In short
- New deadline: The Federal Trade Commission extended the public comment period on its proposed Enforcement Policy Statement Regarding Personalized Pricing by seven days. Comments are now due September 25, 2026, not September 18.
- Docket: The proceeding sits on the public record as FTC-2026-1057. The agency announced the extension on September 3, 2026, without stating a reason or naming a requester.
- What it asks for: Retailers that vary a displayed price using an individual shopper’s personal data would have to disclose that the price is personalized, why it is personalized, and what categories of data feed it.
- What it does not touch: Ordinary supply and demand movement, local market pricing, tax differences and regulatory cost pass-through sit outside the proposal’s scope, as do the long established personalized pricing norms in insurance and credit.
- Why it matters now: A federal policy statement is not a rule, but it lands on top of a hardening state patchwork, with Connecticut and Maryland restrictions arriving October 1 and prohibition bills advancing elsewhere.
What the FTC changed on September 3
The Federal Trade Commission announced on September 3, 2026 that it had extended by seven days the public comment period on its proposed enforcement policy statement covering personalized pricing. The new closing date is September 25, 2026. The original window ran to September 18.
The agency’s notice is unusually spare. It records the extension and the length of it, and it does not identify who asked for more time or explain why the request was granted. That silence is itself a modest signal, because comment extensions in consumer protection dockets are most often granted after trade associations or large filers ask for room to coordinate a submission.
The underlying proposal is not new. The Commission opened the docket on August 19, 2026 on a 2-0 vote, starting the original 30-day comment clock that has now been pushed out. What changed this week is only the calendar, but for retailers weighing whether to file, the calendar is the operative fact.
The docket identifier for anyone tracking submissions is FTC-2026-1057. Comments filed to that docket become part of the public record and are readable by competitors, plaintiffs’ counsel and state attorneys general, which is a consideration in itself for any retailer describing its own pricing stack.
Why a seven-day extension is worth noting
Seven days is a short extension by the standards of federal consumer protection dockets, where 30 and 60 day extensions are common. The brevity suggests the Commission wants to keep the proceeding moving rather than reopen it broadly.
For retailers, that shortens the strategic question. There is no realistic prospect of the deadline sliding into October, so the decision to file or stay silent has to be made in the next two weeks.
What the proposed policy statement actually requires
The proposal is built on Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices in commerce. It does not create a new prohibition. It states how the Commission intends to apply an existing one to a specific pricing technique.
Personalized pricing, in the Commission’s framing, means using personal data about an identified individual to set the price that individual is shown, based on what the seller believes that person is willing to pay. That is distinct from pricing that moves for everyone at once.
The three disclosures
Per the proposed statement as summarized by law firm analyses of the docket, a business that personalizes prices would need to communicate three things clearly. First, that the price shown is personalized rather than uniform. Second, the basis on which it was personalized. Third, the type or types of data used to determine the price.
The third element is the operationally hardest. Naming data categories requires a retailer to know, and to be able to state accurately, which inputs its pricing model consumes, including inputs supplied by third party intermediaries.
Many retailers do not have that mapping in a form they could publish. Pricing models assembled over several years frequently draw on vendor scores whose underlying features are contractually opaque to the retailer buying them.
Deception and unfairness as two separate theories
The proposal advances two distinct routes to liability, and retailers should read them separately because the defenses differ. The deception theory targets representing a price as static or generally available when it is in fact individualized, and it also reaches mischaracterizing personalization, for example presenting a higher personalized price as though it were a loyalty benefit.
The unfairness theory targets the collection or use of personal data for pricing purposes without adequate disclosure or consent. Unfairness under Section 5 carries its own statutory test, including substantial consumer injury that is not reasonably avoidable and not outweighed by countervailing benefits.
The practical effect of pleading both is that a purely cosmetic disclosure fix may not resolve exposure. A retailer could disclose personalization accurately and still face an unfairness argument about the data collection that made it possible.
What the proposal explicitly does not cover
The scope carve-outs matter as much as the requirements, and they are the part most often lost in summary coverage. The Commission draws a line between pricing that responds to market conditions and pricing that responds to a person.
