The Federal Trade Commission’s public record on personalized pricing closed on Friday, September 25, 2026, and the filings now on the docket describe two irreconcilable views of what a price is allowed to know about the person paying it.
On one side sit technology and advertising trade groups who say the agency’s proposed enforcement policy statement is so loosely drawn that it would force a disclosure onto almost every discount in American retail. On the other sit consumer organizations who say the proposal does not go nearly far enough, because disclosure alone leaves discriminatory pricing intact as long as it is announced. Between them, a litigation shop has already laid down a First Amendment marker.
The same day the record closed, Walmart’s chief executive published a letter promising customers the company will never set prices from their personal data. That timing was not accidental.
In short
- The comment window closed September 25, 2026 on the FTC’s proposed enforcement policy statement regarding personalized pricing, docket FTC-2026-1057, after a seven-day extension granted on September 3.
- The proposed standard is disclosure, not prohibition. The FTC concedes it lacks authority to ban personalized pricing outright, so the statement targets sellers who fail to say clearly that a price was built from the buyer’s data.
- Industry called the standard unworkably vague. The ANA, the 4A’s and the AAF warned it could reach a free-shipping coupon for an abandoned cart, while the Washington Legal Foundation framed the disclosure duty as compelled speech.
- Consumer groups wanted more. The Consumer Federation of America pressed the FTC to keep using the Equal Credit Opportunity Act, arguing pricing data acts as a proxy for race, age and income.
- Walmart moved first commercially. CEO John Furner’s September 25 letter pledged “we price the product, not the person,” extending the promise to the Sparky assistant and to digital shelf labels.
What exactly closed on September 25
The FTC announced the proposed enforcement policy statement on August 19, 2026, and put it out for public comment rather than adopting it immediately. The Commission approved the release on a 2-0 vote, a tally that reflects how thinly staffed the agency’s top table has become. The original deadline was September 18.
On September 3 the Commission extended that window by seven days, moving the close to September 25. The extension was modest by the standards of what had been requested: the National Association of Convenience Stores had asked for 60 additional days. The docket number for the proceeding is FTC-2026-1057, and the underlying matter carries the internal number P034101.
A policy statement is not a rule. It does not create new obligations by itself, and it is not subject to the procedural requirements that govern formal rulemaking under the Magnuson-Moss framework. What it does is tell the market how the agency reads existing law, which in practice sets the baseline for enforcement risk.
That distinction is the crux of most of the objections filed. Several commenters argued that if the FTC intends to impose a disclosure duty on the entire retail sector, the correct instrument is a rule with a full record, not a statement of enforcement intent.
What the proposed standard actually requires
The FTC defines personalized pricing as the use of personal data to set a price at the amount the company believes a particular consumer is willing to spend. The agency grounds its authority in Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices, and it builds two separate theories on that base.
Chairman Andrew Ferguson framed the consumer expectation plainly when the proposal landed in August. Consumers who see a listed price expect it to be “the same price that everyone else sees, not the retailer’s estimate of how much they are willing to pay based on their personal data,” he said. The proposal the FTC published in August is the agency’s first formal articulation of that position.
The three elements a disclosure must carry
Under the proposed statement, a disclosure must be clear and conspicuous, and it must convey three things. First, that the price is personalized. Second, the basis on which it was personalized. Third, the types of data used to do it.
The agency explicitly rejected soft euphemisms. Telling a shopper they have received a “specially selected” price does not satisfy any of the three elements, because it discloses neither the mechanism nor the inputs. That rejection matters commercially, because “just for you” language is standard furniture in retail apps.
The FTC did offer a model of what adequate looks like. A disclosure explaining that “a personalized price is based on a consumer’s estimated willingness to pay derived from data about that consumer’s previous purchases from the same retailer through the same login account” would likely clear the bar. Read closely, that sentence is a confession, and that is precisely why industry objects to it.
The deception theory
The deception route is the more conventional of the two. A seller deceives when it represents a price as static or widely offered while in fact tailoring it, when it stays silent about personalization in a context where shoppers reasonably believe prices are standard, or when it misstates the basis for a price difference.
This theory needs no showing of harm beyond the misrepresentation itself. It also travels well: the same logic the FTC has applied for years to drip pricing and undisclosed fees maps onto a price built from behavioural data.
