We Price the Product, Not the Person: Walmart CEO Draws a Hard Line Against AI Surveillance Pricing

On September 25, 2026, John Furner — president and CEO of Walmart Inc., the largest retailer in the United States — published an unusual open letter to “our customers and members.” Its centerpiece was a single, quotable sentence: “We price the product, not the person.”

The letter is Walmart’s most direct response yet to a controversy that has been building for two years: the convergence of electronic shelf labels (ESLs), AI shopping assistants, and the practice regulators and unions now call surveillance pricing — using a shopper’s income, browsing history, location, or “moment of need” to quote them a different price than the person standing next to them.

Furner’s answer, in effect: not at our stores. And the pledge explicitly covers the company’s AI systems, not just its price tags.

What the letter actually commits to

The letter lays out three commitments, each aimed at a distinct fear consumers have voiced about retail AI:

1. No identity-based or time-based differential pricing. “We don’t set different prices based on who you are or the time of day, and we won’t,” Furner writes. “Your income, shopping history, urgency or what we think you could pay won’t change the price” — whether someone is “buying groceries or electronics on a hot afternoon or in a sudden rush for an item.”

2. The same promise binds the AI tools. Walmart’s generative-AI shopping assistant, Sparky, gets called out by name. “When you engage with Sparky, that’s an invitation to serve you better, not to use your personal information to set a personalized price,” the letter states. Furner adds a striking analogy: “We’ve never used the relationships our associates have with customers to charge more, and we won’t do that with AI.” The commitment also bars using information shared with shopping tools to “hide lower-priced options that meet your needs” — a direct acknowledgment of fears that AI assistants could quietly steer users toward pricier items.

3. Humans stay accountable for pricing. “Our people will continue to oversee pricing, and we’ll monitor and test our technology against these commitments.” The CEO positions this as an ongoing verification exercise, not a one-time promise.

The framing is anchored in Walmart’s decades-old Every Day Low Prices (EDLP) doctrine. Furner argues that personalized pricing wouldn’t merely be creepy — it would be a category error for the business model: “using someone’s income, shopping history or moment of need to charge them more would violate the EDLP promise our business model is built on. We won’t do it.”

Why this letter, why now

The timing is not accidental. Three pressures have converged on US grocers in 2026:

The FTC’s surveillance-pricing investigation has teeth. The Federal Trade Commission voted unanimously in 2024 to compel eight intermediaries to detail their surveillance-pricing products, and its January 2025 interim findings described “a wide range of personal data” — browsing and shopping history, income estimates, location, device signals — feeding systems capable of serving individualized prices, discounts, and targeted ads. The FTC docket (FTC-2026-1057) remains open, with consumer groups filing comments as recently as September 9, 2026, arguing that disclosure alone won’t solve the problem.

Electronic shelf labels reached scale. Walmart, Kroger, Whole Foods, and Target have all deployed ESLs across large parts of their US footprints. The tags themselves are innocuous — they sync shelf and register prices and spare staff from swapping paper tags, a chore Furner writes he has personally done. But the same infrastructure that fixes pricing errors also makes it technically trivial to change a price centrally, instantly, and — in principle — selectively.

The AI layer raised the stakes. A digital price tag is passive. An AI shopping assistant is not. Sparky sits inside the Walmart app, learns what customers ask for, knows their purchase history, and can act on their behalf. Critics noted that this is precisely the combination that makes surveillance pricing actionable at scale: an agent that knows what you need, how urgently, and what you can probably afford. An August 2026 report from the Groundwork Collaborative, “Your Data, Their Profit,” tied AI deployment at Kroger and Walmart directly to the surveillance-pricing risk, and the AFL-CIO has called for an outright ban on ESLs, citing both consumer harm and job displacement. A national UFCW campaign against grocery surveillance pricing is underway, and state legislatures have introduced bills restricting the practice.

Into that storm, Walmart’s CEO sent a memo — to employees and customers — deliberately preempting the worst-case narrative before a competitor, regulator, or lawsuit defines it for them.

The spectrum nobody wants to be on

FTC staff and academic researchers generally describe a spectrum of practices: targeted discounts at the benign end; individualized prices at the hostile end; and in between, algorithmic manipulation of search results, product ranking, and what an AI assistant chooses to show you. Walmart’s letter is carefully drafted to address the full span — which is why the “hide lower-priced options” clause matters as much as the headline pledge.

It’s also why verification is the real test. A promise that “our people will continue to oversee pricing” leaves open how Walmart will demonstrate compliance. Will it publish audits? Allow third-party testing of Sparky’s recommendations? Commit to the kind of pricing-transparency APIs that regulators have floated? The letter offers “monitor and test our technology against these commitments” but names no mechanism, no timeline, and no external auditor.

Skeptics make two further points. First, independent researchers who studied more than 180 million Kroger prices after that chain’s digital-tag rollout found no evidence of surge pricing — suggesting the industry’s most-feared scenario is not yet common practice, and that Walmart may be pledging away something it wasn’t going to do anyway, at low cost and high reputational yield. Second, “we don’t set different prices based on the time of day” governs shelf prices, not the personalized coupons, rollbacks, and app-only offers that remain core retail tooling; the line between a personalized discount and a personalized price is thinner than any CEO letter would like to admit.

Why it still matters

Even with those caveats, the letter is a milestone in how AI-era retail is being governed — not by statute first, but by public commitment under scrutiny.

For one thing, it is the largest US retailer explicitly binding its AI agent to a pricing-neutrality promise. Sparky is not a website; it’s an assistant that acts for you. Walmart has now publicly stated that the data customers hand to an AI assistant cannot be converted into price discrimination — a formulation other companies deploying shopping agents will find hard to avoid matching.

For another, it reframes the question regulators have been asking. The FTC’s study asked: what data are intermediaries using, and how? Walmart’s letter answers a different, sharper question: what will you commit to never doing? In a policy environment where a ban on ESLs is being seriously debated by labor and where FTC Chair Andrew Ferguson has separately suggested AI developers should bear liability for their agents’ actions, a self-imposed bright line from the industry’s biggest player changes the baseline. After September 25, “everyone does it” is no longer available as a defense — to anyone.

And strategically, it’s a textbook trust play. Walmart’s letter leans on a 50-year pricing identity, personal history changing paper tags by hand, and the store associate relationship. The subtext: technology may be new; the responsibility isn’t. For a company betting heavily on AI — from supply-chain optimization to the Sparky assistant itself — establishing that AI serves the EDLP promise rather than undermining it is worth a great deal.

The commitments bind one company, for now. But they set a reference point for the AI-agents-in-commerce era: the biggest retailer in America just put in writing that its AI will serve you better, charge you the same, and show you the cheap option even when it knows you’d pay more. Whether “monitor and test” proves enforceable — internally or externally — is the story to watch from here.