Personalized Pricing Enforcement Policy [What It Changes]

TL;DR

• Personalized Pricing Enforcement Policy is the FTC's August 19, 2026 proposal that a business using personal data to set an individual price must clearly tell the buyer three things: that the price is personalized, what drove that personalization, and which types of personal data were used.

• The Commission is not claiming power to ban personalized pricing outright; skipping those disclosures is what it says is likely an unfair or deceptive act under Section 5 of the FTC Act.

• A 2-truck HVAC shop, a 10-person agency, or a solo clinic is in scope if quote, cart, or booking software uses personal data to change the number the buyer sees.

• Comments run 30 days after Federal Register publication on docket FTC-2026-1057; Maryland and Connecticut food-retail rules take effect October 1, 2026, and New Jersey signed a grocery ban with $50,000-per-violation penalties.

Key Takeaways

• Treat "dynamic pricing" (supply, demand, time of day) as a different bucket from "personalized pricing" (who the buyer is).

• Vague copy such as "specially selected" is called out as inadequate; naming estimated willingness to pay from prior purchases on the same login is closer to adequate if it is accurate.

• State grocery and delivery law is already stricter: New Jersey bans the practice, Maryland restricts large food stores, and Connecticut requires a specific label.

• The checkout path that already has to show total price under 16 CFR Part 464 is the natural place to attach the three facts.

• Audit third-party pricing vendors the same way the FTC's 2024 6(b) orders treated intermediaries: the merchant still owns the customer-facing sentence.

What Personalized Pricing Enforcement Policy means

Personalized Pricing Enforcement Policy is the Federal Trade Commission's proposed statement that, where a buyer reasonably expects a listed price not to vary with personal data, a business that uses that data to set an individual price must disclose the fact of personalization, the basis for it, and the types of data used.

It is not a ban. It is a disclosure duty bolted onto Section 5 of the FTC Act, which already makes unfair or deceptive acts or practices in commerce unlawful.

A 2-truck HVAC shop should care because after-hours quotes that rise when the CRM shows a senior on a fixed income or a remote address are using personal data, not just time-of-day surge. A 10-person agency should care because retainer calculators that lift the number after a prospect's title or site-visit count are doing the same thing on a proposal PDF. A solo clinic should care because cash-pay prices that move with wait-list urgency match the FTC's own grocery examples.

If the price is the same price anyone else would see for the same SKU, slot, or job at that moment, this policy is not the issue. If the price moves because of who they are, the three disclosures are the issue, as of August 19, 2026.

Operators already mapping quote fields through form-to-CRM automation will feel this first: the CRM already holds the personal data the pricing tool is reading.

What happened on August 19, 2026

The Commission announced a Proposed Enforcement Policy Statement Regarding Personalized Pricing and asked for public comment.

According to Holland & Knight, the FTC voted 2-0 to issue the proposed statement. The FTC voted 2-0 on August 19, 2026. Chairman Andrew Ferguson said consumers expect a listed price to be the same price everyone else sees, not a retailer's estimate of willingness to pay, and he repeated that the FTC does not have authority to ban personalized pricing in all circumstances.

According to Holland & Knight, the public comment window is 30 days from Federal Register publication on docket FTC-2026-1057. Comments run 30 days after Federal Register publication. That is a short clock for anyone who needs to describe a live pricing model to the agency.

Consumer Finance Monitor ties the date to a Senate Judiciary subcommittee hearing two weeks earlier and to statutes already on the books in Connecticut, Maryland, and New Jersey. The federal move is disclosure-first. Several states are already in ban-or-label territory.

The legal hook is not new. 15 U.S.C. § 45 already declares unfair or deceptive acts or practices unlawful. The proposed policy is the agency saying how it intends to apply that statute to AI-set, person-specific prices.

The three disclosures

Where consumers reasonably expect that a price will not vary based on personal data, the proposal says a business using personalized pricing should clearly and conspicuously disclose (1) that the price is personalized, (2) the basis for the personalization, and (3) the types of data used.

Failure to provide those three facts is, in the FTC's view as reported by Consumer Finance Monitor, likely a deceptive or unfair practice under Section 5.

Holland & Knight's walkthrough is specific about copy quality. Telling a buyer they have been shown a "specially selected" price is treated as misleading. A disclosure that a personalized price is based on estimated willingness to pay derived from previous purchases from the same retailer through the same login is treated as likely sufficient if it is accurate and complete.

The deception theory is the familiar three-part test: a representation, omission, or practice that is material, likely to mislead a reasonable consumer, and to that consumer's detriment. The unfairness theory is substantial injury, not reasonably avoidable, and not outweighed by countervailing benefits. A higher personalized price can be the injury; concealment can make it unavoidable.