The unfairness theory
The unfairness route is more aggressive and more contested. The FTC argues that personalized pricing which results in a higher price can be unfair when the consumer could not reasonably have avoided that price, because the personalization was concealed from them.
Concealment is doing the work in that sentence. If a shopper does not know the price moved, they cannot shop around, clear their history, or log out to test the alternative. The FTC treats that inability to avoid as the substantial injury the unfairness test requires.
The six examples the FTC put on the record
The proposed statement is unusually concrete about what it is aiming at. The examples the agency listed share a single thread: each uses data that signals a specific vulnerability rather than a general preference.
- A food delivery company charging more when its data suggests the customer cannot leave the house.
- A grocery chain pricing milk higher for households whose composition implies children.
- A hotel charging more for travel it can identify as funeral-related.
- A rideshare service pricing trips to medical facilities higher, or raising fares when no competitor app is detected on the handset.
- A retailer pricing home-security systems higher for customers who appear in crime victimization records.
- A website setting prices from a shopper’s physical location inside a store while they browse.
None of these are hypotheticals invented for a policy document. Each maps to a practice that has been reported, litigated, or marketed as a feature by pricing-technology vendors over the past three years. Listing them was a deliberate act of signalling.
The examples also reveal the limit of the FTC’s ambition. Every one of them would remain lawful under the proposed statement if the seller disclosed it properly, which is the concession that drew the sharpest criticism from the consumer side of the docket.
Where industry drew its line
The trade association filings converge on a single complaint: the FTC never defined the consumer expectation its whole theory depends on. Without parameters for assessing what a shopper reasonably believes about a price, the objection runs, the standard becomes whatever an enforcement attorney later says it was.
The Association of National Advertisers, the American Association of Advertising Agencies and the American Advertising Federation filed jointly. They argued the proposal would “create confusion and uncertainty in the business community and likely will deprive consumers of advertising offers they want and use,” and that its “vague enforcement standards” would chill responsible practices.
The most commercially loaded argument is about scope. The advertising groups pointed to a personalized coupon for free shipping sent to a customer who abandoned a cart, and asked whether that now requires a disclosure explaining that the offer was generated from browsing data.
If the answer is yes, the practical consequence is over-notification. Retailers would attach data-use language to essentially every price a shopper encounters, which the filings argue would “dilute the effectiveness of appropriate disclosures” until nobody reads them. The same reasoning underpins arguments about loyalty-led personalized offers now being built into major chains’ 2027 marketing plans.
The Computer and Communications Industry Association made the point more narrowly. Krisztian Katona, its vice president of global competition and regulatory policy, said the FTC “should ensure that its proposed policy statement does not inadvertently discourage discounts and loyalty programs that benefit consumers.”
CCIA asked the agency to separate personalized price increases from discounts, coupons, loyalty rewards and membership benefits, and to apply case-specific deception and unfairness analysis rather than treating price variation as harmful on its own. That is a request for the burden of proof to stay where it has traditionally sat.
A second industry ask is about allocation. CCIA urged the FTC to confirm that any disclosure duty rests with the entity that determines the challenged price, not with platforms and vendors that supply pricing tools and infrastructure to sellers.
For marketplaces, this is the whole argument. A third-party seller on a marketplace sets its own price using tools the marketplace provides, and the two parties have very different compliance capacity. NetChoice, whose comments were filed on September 25 by director of policy Patrick Hedger, pressed related concerns about the proposal’s reach.
The constitutional challenge already taking shape
The filing that most clearly telegraphs litigation came from the Washington Legal Foundation. It attacks the proposal on two constitutional fronts rather than on policy grounds.
Compelled speech
The first is the First Amendment. WLF characterises a mandated disclosure about algorithmic pricing as compelled speech, arguing that requiring a business to announce how it arrived at a price forces it to say something it would not otherwise say.
That argument has been tested before, and it lost. The National Retail Federation raised substantially the same theory against New York’s Algorithmic Pricing Disclosure Act in the Southern District of New York, and the court declined to block the statute. The compelled disclosure was “plainly factual,” the court reasoned, and was not rendered controversial merely because the regulated entity would prefer not to make it.
