Scott Olson

Illinois resident Michael Thomas filed a lawsuit in federal court in Chicago on October 2, accusing McDonald's (New York Stock Exchange: MCD) of using an internal AI pricing system to coordinate menu prices across nearly 14,000 US outlets in breach of antitrust law. McDonald's denies the allegations, saying the AI does not set menu prices and that the decision rests with franchisees. The case remains at an early stage: the court has not found either side to have engaged in wrongdoing, nor has it approved class-action status.

At the heart of the case: the dispute is over data sharing, not AI itself

According to the complaint, about 95% of McDonald's outlets in the US are operated by independent franchisees which, in theory, should set prices independently and compete with one another. The plaintiff alleges that, since at least 2019, McDonald's has used a centralised machine-learning pricing engine that pools non-public sales data from company-owned and franchised restaurants before recommending an "optimal price" to each outlet.

According to an Associated Press report, the lawsuit centres on the tool's collection of data that competing franchisees would not normally share, including outlet-level sales. The complaint describes McDonald's as the sole channel through which one franchisee's data could influence a rival's pricing. A summary of the court filings by a litigation information website said the plaintiff had brought four causes of action, including price fixing under Section 1 of the Sherman Act and the unlawful exchange of competitively sensitive information. The other two claims are based on Illinois law. The plaintiff is seeking treble damages and an injunction against the arrangement.

McDonald's said the complaint contained numerous inaccuracies and that it would vigorously defend itself. The company said the pricing recommendation tool had been in use for more than a decade, taking into account factors including outlet sales, location and competitors' prices. Franchisees, it said, could decide for themselves whether to use the recommendations.

Reuters investigation: how "voluntary" are the recommendations?

The complaint cites a September 29 Reuters investigation. The report said the engine, operated by Tiger Analytics, analyses millions of transactions at nearly 14,000 outlets every day and calculates an "optimal price" for each item at every outlet. One key input is an estimate of how much customers at that outlet are willing to pay.

  • Pressure: Five franchisees told Reuters that McDonald's had pressured them to use the tool. Company documents showed that, from January this year, operating standards required franchisees to "constructively engage" with approved pricing consultants and tools. The company also recorded instances in which franchisees deviated from recommended prices. Chief executive Chris Kempczinski told investors that, in some cases, pricing compliance would form part of business review discussions with franchisees.
  • Price gaps: Three franchisees said the engine had widened price differences between outlets. In September, Reuters checked the McDonald's mobile app and found that a Big Mac cost US$5.69 (about HK$44.40) at a company-owned outlet in Fresno, California, while another company-owned outlet about two miles away charged US$6.89 (about HK$53.70), around 21% more. Reuters said it could not determine whether the difference had been caused by the engine's recommendations.
  • McDonald's response: The company said the system was a "tool, not a mandate" and that franchisees made the final pricing decisions.

As for who benefits from higher prices, the report showed that the picture was not simply one of "head office raising prices". McDonald's headquarters mainly takes a share of franchisees' sales and therefore tends to favour lower prices to drive traffic, while franchisees bear the costs and have a greater incentive to raise prices. Reuters also reported that, amid weak consumer spending in recent months, the engine had instead tended to recommend more conservative prices or even price cuts, which some franchisees facing cost pressures resisted.

Uncharted legal territory, with outcome uncertain

Jeffrey Shinder, a lawyer for the plaintiff who is not involved in the case, said algorithmic pricing lawsuits were testing whether the law could keep pace with technological development. Similar litigation could increase as public opposition to AI grows, he said. Some reports have described the case as the first lawsuit to allege that franchisees under the same brand used shared pricing technology to overcharge customers.

However, Reuters quoted some experts as saying that because the competitors involved were franchisees, courts had traditionally given brand owners greater latitude to control franchisee pricing, meaning McDonald's could face relatively limited legal risk. Outcomes in similar cases have varied. A US appeals court once revived a lawsuit alleging that casino hotels had used shared software to inflate room rates. The US Department of Justice also proposed a settlement on November 24, 2025, in an algorithmic pricing case involving rental software provider RealPage. A California law that took effect in January has also banned "shared pricing algorithms" that use competitors' data for anti-competitive arrangements under the state's antitrust rules.

Another front: personalised pricing and regulation

It is important to note that this case concerns alleged price coordination among franchisees, which is a separate legal issue from charging individual customers different prices based on personal data. McDonald's told TheStreet that its system did not involve real-time dynamic pricing and did not calculate how much an individual customer was willing to pay.

The latter issue is now a focus of legislation in several states:

  • New York state has required disclosure of algorithms that set prices using personal data.
  • Maryland signed a law on April 28 banning food retailers and food delivery platforms from using personal data to charge individual customers higher prices. It takes effect on October 1.
  • Connecticut signed a broader ban on June 4 covering general retailers.
  • New Jersey signed a ban targeting groceries on July 23, with most provisions taking effect on August 1, 2027.
  • More than 40 similar bills are pending in about 24 states.

Academics and central banks are also paying attention. Research by Dr Miroslava Marinova of the University of East London found that AI could shift pricing from the "market level" to each buyer's individual limit. Once pricing becomes invisible, she said, fairness becomes a central concern. A Bank of England official wrote in April that although dynamic and personalised pricing had increased, it had not so far pushed up inflation. It did, however, make the consumer price index (CPI) harder to measure because it was becoming increasingly difficult to define a "representative price".

Supermarkets are also deploying similar technology. Walmart has announced plans to install electronic shelf labels at 2,300 stores by 2026, while Kroger has used similar labels since 2018 and expanded them to about 500 stores. Neither initiative amounts to dynamic pricing, and Kroger has publicly denied using dynamic pricing. Some US senators are nevertheless concerned that the technology could pave the way for more frequent price changes in future.

What to watch next

  • Court developments: McDonald's has not yet filed a response in court, and the court has not decided whether to grant class-action status. The case is being heard in the US District Court for the Northern District of Illinois.
  • State legislation: Existing laws mostly target pricing based on personal data and may not directly cover price coordination between brand owners and franchisees. Whether more targeted legislation follows will be closely watched.
  • Federal action: Reuters said the Federal Trade Commission and the Department of Justice were closely examining whether algorithmic pricing could evolve into covert price collusion between competitors.