The policy also points at neighboring statutes the FTC already enforces: the Restore Online Shoppers' Confidence Act (15 U.S.C. §§ 8401-8405), the Rule on Unfair or Deceptive Fees at 16 CFR Part 464, and the Fair Credit Reporting Act as an analogy for industries that already disclose when a personal file changes a price.

Part 464 is useful even if you do not sell tickets or lodging. As of the eCFR display dated August 31, 2026, an interactive electronic disclosure must be unavoidable and must not be contradicted by other copy. Personalized-price language that sits behind a tooltip is fighting that definition.

The FTC Endorsement Guides FAQ uses the same "clear and conspicuous" idea for a material connection a significant minority of consumers would not expect. Personalized pricing is that kind of connection between the buyer's data and the number on the screen.

How the FTC got here

On July 23, 2024 the Commission issued 6(b) orders to eight firms that sell surveillance-pricing products: Mastercard, Revionics, Bloomreach, JPMorgan Chase, Task Software, PROS, Accenture, and McKinsey & Co. According to the Federal Trade Commission, the orders went to 8 companies and the vote to issue them was 5-0. The questions covered product types, data inputs, customers, and effects on the prices people pay.

Staff later published research summaries dated January 2025, labeled a staff perspective, not a finding of lawbreaking.

Holland & Knight's April 27, 2026 alert records an Advance Notice of Proposed Rulemaking covering total price disclosure, fee transparency, personalized pricing disclosure, and unauthorized billing. The same firm's August 5, 2026 brief dates that ANPRM to April 14, 2026 and notes a May 18, 2026 letter from 16 state attorneys general, led by New York and Tennessee, asking the FTC to issue a separate surveillance-pricing rule.

NIST's AI Risk Management Framework 1.0 (January 26, 2023, NIST AI 100-1) is a voluntary overlay; it does not replace Section 5. NIST's Center for AI Standards and Innovation is a testing door, not a checkout-copy rule.

Congress and the House letters

On March 5, 2026 the House Committee on Oversight and Government Reform opened an inquiry into AI used to set consumer prices. Chairman James Comer's letters went to Booking Holdings, Expedia Group, Uber, Lyft, and Instacart.

According to the House Oversight Committee, one cited report showed a 221 percent fare gap on the same trip in the same window ($76.82 versus $23.92), and another cited an average 11 percent difference on identical products. One cited Uber pair was $76.82 versus $23.92. Those figures are the committee's cited examples, not an FTC finding.

The Senate Judiciary Subcommittee on Crime and Counterterrorism held "Your Data, Their Profit: The Consumer Cost of AI Surveillance Pricing" on August 4, 2026 in Dirksen 226, with five witnesses.

According to Consumer Finance Monitor, 4 of 5 witnesses backed new legal restrictions and Senator Richard Blumenthal said he and Senator Josh Hawley already had a legislative framework. Hawley also said the FTC should use existing authority — which is what the August 19 proposal does.

S.3387, the One Fair Price Act of 2025, was introduced December 8, 2025 by Sen. Ruben Gallego and remains at introduced on Congress.gov. Those bills are stricter than the FTC's disclosure theory and are not the rule a checkout team can implement this month.

State law is already ahead on groceries

New Jersey's Fair Price Protection Act was signed July 23, 2026. Holland & Knight's August 5, 2026 alert reports penalties of up to $50,000 per violation, treble damages, and a private right of action, with effect about one year after enactment. New Jersey penalties reach $50,000 per violation. The Act covers retail food stores and third-party grocery delivery and preserves loyalty programs, bona fide group discounts, and cost-based differences.

Maryland's Protection From Predatory Pricing Act (House Bill 895) was signed April 28, 2026 and takes effect October 1, 2026. Consumer Finance Monitor reports the food-retail piece applies to establishments of at least 15,000 square feet and to third-party food delivery, with a required merchant disclosure: "THIS PRICE WAS SET BY AN ALGORITHM OR BY USING YOUR PERSONAL DATA."

Connecticut, as described in the August 20 Consumer Finance Monitor piece and the August 5 Holland & Knight brief, requires the label "THIS PRICE WAS INCREASED BY A PRICE SETTING DEVICE USING YOUR PERSONAL DATA," effective October 1, 2026, and its Unfair Trade Practices Act tells courts to be guided by FTC Section 5 interpretations.

According to Consumer Finance Monitor, 11 other states had introduced surveillance-pricing bills in 2026 as of that article. Holland & Knight's August 5 count is broader: more than 40 surveillance-pricing bills pending across 24-plus states, and 28 digital-shelf-label bills across 16 states, with four states enacted and a fifth (New York's One Fair Price Act, passed June 10, 2026) awaiting a governor's signature.

California is not waiting on a grocery-specific ban. The California Attorney General's CCPA page, updated August 28, 2026, restates the covered-business tests: over $25 million gross annual revenue; 100,000 California residents or households; or 50% of annual revenue from selling California residents' personal information. Civil Code Title 1.81.5 requires purpose notice at collection. The California Privacy Protection Agency is the complaint door; Holland & Knight reports AG Bonta has opened investigations treating pricing-related data use as potentially exceeding disclosed purposes.