That precedent does not settle the federal question. A state statute with a fixed disclosure text is a different target from a federal enforcement policy that leaves the wording to the seller, and the two cases would be litigated in different postures.
Regulation by litigation
WLF’s second line of attack is procedural due process. By announcing an enforcement posture instead of promulgating a rule, the argument runs, the FTC gets the practical effect of a regulation without the record, the cost analysis, or the judicial review that a rule would attract.
Zac Morgan, WLF senior litigation counsel, put the substantive defence of the practice alongside the procedural objection. “Individualized pricing helps markets clear efficiently and increases consumer bargaining power,” he said. “It shouldn’t be discouraged through unfounded regulation-by-litigation.”
The procedural argument is the more dangerous of the two for the agency, because it does not require a court to decide that personalized pricing is good. It only requires a court to decide the FTC took a shortcut.
What consumer groups asked for instead
The consumer filings treat disclosure as the floor rather than the ceiling, and several argue the FTC is retreating from tools it already has.
The Consumer Federation of America made the discrimination case most directly. “Discrimination is the most pernicious consequence of surveillance pricing,” its comment argued. “Data used to set prices can easily stand in for race, age, income, and other protected characteristics.”
CFA, whose competition and market fairness work is directed by Emily Peterson-Cassin, urged the Commission to keep deploying the Equal Credit Opportunity Act against discriminatory pricing rather than narrowing itself to a disclosure theory. That framing moves the question from consumer protection into civil rights enforcement, where remedies are stronger.
The Open Markets Institute filed on the same day, urging the FTC to treat AI-enabled pricing as an exploitation problem rather than an information problem. The Center for American Progress published its own set of recommendations the day before the deadline. The News/Media Alliance and the Lexington Institute also entered comments during the final week, as did the American Action Forum.
The political layer arrived a week earlier. On September 18 the House Oversight Committee pressed the FTC to explain how its proposed disclosure requirements would work in practice, a request that puts congressional attention on the mechanics rather than on the principle.
How the record breaks down
The filings sort into three camps rather than two, because the constitutional challengers want something different from the trade associations. Industry groups want the standard narrowed and clarified. The litigation shops want it withdrawn. Consumer groups want it widened.
| Commenter | Type | Core ask of the FTC |
|---|---|---|
| ANA, 4A’s, AAF (joint) | Advertising trade | Define the consumer expectation standard; exclude routine discounts and coupons |
| CCIA | Tech trade | Separate price increases from discounts; place the duty on whoever sets the price |
| NetChoice | Tech trade | Narrow the proposal’s reach across platforms and third-party sellers |
| Washington Legal Foundation | Litigation | Withdraw it: compelled speech plus due process defects |
| NACS | Retail trade | Asked for 60 more days rather than the seven granted |
| Consumer Federation of America | Consumer | Keep using the ECOA; treat proxy discrimination as the core harm |
| Open Markets Institute | Consumer | Move beyond disclosure to substantive limits on AI pricing |
| Center for American Progress | Policy | Broaden the enforcement toolkit against AI-driven pricing |
The asymmetry worth noting is that the industry and consumer camps agree on one thing: a disclosure-only standard is unstable. They disagree about which direction it should collapse.
Walmart’s letter and the corporate hedge
Walmart chose the day the record closed to publish a commitment the FTC has not asked for. CEO John Furner addressed a public letter to customers and Sam’s Club members ruling out personalized pricing outright.
“We price the product, not the person,” Furner wrote. “Using someone’s income, shopping history or moment of need to charge them more would violate the EDLP promise our business model is built on.” Asked about dynamic personalized pricing, his answer was reported as three words: we won’t do it.
The letter extends the pledge to two technologies that have drawn the most suspicion. On the Sparky assistant, Furner wrote that engaging with it “is an invitation to serve you better, not to use your personal information to set a personalized price.” On AI generally, he wrote that the company has “never used the relationships our associates have with customers to charge more, and we won’t do that with AI.”
The most operationally significant part of the letter concerns electronic shelf labels. Furner framed them as a consistency and labour tool: they keep the shelf price aligned with what rings up at checkout and stop associates replacing paper tags by hand.
