The Algorithm Sets the Menu: Inside McDonald's AI Pricing Engine Across 14,000 US Restaurants

The next time a Big Mac costs a dollar more two miles down the road, there is a good chance a machine-learning model made that call. A Reuters investigation published September 29 pulls back the curtain on one of the largest deployments of AI-driven pricing in American retail: McDonald’s now uses machine-learning algorithms to recommend “the optimal price” for every menu item at each of its nearly 14,000 US restaurants, and it is pushing franchisees hard to follow those recommendations.

The report, by Reuters reporter Waylon Cunningham, is based on screenshots of the company’s pricing engine taken in August 2026 and interviews with nine sources with first-hand knowledge of the burger chain’s strategy. The details of how the system works — and the pressure campaign around it — had not been previously reported.

How the engine works

McDonald’s pricing engine continuously analyzes data from millions of daily transactions across the chain’s US footprint and generates location-specific price recommendations for everything from Big Macs to discounted senior coffee. Among the factors supercharged by AI is an estimate of how much each store’s patrons are willing to pay.

Screenshots of the franchisee interface reviewed by Reuters show messages like “Your restaurant is showing MEDIUM SENSITIVITY to Price,” derived in part from “customer willingness to pay in your area.” The platform also ingests public price data scraped from the online menus of nearby competitors, including Wendy’s and Burger King — both of which told Reuters they do not use AI in their pricing decisions.

The AI platform is run by Tiger Analytics, according to two former Tiger employees, and McDonald’s does not simply let the model run free. The company regularly feeds it rules and corporate targets — attracting more customers, or boosting profits. Documented examples of those parameters include concentrating increases on items that had not been raised in at least two years, raising prices only on items that had already gone up in at least 30% of stores, and excluding ice creams and drinks from increases during the summer months.

Twenty-one percent, two miles apart

The engine’s most visible output is price divergence between neighboring stores. A Reuters check of the McDonald’s app in September found a company-run store in Fresno, California selling a Big Mac for $5.69, while another company-run restaurant two miles away charged $6.89 for the same sandwich — a 21% premium. Reuters could not confirm whether that specific gap resulted from the engine’s recommendations, but three franchisees said the system has widened existing price differences for the same product between restaurants, including from neighborhood to neighborhood within the same area.

McDonald’s frames the system as optional. Its pricing portal is “a tool, not a mandate, designed to provide restaurant-specific recommendations to help franchisees deliver value for customers and make informed business decisions,” the company said in a statement, dismissing Reuters’ reporting as “speculative and uninformed” claims that “attempt to recast a standard business practice as something controversial.”

The franchisees tell a different story. Five store owners described various forms of corporate pressure to adopt the AI pricing. In January, as part of new business standards, McDonald’s began requiring franchisees to be “constructively engaging with McDonald’s approved Pricing Consultant and Tools,” according to an internal communication reviewed by Reuters. A franchisee document from June shows the company records deviations from the recommendations in detail. CEO Chris Kempczinski told investors in August that “pricing non-compliance in certain cases is part of” the company’s business-review conversations with owners — reviews that determine eligibility to renew or open stores.

“You don’t really have much of a choice anymore,” said Karen King, a former franchisee who retired from the system last summer, describing phone calls from corporate officers whenever she strayed from the recommendations.

The economics behind the squeeze

The friction between headquarters and franchisees is structural. McDonald’s corporation makes the vast majority of its money by taking a percentage of each franchisee’s total revenue — regardless of the store’s profit margins. That gives corporate a strong incentive to push prices down to drive customer volume. Franchisees, meanwhile, face wages, rent, and other operating costs that are up an estimated 36% since 2019 per the National Restaurant Association, giving them every incentive to raise prices.

Lately the engine has been recommending more conservative pricing — including some decreases — which has intensified the tension. At the August earnings call, Kempczinski publicly called out the roughly one-third of franchisees who did not comply with the chain’s under-$3 menu pricing guidance, saying their “business results (were) a lot softer” — without disclosing how the company reached that conclusion. Despite the push for affordability, US foot traffic has declined year-over-year for every complete month since March, according to estimates from Placer.ai.

The antitrust shadow

The most legally sensitive dimension is that McDonald’s own portal terms acknowledge the risk. The terms of service, reviewed by Reuters in August, warn that franchisee users “may be competitors of each other” and that it is “particularly important for all Users of the Tool to understand and comply fully with anti-trust and competition laws” — advising owners to consult their own attorneys.

That language is “an acknowledgment there’s a potential problem,” said William Kovacic, director of the competition law center at George Washington University and a former FTC commissioner, noting recent FTC and Justice Department scrutiny of pricing algorithms and collusion concerns. Other experts countered that the legal risk is low because courts have given brands wide latitude to control franchisee pricing. The FTC and DOJ declined to comment to Reuters.

A short, instructive history

McDonald’s has used some form of AI pricing tool since at least 2019, and CEO Kempczinski first touted proprietary restaurant-level pricing tools to investors in 2023. That same year, the tools reportedly suggested a Connecticut franchisee charge about $18 for a Big Mac meal at a turnpike location, according to a discrimination lawsuit the owner filed against the company — a price that went viral and sparked outrage, though a December filing claimed it “caused no loss of sales.” McDonald’s disputes the suit.

The industry context is equally cautionary. Wendy’s endured a wave of criticism in 2024 after its CEO announced plans to test “dynamic pricing” (the company later said the comments were misconstrued and the system was never implemented). Instacart ended its use of AI tools that showed different grocery prices to different shoppers last December, after a study exposing the practice drew fire from consumers and lawmakers. Yum Brands, the largest restaurant company by location count, is also turning to AI for pricing and operations — suggesting the McDonald’s model is a preview, not an outlier.

Why it matters

The Reuters report lands at the intersection of the industry’s biggest unresolved questions: as AI pricing quietly becomes standard infrastructure in the physical economy, the gap between “recommendation” and “mandate” is where both consumer trust and legal exposure live. McDonald’s simultaneously tells franchisees the tool is optional and tracks their “non-compliance” with it — while warning them, in the fine print, that using it might implicate them in antitrust violations. A system that estimates every neighborhood’s willingness to pay may be good business, but it is exactly the kind of algorithmic differentiation that regulators, courts, and customers are only beginning to understand. The company told investors last week the “industry-leading” engine is central to its affordability bet. Whether customers share that framing, the next price check on the app will tell.