Are Companies Using AI to Gouge Their Customers?

Artificial intelligence can write computer code, diagnose diseases, and generate realistic images in seconds. But it could also be quietly helping companies gouge their customers on everything from plane tickets and hotel rooms to a carton of eggs.

That was the shocking conclusion from a recent Senate hearing on what lawmakers are calling “AI surveillance pricing” – the use of personal data, sophisticated algorithms, and AI to determine what individual consumers are willing to pay. Under surveillance pricing schemes, rather than one fair price for everyone, the price you pay for goods and services depends on what your individual customer profile says you would be willing to pay.

Senator Josh Hawley (R-MO), chairman of the Senate Judiciary Subcommittee on Crime and Counterterrorism, described surveillance pricing as a combination of “spying on people, ripping them off, and taking away jobs.”

The Federal Trade Commission has been studying surveillance pricing for several years. In early 2025, the agency reported that pricing intermediaries can draw on information including a customer’s precise location, demographics, browsing history, shopping behavior, and even computer cursor movements to help retailers tailor prices or promotions.

AI makes processing those enormous quantities of information faster and cheaper.

During the Senate hearing, Robert Hedges, a former chief data officer for Visa who now works at MIT, explained that companies can combine information about a consumer into detailed profiles designed to calculate one especially important number: their “willingness to pay.”

In plain English, the question is no longer simply, “What should this product cost?” It is now, “What is the most that this particular customer might pay for it?”

That distinction is at the heart of surveillance pricing.

One of the most striking examples raised during the hearing involved a JetBlue customer attempting to book a flight to attend a funeral. The customer complained on social media that the ticket price had increased by $230 in a single day. JetBlue’s official account responded by suggesting that he clear his browser cache and cookies or use an incognito window.

The exchange immediately fueled suspicion that browsing information was influencing the fare. JetBlue subsequently said the advice was erroneous and denied using customers’ personal information to set prices, according to Reuters. But lawmakers argued that the incident illustrated why consumers are increasingly concerned about what companies know about them and how that information might be used.

Other examples show how sophisticated data collection can become.

DoorDash, for instance, filed a patent application for technology that can analyze real-time and historical device data to calculate what it calls an “agitation score.” For instance, if someone has a low battery on their phone, they might be willing to pay more for a ride.

The patent does not establish that DoorDash uses agitation to raise prices, but the technology demonstrates just how precisely companies may be able to assess a customer’s behavior and state of mind.

Location, purchase history, loyalty-program activity, device information, browsing patterns, abandoned online shopping carts, and how long someone lingers on a product page can all potentially offer clues about how urgently someone wants something and how price-sensitive they are.

That creates an enormous information imbalance between buyer and seller.

A company might know where a customer lives, what he has purchased before, whether he has comparison-shopped, how often he has returned to a website, and whether his behavior suggests an urgent need. That customer, meanwhile, has no idea why the price on his screen is what it is – or whether someone else is seeing a different price.

“Surveillance pricing is good for corporate profits and bad for consumers,” Lee Hepner of the American Economic Liberties Project told senators at the hearing.

The phenomenon also extends beyond online shopping. Retailers are rapidly adopting electronic shelf labels, or ESLs, which replace traditional paper price tags with digital displays that can be changed remotely within seconds. Walmart, for example, plans to deploy the technology across its stores.

The labels themselves are not surveillance pricing. They can reduce labor costs, make updating prices easier, and allow stores to respond quickly to inventory changes. But critics worry about what becomes possible when those labels are connected to increasingly sophisticated pricing systems and individual customer profiles.

Hillary Caron of the United Food and Commercial Workers International Union testified that ESLs can connect with inventory, point-of-sale, and pricing systems that use consumer data and AI. They can also allow retailers to change prices far more frequently than was practical when employees had to manually replace thousands of paper tags.

In the near future, grocery stores could even change prices based on when individual customers are walking past a certain product. If Walmart’s data profile on you says that you are willing to pay more for a carton of eggs than the person in front of you, the ESL could theoretically show you a higher price.

“Families expect predictable pricing,” Caron warned, arguing that constantly changing prices could make household budgeting more difficult.

“Companies’ attempts to conceal this practice suggest they know it is unfair,” said Dr. Gustav Mårtensson, a former commercial lawyer. “Most consumers would not stand idly by if they discovered their personal data was being used against them just to boost retailers’ profits.”

The airline industry provides another glimpse of where AI pricing may be heading. Delta has partnered with AI firm Fetcherr on an effort its president described as a “full reengineering of how we price,” with AI increasingly assisting the airline in determining optimal fares. Delta has pushed back on suggestions that it uses customers’ personal information to set individualized prices, but the broader trend toward automated pricing is unmistakable.

What makes surveillance pricing alarming is the possibility that the variable being measured is not merely demand for the product but the individual customer standing in front of it.

Federal regulators are now grappling with that distinction. Shortly after the Senate hearing, the FTC announced that it was considering an enforcement policy on personalized pricing. FTC Chairman Andrew Ferguson said consumers generally expect a listed price to be the same price everyone else sees, rather than a retailer’s estimate of how much they personally might be willing to spend.

The FTC acknowledged that it does not have blanket authority to prohibit personalized pricing. But the agency warned that companies could violate existing law if they secretly use customers’ personal information to determine prices without adequately disclosing that practice.

That may ultimately be the most important issue for Congress to address.

AI will continue to make businesses more efficient and give companies extraordinary new tools to understand their customers. But there is a fundamental difference between using technology to determine what a product is worth and secretly determining how much an individual consumer can be persuaded – or pressured – to pay.