That framing is a direct response to the legislative record. Multiple state bills now treat electronic shelf labels as surveillance pricing infrastructure and impose moratoria on new deployments, which puts a large capital programme at risk on a theory about what the hardware might later enable.
A voluntary pledge is not a legal commitment, and it can be revised. But it is admissible in a consumer protection case, and a retailer that publishes “we price the product, not the person” has handed regulators a deception hook if the practice ever diverges. That is the calculated part of the move.
The state patchwork retailers already face
Whatever the FTC finalises, it lands on top of state law that is already in force and already inconsistent. Four states have enacted statutes, and the drafting approaches differ enough that a national retailer cannot satisfy them with one template.
| Jurisdiction | Approach | Scope | Enforcement | Status |
|---|---|---|---|---|
| New York, Algorithmic Pricing Disclosure Act | Mandated disclosure text | Any algorithmic pricing using personal data | Attorney General only | In force since November 10, 2025 |
| Maryland, Protection from Predatory Pricing Act | Prohibition with limited exceptions | Large food retailers, third-party food delivery | Attorney General only | Enacted |
| Connecticut, omnibus privacy law | Hybrid: disclosure plus ban | Retail sellers, food delivery services | Attorney General only | Prohibition generally from July 1, 2027 |
| New Jersey, Fair Price Protection Act | Prohibition with cost-based and bona fide discount exceptions | Groceries and foodstuffs | AG plus a private right of action | Signed July 23, 2026; enforcement from August 1, 2027 |
| New York, One Fair Price Act | Prohibition beyond grocery | Broader retail | To be set | Passed both chambers June 2026; awaiting signature, effective 180 days after |
| California, AB 2564 | Prohibition on surveillance-derived custom prices | Retailers using personally identifiable information | To be set | Passed Assembly May 2026; pending in Senate |
| Washington, HB 2481 and SB 6312 | Grocery ban, surge limits, ESL moratorium | Groceries and essential goods | AG plus private right of action | Companion bills pending |
| Seattle, CB 121267 | Municipal ban | Larger grocers, delivery platforms | City enforcement | Passed September 22, 2026 |
New York requires a specific sentence: “THIS PRICE WAS SET BY AN ALGORITHM USING YOUR PERSONAL DATA.” Connecticut requires a different one: “THIS PRICE WAS INCREASED BY A PRICE SETTING DEVICE USING YOUR PERSONAL DATA.”
Those two sentences carry different meanings. New York’s is neutral about direction, while Connecticut’s asserts that the price went up. A retailer operating in both states cannot use a single string, and the FTC’s three-element framework matches neither. Our earlier coverage tracked how state surveillance pricing bans began taking effect with these divergent label requirements.
New Jersey is the first state to attach a private right of action, and violations also count as violations of the state Consumer Fraud Act. Consumers can pursue treble damages plus attorneys’ fees and costs. The Attorney General can recover actual damages or $50,000 per violation, whichever is greater.
That structure converts a compliance question into a class action exposure. It also brings an electronic shelf label provision: a one-year moratorium on new deployments beginning February 1, 2027.
The wider legislative picture explains why general counsels are treating this as a 2027 problem rather than a 2026 one. More than 40 personalized pricing bills have moved across roughly 24 states in 2026, already exceeding the whole of 2025. Colorado’s governor vetoed one of them, HB 26-1210, in June. California separately amended the Cartwright Act through AB 325 and SB 763, effective January 1, 2026, to reach common pricing algorithms used as part of a restraint of trade.
Municipal action has now started too, with Seattle’s grocery surveillance pricing ban clearing the city council on September 22, three days before the federal comment record closed.
What happens next
The FTC is now in the review phase. It can finalise the statement as drafted, revise it in response to the record, convert the exercise into a formal rulemaking, or let it sit. There is no statutory deadline forcing any of those outcomes.
The most likely path, on the shape of this record, is a revised statement with clearer carve-outs for discounts and loyalty programmes. That would cost the agency little, since the FTC has already conceded it cannot ban disclosed personalization, and it would remove the single objection that appears in nearly every industry filing. Whether the standard then hardens into practice is the question our earlier analysis of whether the framework would become the de facto US standard examined in detail.