Movement driven by supply and demand falls outside the proposal. So does location-based pricing tied to local market conditions, along with differences created by taxes and by regulatory compliance costs that vary between jurisdictions.
The statement also acknowledges that individualized pricing is a long standing and openly understood norm in insurance and in credit, where risk-based pricing is the product. That acknowledgement is a meaningful limiting principle rather than a throwaway line.
What the proposal does not do is create a safe harbor for loyalty programs, coupons or negotiated prices. Those mechanisms are not named as exempt, which leaves an unresolved question at exactly the point where most retail personalization actually lives.
| Pricing behavior | Inside the proposal’s scope? | Practical read for retailers |
|---|---|---|
| Price moves for all shoppers on supply and demand | No | Conventional dynamic pricing is untouched |
| Different price by store location or region | No | Treated as local market conditions |
| Price differences from taxes or compliance costs | No | Explicitly excluded |
| Risk-based pricing in insurance and credit | Acknowledged as established | Recognized norm, not a target |
| Price set from an individual’s browsing or purchase history | Yes | Core of the proposal |
| Higher personalized price framed as a loyalty discount | Yes | Named as a deception risk |
| Loyalty offers, coupons, negotiated prices | Not addressed | No stated safe harbor, the open question |
Where the evidence base came from
The proposal does not arrive from nothing. It rests on a study the Commission has been running since 2024, and the study’s findings supply most of the factual predicate in the docket.
In July 2024, the FTC issued orders to eight intermediary firms that build and sell pricing products, requiring them to describe how those products work. The agency released preliminary observations from that study on January 17, 2025.
What the intermediaries described
Staff reported that the firms examined worked with at least 250 clients selling goods and services, spanning grocery retailers through apparel sellers. That client count is the most cited number in the record because it establishes reach rather than isolated experimentation.
The behavioral inputs described were granular. Staff found that signals ranging from mouse movements on a page to the specific products a shopper abandons in a cart can be tracked and used to tailor what that shopper is charged.
One example in the findings involved a cosmetics company targeting promotions by skin type and skin tone. That illustration has done a large amount of work in the policy debate, because it moves the discussion from abstract data use toward characteristics that map onto protected categories.
Commenting on the study when it was launched in July 2024, then Chair Lina M. Khan said Americans deserve to know whether businesses are using detailed consumer data to deploy surveillance pricing. The agency’s own research framing put it more plainly, stating that consumers expect prices to change based on supply and demand, not their web surfing habits or buying history.
A record with a dissent in it
The evidentiary base is not unanimous. When the preliminary findings were published in 2025, then Commissioner Andrew Ferguson and Commissioner Melissa Holyoak dissented from their release, characterizing the publication as premature and reflecting incomplete early impressions.
Ferguson now chairs the Commission and voted to open this docket. That progression, from dissenting on the study’s publication to advancing an enforcement policy built partly on it, is a detail commenters are likely to probe.
The Chairman’s framing of the harm has been consistent and narrow. When consumers see a listed price, Ferguson said, they expect it to be the same price everyone else sees, not the retailer’s estimate of how much they are willing to pay based on their personal data.
How personalized pricing is actually built
Understanding the compliance problem requires understanding the plumbing, because the disclosure the Commission proposes is a statement about a system most retailers do not fully own. Very little individual-level pricing is built end to end inside a retailer.
The structure the FTC study examined is a three layer one. A retailer sits at the top, an intermediary vendor supplies the pricing or offer logic, and a wider data supply chain feeds that vendor with behavioral and demographic signals.
That is why the agency issued its 2024 orders to intermediaries rather than to retailers. The firms that build the products had the fullest view of how the technique works in practice, and the study’s reach figure of at least 250 client businesses came from that layer rather than from surveying merchants.
Why the data-type disclosure is hard to write
A retailer asked to name the data types feeding its pricing must ask its vendor, and vendor answers are frequently constrained by the vendor’s own commercial position. Feature lists inside a scoring product are treated as proprietary.
The result is a documentation gap that is contractual rather than technical. The retailer holds the customer relationship and the legal exposure, while the party that can accurately describe the inputs has an incentive not to.
