The FTC Wants to See the Math: What Its Nine Personalized Pricing Orders Ask Delta, Uber, Walmart and Others

The Federal Trade Commission sent nine companies a set of questions on October 28, 2026, and every one of them goes to the same issue: does the price you see, or the pay a driver is offered, depend on what the company knows about you? Delta, United, JetBlue, Lyft, Uber, DoorDash, Instacart, Target and Walmart each received a special order under Section 6(b) of the FTC Act. They have 45 days from receipt to answer.

These are study orders, not complaints. The FTC has not accused any of the nine of breaking the law, and its own announcement says Section 6(b) lets the agency run wide-ranging studies that have no specific law enforcement purpose. That distinction matters, but it should not make anyone relax. A 6(b) order is compulsory, the Commission can go to federal court if a recipient does not comply, and the answers will feed a public study that lawmakers in Washington and the states are already waiting for.

What the FTC is asking for

According to the FTC’s announcement, each company must turn over enough information for the agency to judge whether the price of a ticket, a ride, a delivery or another product varies from one person to the next, and why. The release names two kinds of inputs it wants explained: participation in loyalty programs, and analysis of personal data by algorithms. The orders ask the companies to document the inputs, tools and procedures they use to set or adjust prices.

The second half of the inquiry is about workers. Airlines usually pay staff salaries or fixed hourly wages, but ride-hailing and delivery platforms make shifting pay offers to drivers. The FTC says those offers may be based on a driver’s personal data, such as phone usage or the history of offers they have accepted before. That puts compensation inside the same review as consumer pricing, which is a wider frame than most privacy programs have used for this topic.

Chairman Andrew Ferguson’s statement was blunt: consumers “deserve to know whether companies use their web surfing habits or other personal data to estimate the highest price they can charge each consumer,” and workers “deserve to know if their compensation is affected by how companies use their personal data.” The Commission voted 2-0 to issue the orders.

Why this study looks different from the 2024 one

This is the FTC’s second pass at personalized pricing. In mid-2024 the agency sent 6(b) orders to intermediaries, the consultancies and software vendors that sell pricing tools, and staff released initial findings in January 2025, days before the change of administration. The new release is openly critical of that work. It says the earlier study looked only at “middlemen” and that the agency “learned nothing about actual pricing decisions” from it. This time the orders go to the businesses that set prices themselves.

That shift changes who carries the exposure. A vendor can describe what its tool is capable of. Only the company using it can say which data fields actually went into a price, whether a loyalty tier changed the number on the screen, and whether anyone tested the outcome. Those are the facts the nine recipients now have to produce, in writing, to a regulator.

The companies’ public positions so far

Several recipients have already said they do not do what the FTC is asking about. The Associated Press reported statements from Delta, JetBlue and Instacart. Delta said it has no fare product in use, in testing or planned that targets customers with individualized offers based on personal data. JetBlue said it does not use personal information or browsing history to set individual fares. Instacart said prices on its platform are set by retailers and that shopper pay depends on job factors such as distance and item count.

Those statements may well be accurate. They also set a benchmark. If a company tells the public it does not personalize prices, the documents it gives the FTC will need to match, and the FTC’s Section 5 authority over deceptive statements applies to public claims as much as to privacy notices. One day after the orders went out, DoorDash and Instacart announced a deal for DoorDash to buy Instacart. The FTC release does not mention the transaction.

The wider pressure on personalized pricing

The orders land in a crowded field. The FTC published a policy statement on August 19, 2026, saying personalized pricing without adequate disclosure “likely violates Section 5.” In Congress, House Energy and Commerce ranking member Frank Pallone has sent several rounds of letters, the latest on October 22 to ticketing, travel booking, rideshare and delivery companies. We covered his earlier inquiry to 25 retailers when it began.

The states are moving faster. California Attorney General Rob Bonta opened an investigative sweep in January, which we discussed in our piece on Bonta’s surveillance pricing focus. Maryland, New York and Connecticut have passed pricing rules, Maryland’s being the first to target grocery prices, as explained in our coverage of Maryland’s law and of New York’s vote. Seattle approved a citywide grocery measure in September. A December 2025 Consumer Reports study found some grocery items priced up to 23% differently between customers on Instacart’s platform, after which Instacart stopped offering the tool to retailers.