A gap runs through the whole proceeding. The proposed statement addresses sellers who personalise prices, but it says little about AI shopping agents that negotiate or transact on a consumer’s behalf, and nothing about who discloses what when an agent is the party seeing the price.
Ferguson addressed adjacent ground at a Reuters conference on September 25, suggesting that AI developers should bear liability for the conduct of their agents and rejecting the idea that agents act on their own account. That is a coherent position, but it is not in the pricing document, and agentic commerce is scaling faster than the docket is moving.
What retailers should be doing now
Three steps are defensible regardless of how the FTC lands. First, inventory which pricing inputs are personal data: login-linked purchase history, device signals, location, inferred household composition. The FTC’s examples show that the sensitivity of the input, not the sophistication of the model, determines exposure.
Second, separate the discount stack from the price stack in system architecture. Nearly every industry filing argues that coupons and loyalty rewards should sit outside the disclosure duty, and a retailer that cannot demonstrate the separation technically will struggle to claim it legally.
Third, treat the state texts as the binding constraint and the FTC framework as the overlay. New York and Connecticut already mandate specific wording, New Jersey adds private litigation from August 2027, and none of that waits for a federal policy statement to be finalised.
Frequently asked questions
Is personalized pricing now illegal in the United States?
No. The FTC’s proposed enforcement policy statement is explicit that the agency lacks the authority to ban personalized pricing in all circumstances. Its theory targets sellers who conceal the practice, meaning a fully disclosed personalized price would remain lawful under federal law. Several states go further and prohibit the practice outright in grocery and food delivery.
What is the difference between dynamic pricing and personalized pricing?
Dynamic pricing moves a price for everyone in response to conditions such as demand, time of day or inventory. Personalized pricing moves the price for one individual based on data about that person. The FTC’s proposal is aimed squarely at the second, though the examples it lists show the boundary blurs when location or device data is involved.
What must a disclosure say under the FTC’s proposed standard?
It must be clear and conspicuous and must convey three things: that the price is personalized, the basis for the personalization, and the types of data used. The FTC expressly rejected phrases such as “specially selected” as insufficient because they disclose neither the mechanism nor the inputs.
Do loyalty programmes and coupons trigger the disclosure duty?
That is the central unresolved question in the docket. The advertising and technology trade groups asked the FTC to state clearly that discounts, coupons, loyalty rewards and membership benefits fall outside the duty, warning that otherwise a free-shipping coupon for an abandoned cart could require a data-use notice. The proposal as drafted does not draw that line sharply.
Could the policy statement be struck down in court?
A challenge is likely. The Washington Legal Foundation argues the disclosure duty is compelled speech under the First Amendment and that proceeding by policy statement rather than rulemaking raises due process problems. The First Amendment theory already failed once against New York’s disclosure statute, where the court found the mandated text plainly factual, but the procedural argument is untested against this instrument.
Which states already regulate surveillance pricing?
New York, Maryland, Connecticut and New Jersey have enacted statutes, using a mix of mandated disclosure and outright prohibition. New York’s disclosure law has been in force since November 10, 2025. Connecticut’s prohibition generally applies from July 1, 2027, and New Jersey’s enforcement begins August 1, 2027. Seattle passed a municipal grocery ban on September 22, 2026.
Why does New Jersey’s law matter more than the others?
It is the first to include a private right of action, and violations also constitute violations of the state Consumer Fraud Act. Consumers can seek treble damages plus attorneys’ fees, and the Attorney General can recover actual damages or $50,000 per violation, whichever is greater. That converts a regulatory compliance exercise into class action exposure.
What did Walmart actually commit to?
In a September 25 letter to customers and Sam’s Club members, CEO John Furner said Walmart will not use income, shopping history or moment of need to set prices, summarised as “we price the product, not the person.” The pledge extends to the Sparky assistant and to digital shelf labels, which he framed as tools for price consistency and labour savings rather than personalization.
When will the FTC decide?
No statutory deadline applies. The Commission is reviewing the submissions and can finalise, revise, convert the exercise into a formal rulemaking, or take no further action. Given how consistently the record asks for clearer discount carve-outs, a revised statement is the most probable next step, and the state statutes will bind retailers in the meantime regardless.
The FTC’s proposed enforcement policy statement and the full public docket are available on the Commission’s announcement page.