Closing that gap generally means reopening vendor agreements to add disclosure and audit rights. That is slow work, and it is the reason the September 25 filing question and the engineering question are on different timescales.
Segments, individuals and the line between them
Most retail personalization operates on segments rather than on named individuals, and retailers often assume segment-level targeting sits outside the proposal. That assumption deserves care.
The Commission’s framing turns on using personal data to estimate what a person will pay, and a segment of one, or a segment narrow enough to function as one, does not obviously escape that description. The cosmetics example in the study findings was a segment-level practice.
Where a segment is defined by attributes drawn from an individual’s own behavior, the distinction between segment and individual pricing becomes difficult to defend in a disclosure. That is a scope question the docket has not resolved, and one worth raising in a comment.
How the federal proposal sits against the state patchwork
The FTC is not moving into empty territory. States have been legislating on this for roughly a year, and several of those measures are already operative or about to be.
New York moved first on disclosure. Its Algorithmic Pricing Disclosure Act, passed on November 10, 2025, requires covered sellers to display a specific notice stating that the price was set by an algorithm using the shopper’s personal data, with penalties reported at up to $1,000 per violation.
Prohibition followed disclosure. New Jersey enacted a surveillance pricing ban rather than a labeling requirement, and Connecticut and Maryland restrictions are next on the calendar. Those two take effect October 1, 2026, which puts them roughly a week after the FTC docket closes.
California has pursued both tracks at once. Attorney General Rob Bonta announced an investigative sweep on January 27, 2026, sending letters to businesses with significant online presence across retail, grocery and hotels, while Assembly Bill 2564, introduced February 20, 2026, would prohibit surveillance pricing outright with civil penalties reported at up to $12,500 per violation and treble that amount for intentional conduct.
Disclosure states versus prohibition states
The split between the two models is the central compliance problem. A disclosure regime and a prohibition regime cannot be satisfied by the same build, because one requires a label on a practice the other forbids entirely.
The FTC proposal is a disclosure instrument. It tells retailers how to personalize lawfully rather than telling them not to, which means federal compliance does not resolve exposure in prohibition states.
New York illustrates the tension inside a single jurisdiction. Trackers report that its legislature passed a One Fair Price Act in June 2026 that would prohibit using personal data driven algorithms to charge different customers different prices, going substantially further than the disclosure regime the same state adopted seven months earlier.
Industry trackers count roughly two dozen states that have moved to regulate algorithmic or surveillance pricing in some form during 2026. That figure includes introduced bills as well as enacted law, so it should be read as a measure of legislative attention rather than of live obligations.
| Jurisdiction | Instrument | Model | Status or effective date | Reported penalty |
|---|---|---|---|---|
| Federal (FTC) | Enforcement policy statement | Disclosure | Proposed, comments close September 25, 2026 | Section 5 remedies |
| New York | Algorithmic Pricing Disclosure Act | Disclosure | Passed November 10, 2025 | Up to $1,000 per violation |
| New York | One Fair Price Act | Prohibition | Passed legislature June 2026 | Not confirmed |
| New Jersey | Surveillance pricing ban | Prohibition | Enacted | Not confirmed |
| Connecticut | SB 4 | Restriction | Effective October 1, 2026 | Not confirmed |
| Maryland | HB 895 | Restriction | Effective October 1, 2026 | Not confirmed |
| California | AB 2564 | Prohibition | Introduced February 20, 2026 | Up to $12,500, treble if intentional |
What a policy statement can and cannot do
A policy statement is not a rule. It does not go through notice and comment rulemaking under the Administrative Procedure Act in the way a trade regulation rule does, and it does not create independently enforceable obligations or civil penalties of its own.
What it does is announce how the agency reads an existing statute. That has real consequences even without the force of law, because it tells the market what conduct the Commission will treat as actionable and it gives staff a published framework to cite when opening an investigation.
It also shapes litigation outside the agency. Plaintiffs’ firms and state attorneys general routinely borrow federal enforcement framings, and state unfair and deceptive practices statutes in many states are interpreted with reference to Section 5 jurisprudence.
