make sure the models that you're using are trustworthy and working as expected.”

Notably, there has been a greater focus on deploying AI behind the scenes than in front of customers. Caroline Reppert, senior director of AI and technology policy at the National Retail Federation, told Retail Brew “a lot of investment” had been in “more internal-facing use cases,” such as supply chain optimization and employee applications.

Picking a side: From the perspective of AI firms trying to generate revenue that would justify their sky-high market caps, the challenge is figuring out where exactly to take profit in the retail space, Baird said.

Right now, OpenAI is generating about 75% of its revenue off consumer subscriptions, but Baird predicts it will look more to B2B revenue generators such as advertising—which Friar hinted was on the agenda in the blog post laying out its priorities for the coming year.

“With open AI going after advertising they've said, ‘We think we have a better chance of monetizing the business side of it as opposed to the consumer side,’” Baird said, adding that it’s “extremely difficult to maintain your credibility and monetize both sides at the same time.”

This pivot doesn’t apply to the entire industry, however. At the same time that OpenAI is leaning more into B2B, competitor Anthropic is leaning more toward consumers. The company notably ran Super Bowl ads, and its CEO critiqued OpenAI for its embrace of advertising.

“For all these businesses, I think it's still kind of an early nascent market where they're all trying to figure out what's going to be the biggest revenue driver,” Kharazian said.