Both companies say their technology does not determine what individual customers pay. Lawsuits, patents and growing scrutiny are shifting the debate toward what these systems are capable of doing.

Walmart and McDonald’s have spent the past several weeks trying to reassure consumers that technology is not quietly deciding how much they should pay. Now both companies are having to defend that boundary more aggressively. McDonald’s is facing a proposed nationwide antitrust class action alleging that its AI-powered pricing system facilitates coordination across thousands of restaurants. Walmart, meanwhile, has issued a public rebuke of a consumer advocacy group that says the retailer’s patents reveal technology capable of much more sophisticated pricing than Walmart acknowledges. Neither development proves that the companies are individually adjusting prices based on who is standing at the register. But together, they are moving the debate beyond whether dynamic pricing is happening today and toward a harder question: how much should consumers care about what a company’s pricing infrastructure could eventually allow it to do?

When SSC previously examined the fight in “Walmart and McDonald’s Are Drawing a Public Line Around AI Pricing”, both companies were already trying to establish a clear boundary between technology that helps businesses decide what to charge and technology that actually determines a customer’s price. That boundary remains central to McDonald’s defense. The company says AI does not set the price of a Big Mac, change menu prices or decide what an individual customer pays. Franchisees ultimately control pricing. But Reuters reported that a federal lawsuit filed in Chicago alleges McDonald’s uses an AI-powered pricing system built from nonpublic data and machine-learning algorithms drawing on millions of transactions across roughly 14,000 U.S. restaurants. The plaintiffs argue the system improperly coordinates prices among franchisees. McDonald’s has called the allegations speculative and says pricing recommendation tools are common across industries.

That difference between a recommendation and a decision may eventually become important in court. About 95% of McDonald’s U.S. restaurants are independently owned, according to Reuters, meaning the case raises an unusual antitrust question: when businesses that ostensibly make their own pricing decisions rely on the same technology and the same pool of data, when does assistance begin to resemble coordination? The lawsuit has not established that McDonald’s violated antitrust law, and other algorithmic price-fixing cases have produced mixed outcomes. But it puts legal pressure behind a concern that until recently had largely been argued as a question of consumer trust.

Walmart is fighting a different version of the same battle. In an October 5 public letter, the retailer said it “price[s]products, not people” and does not use an individual customer’s income, shopping history, urgency or willingness to pay to determine an individualized price. Walmart also rejected claims that its digital shelf labels are infrastructure for real-time dynamic pricing. The company says the electronic labels replace paper tags and make already-approved price changes faster to display; they do not identify shoppers or independently decide what an item costs. Walmart went further, saying it does not and will not use those labels for individualized or dynamic pricing.

The argument intensified because Groundwork Collaborative is not simply pointing to the shelf labels. The organization analyzed Walmart patents filed over more than a decade and says some describe systems capable of using customer identity, shopping behavior and the contents of a shopper’s cart in pricing or product recommendations. Groundwork argues that those patents show Walmart is building an infrastructure that could eventually be used to profile customers and influence what they pay. Walmart rejects that interpretation, arguing that a patent establishes what technology can do, not what the company is actually doing or intends to do. Groundwork CEO Lindsay Owens has since called for Walmart to relinquish certain patent rights and submit its pricing technology to an independent audit.

That leaves consumers confronting a problem that is increasingly common in the AI economy: capability and practice are not the same thing, but the gap between them is difficult for outsiders to evaluate. McDonald’s can accurately say that a recommendation engine is different from software automatically changing a menu price. Walmart can accurately say that a digital shelf label capable of changing a price within seconds does not prove that the company is using it to surge prices when demand changes. The concern raised by critics is what happens when those systems are combined with increasingly detailed customer data, predictive models and technology capable of updating prices almost instantly. Consumers are effectively being asked to trust that the technological line companies draw today will remain in place tomorrow.

That is why the fight has moved since SSC first examined it. The initial debate was largely about whether Walmart and McDonald’s were embracing personalized or dynamic pricing. Both companies said they were not. Now the conversation is becoming more consequential: who gets to verify those assurances, what rules govern the systems behind them and whether technological capability deserves scrutiny before it becomes standard business practice. For consumers, the question may no longer be simply whether an algorithm set today’s price. It is whether the infrastructure being built around pricing will eventually make the answer much harder to know.