The Commission has been willing to litigate pricing presentation questions directly, as seen in its case against Amazon over hidden ad auction surcharges. A policy statement is the cheaper instrument, but the enforcement appetite behind it is not theoretical.
There is a third audience beyond retailers and staff. Publishing a framing tells the vendor layer what its retail customers will start demanding contractually, which tends to move faster than enforcement does.
The durability question
Policy statements can be withdrawn by a later Commission more easily than rules can be repealed. That cuts both ways for retailers planning multi-year investment in pricing infrastructure.
The state laws, by contrast, do not depend on the composition of a federal commission. For most retailers, the state calendar is the harder constraint and the one that should drive engineering sequencing.
What retailers should do before September 25
There are two separate decisions in front of retail legal and pricing teams, and conflating them wastes the remaining time. One is whether to file a comment. The other is what to fix regardless of the docket’s outcome.
The filing decision
Filing is worth considering where a retailer has a concrete scope problem that the proposal leaves unresolved. The loyalty and coupon question is the strongest candidate, because a targeted offer to a segment sits uncomfortably between the personalization the Commission is describing and the promotional practice every grocer runs.
Comments that describe a specific mechanism and ask for a clear scope line tend to land better than general objections. Trade associations will file at a level of abstraction that individual retailers can usefully supplement.
The countervailing consideration is disclosure risk. Anything submitted becomes public, and a detailed description of a personalization stack is discoverable material in exactly the states now moving toward prohibition.
The audit work that is worth doing either way
The compliance steps recommended in law firm analyses of the docket are largely independent of whether the statement is finalized as drafted. They also overlap heavily with what the October 1 state restrictions will require.
- Map every point where consumer-specific data affects a displayed or charged price, including prices set by third party vendors.
- Identify the data sources feeding those models and confirm each is covered by an existing consent framework.
- Audit customer-facing pricing language for accuracy, with particular attention to anything describing a price as a discount, a member rate or a loyalty benefit.
- Put controls, testing and complaint handling around pricing models rather than treating them as pure merchandising tools.
- Assess where jurisdictional differences require different behavior, not just different copy.
Step three is where most immediate exposure sits. A personalized price presented as a loyalty discount is the fact pattern the deception theory names most directly, and it is common in production because it tests well.
Which retail categories carry the most exposure
Exposure is not evenly distributed. It tracks how much individual-level data a category holds and how routinely it varies prices at the individual level.
Grocery sits at the front, which is why grocery retailers appear in both the FTC study’s client descriptions and the California sweep. Loyalty penetration in grocery is high, purchase histories are long and dense, and personalized offers are standard practice.
Online-native apparel and beauty follow, for the reason the cosmetics example in the study illustrates. These categories collect attribute-level personal data and have mature offer personalization systems built on it.
Travel and hospitality are exposed through the same mechanics, and hotels were named alongside retail and grocery in the California sweep letters. General merchandise marketplaces sit in a more complex position, because much of the price variation on a marketplace comes from independent sellers rather than the operator.
Grocery: high loyalty penetration, long histories
Grocery’s exposure comes from the depth of its data rather than from aggressive pricing. A weekly shop generates a purchase record that is unusually complete, covering categories from household staples to pharmacy adjacent items.
The category also runs the densest promotional machinery in retail. Personalized offers delivered through an app are standard, which places grocery squarely in the zone where a targeted price could be read as a personalized one.
The presentational risk is acute here for a specific reason. Grocery offers are almost always framed as savings, and the deception theory names precisely the case where a personalized price is presented as a discount.
Marketplaces: whose price is it
On a marketplace, the operator sets fees, ranking and sometimes buy box logic, while third party sellers set nominal prices. Responsibility for a personalized outcome is therefore genuinely divided.
That division does not obviously provide a defense. Where an operator’s own systems determine which offer a shopper sees, and that determination draws on the shopper’s personal data, the operator is participating in setting the effective price.
Marketplace operators are also the most likely to have the required data mapping already, because platform economics force them to instrument their own ranking systems. The disclosure burden may be lighter there than the legal exposure.