Airlines have been a recurring target since well before this week. Our earlier analysis of Delta’s AI pricing plans walked through why travelers and regulators reacted the way they did.

Dynamic pricing versus personalized pricing

The legal risk depends on a distinction that often gets blurred. Dynamic pricing changes the price for everyone at once, based on demand, time, inventory or seat availability. An airline fare that rises as a flight fills up is dynamic. Personalized pricing changes the price for a specific person, based on who they are or what the company knows about them. The first is ordinary economics. The second is the one regulators are investigating, because it turns a pricing decision into a use of personal data.

Most companies do some of both, and many do things in between: targeted coupons, member-only rates, loyalty tiers, offers shown only to logged-in users. Each of those is a price that differs by person. Some are well disclosed and expected. Others are not, and the difference usually lives in the data inputs rather than in the marketing copy.

Five questions to answer before a regulator asks

Only nine companies received orders, but the questions in them are a fair preview of what any retailer, platform or travel business could be asked by a state attorney general. Working through them now is cheaper than doing it under a deadline.

1. Which data fields can reach a price? List every input to your pricing, promotion and offer systems, including those passed in by vendors. Location, device type, browsing history, purchase history, loyalty status and inferred income are the fields regulators keep naming.
2. Who decides the final number? Separate rules a person wrote from model outputs. If an algorithm sets or adjusts the price, record who approved it, what it optimizes for and what limits it has.
3. Can two customers see different prices for the same item at the same moment? If yes, document why. If you believe no, test it with controlled accounts and keep the results.
4. What did you tell people? Compare privacy notices, loyalty terms and app disclosures with what the systems actually do. A notice that says data is used “to personalize your experience” may not be enough if the experience includes the price.
5. Does any of this apply to workers or contractors? If pay offers vary with personal data, that now sits inside the same regulatory question, and it may also trigger employment and state privacy rules.

What a defensible record looks like

The most useful document a company can have here is a plain data flow for pricing: where each input comes from, which system uses it, for what purpose, how long it is kept, and what controls stop sensitive data from being used. Without that map, a company’s answer to a 6(b) order or an attorney general letter will depend on what individual engineers remember, which is not a strong position.

• Keep an inventory of pricing and offer systems, including vendor tools and the data they receive.
• Tag sensitive categories, such as precise location, health signals and inferred financial status, and block them from pricing inputs unless counsel has approved the use.
• Run periodic tests that compare prices across accounts with different profiles, and keep the results.
• Align disclosures with practice, and update them when the pricing model changes, not once a year.
• Assign an owner who can answer for the system end to end.

Captain Compliance works with privacy and legal teams on this kind of groundwork, including data inventories and the privacy notices that have to describe these uses accurately. Our data privacy resources cover the state laws involved. No tool or document makes a pricing practice lawful on its own; the value is in knowing what your systems do before someone else asks.

Frequently asked questions

Is the FTC accusing these nine companies of illegal pricing?

No. The orders are issued under Section 6(b) for a study. The FTC has made no finding of wrongdoing, and several recipients have publicly said they do not set individual prices from personal data.

When are the responses due?

Each company has 45 days from the date it received its order. The FTC said it will use the information for a study; it has not given a date for publishing it.

Can companies refuse to answer?

Not without consequences. The FTC says it can seek to compel compliance in federal court if a recipient fails to turn over the required information.

Does this affect businesses that did not get an order?

Not directly. But the FTC’s August policy statement on Section 5, state laws in Maryland, New York and Connecticut, and California’s investigative sweep all apply more broadly, and the study’s findings are likely to shape future rules.

Is a loyalty discount a form of personalized pricing?

It can be. The FTC specifically named loyalty program participation as a factor it is asking about. Whether a given discount raises legal issues depends on how it is disclosed and what data drives it.

Forty-five days is not long to explain a pricing system to a federal regulator, and for at least some of these companies it will be the first time anyone has had to put the whole thing on paper. The study that follows will tell lawmakers how common personalized pricing really is. It may also tell some executives what their own systems have been doing.