Apparel and beauty: attribute-level data
These categories collect data that describes the shopper rather than only the shopper’s behavior, including size, fit and, in beauty, skin characteristics. That is a different order of sensitivity from browsing history.
The cosmetics illustration in the study findings landed hard for that reason. Once pricing draws on attributes that correlate with protected characteristics, the analysis extends beyond consumer deception into discrimination exposure that no disclosure resolves.
What happens after the docket closes
Closing the comment period does not start a fixed clock. The Commission may finalize the statement, revise it, or leave it pending, and it is under no statutory deadline to act by a particular date.
The more likely near-term development is enforcement activity that reflects the framing whether or not the statement is finalized. A published theory of liability is usable by staff immediately.
Retailers should therefore watch the October 1 state effective dates more closely than the federal docket. Those dates create obligations on a fixed schedule, and the Connecticut and Maryland restrictions arriving October 1 will bite before any federal finalization plausibly could.
How a case would actually be proven
Proving individualized pricing from the outside is harder than proving most deception cases, which is part of why the agency went to intermediaries for evidence rather than to shoppers. A single consumer cannot observe the price another consumer was shown.
That asymmetry points to where the evidence will come from. Investigations are likely to run on compulsory process against retailers and their vendors, seeking model documentation, and on comparative testing that shows different prices for the same item across controlled profiles.
Retailers should assume their own internal experimentation records are the most probative material in the file. A pricing test that measured willingness to pay by cohort documents both the practice and the intent behind it in a single artifact.
This is also where the disclosure obligation interacts awkwardly with commercial confidentiality. A retailer that documents its inputs well enough to disclose them accurately has also assembled a clean evidentiary record, which is a genuine tension with no clean resolution in the current draft.
The broader direction is not seriously in doubt. Between federal framing and the spread of state surveillance pricing restrictions, undisclosed individual-level pricing is moving from a competitive tactic toward a regulated one across the US market.
The official notice of the extension is published on the Commission’s site for anyone filing to the docket: FTC extends public comment on personalized pricing.
Frequently asked questions
When is the FTC personalized pricing comment deadline?
September 25, 2026. The Commission announced on September 3, 2026 that it had extended the original September 18 deadline by seven days. Comments go to docket FTC-2026-1057.
Is the policy statement a law retailers must follow?
No. It is a proposed enforcement policy statement, which describes how the FTC intends to apply Section 5 of the FTC Act to personalized pricing. It does not create new obligations or its own penalties, but it signals what conduct the agency will treat as actionable.
Does this ban personalized pricing?
No. The federal proposal is a disclosure instrument. It would require retailers to say that a price is personalized, why, and what data types feed it. Several states, including New Jersey, have taken the different approach of prohibiting the practice.
Does normal dynamic pricing fall under this?
Not as drafted. Prices that move on supply and demand, local market conditions, taxes or regulatory compliance costs are outside the proposal’s stated scope. The trigger is using an identified individual’s personal data to set what that person is charged.
Are loyalty programs and coupons exempt?
Not explicitly. Analyses of the docket note that no specific safe harbor exists for loyalty programs, coupons or negotiated prices. That gap is the single strongest reason for a retailer to file a comment before September 25.
What did the FTC’s surveillance pricing study actually find?
The preliminary findings, released January 17, 2025, described intermediaries working with at least 250 client businesses, from grocery to apparel. Staff reported that inputs as granular as mouse movements and abandoned cart contents can be used to tailor pricing.
Which state rules take effect soonest?
Connecticut and Maryland restrictions take effect October 1, 2026, roughly a week after the federal comment period closes. New York’s disclosure act passed in November 2025, and California’s AB 2564 prohibition bill was introduced in February 2026.
Can complying with the FTC proposal satisfy state requirements?
No, and assuming otherwise is the main compliance trap here. Federal disclosure compliance does not help in states that prohibit the practice outright, because a label cannot cure a ban.
What should a retailer do first if it has limited time?
Audit customer-facing pricing language for anything that presents a personalized price as a loyalty benefit, member rate or discount. That fact pattern is named directly in the deception theory and is common in live retail environments.