What the FTC says is in bounds

The proposal does not declare that every different price is unlawful. Insurance and credit are called out as markets where price already reflects individual risk. Ordinary variation from supply, demand, geography, taxes, and other market conditions is distinguished from personalization.

Holland & Knight lists seven illustrative high-risk practices, all "vulnerability or reduced alternatives" patterns, including a food-delivery quote that rises because the buyer cannot leave home, a hotel quote that rises because data indicates funeral travel, and a website price that rises because the buyer is already in the store parking lot.

A small operator does not need a ride-share stack to hit the same pattern. An HVAC after-hours quote that rises because the CRM shows a medical device at the address, or a clinic cash price that rises because the patient is already in the waiting room, is the same theory at a smaller ticket.

Consent is part of the statement. Collecting, using, or disclosing personal data for personalized pricing without adequate disclosures or without consent for that specific purpose may itself violate Section 5. A privacy policy that says "we use data to improve our services" is not a pricing consent.

What a small shop should change

Map every input that can change a quoted or carted price. If the input is inventory, time of day, crew availability, or a published senior/veteran discount, label it dynamic or group-discount and keep the rule in writing. If the input is a person-level field — prior purchases, device, location, household, browsing, inferred income, "likelihood to book" — label it personalized and attach the three facts to the price the buyer sees.

Include vendors. The 2024 6(b) study was aimed at intermediaries. A shop that buys a "smart quote" module is still the business the customer sues or the FTC letters.

Put the disclosure where the price is, not in a footer. Part 464's "unavoidable" test for interactive media is the closest written standard the agency already uses. Teams already routing documents through US Tech Automations workflows can add a disclosure branch when a price is personalized, rather than rebuilding the cart.

Keep the privacy notice and the checkout sentence in the same facts. The state of small-business automation work is the same discipline: write down what the system does before a regulator asks. A 10-person agency that already maps form-to-CRM fields through US Tech Automations can attach the three FTC facts as required fields on the quote object, and operators who log vendor data-use terms in US Tech Automations document routing can reuse that packet for docket FTC-2026-1057.

Do not use "specially selected" unless the basis and the data types are also on the screen. If you sell in New Jersey, Maryland, or Connecticut food retail or delivery, disclosure is not the whole job: those statutes restrict or ban the underlying practice on the dates above. If you only run posted, one-price-for-everyone catalogs, document that.

Home-office operators comparing this to other workflow changes can start at US Tech Automations and the live marketing-agency automation notes; the disclosure is a field and a branch, not a new website.

Workflows that already log who changed a quote — see executive-assistant automation and agency dispatch software — are the audit trail a comment letter will ask for. Plutio-style agency stacks that generate client quotes are the same surface.

Timeline of the pricing-transparency stack

Sources: FTC 6(b) orders; FTC staff summaries; S.3387; House Oversight; Holland & Knight; Consumer Finance Monitor; Senate hearing.

Federal disclosure versus state grocery rules

Sources: Holland & Knight August 21, 2026; 16 CFR Part 464; Holland & Knight August 5, 2026; Consumer Finance Monitor Maryland; S.3387; California AG CCPA.

House Oversight fare example (cited, not an FTC finding)

Source: House Oversight Committee release, March 5, 2026.

USTA analysis

USTA analysis (derived only from the figures cited above): the House Oversight same-trip pair is $76.82 and $23.92. Subtracting gives a $52.90 gap. Dividing $52.90 by $23.92 gives 2.2107, or 221.07%, which is the 221 percent the committee stated. Adding the July 2024 6(b) list (8 companies) to the March 2026 House letter list (5 companies) gives 13 named firms with no overlapping names. That 13 is a count of named counterparties in two public probes, not a liability finding. Holland & Knight's 30-day comment clock starts at Federal Register publication, not at the August 19 announcement, so the close date is still "FR date plus 30 days."

Signal vs Speculation

Demonstrated fact (sourced): On August 19, 2026 the FTC proposed this enforcement policy, voted 2-0, and defined personalized pricing as using personal data to set prices to estimated willingness to pay. The three disclosure elements, the "specially selected" example, the Section 5 theories, the 30-day comment period, and docket FTC-2026-1057 are in the Holland & Knight and Consumer Finance Monitor accounts and on the FTC statement page. The Commission has said it lacks authority to ban personalized pricing in all circumstances. The 2024 6(b) orders, 2025 staff summaries, April 2026 ANPRM, House letters, Senate hearing, New Jersey $50,000-per-violation grocery statute, Maryland and Connecticut October 1, 2026 rules, S.3387, 16 CFR Part 464, ROSCA, FCRA, and the CCPA $25 million / 100,000 / 50% tests are on the linked pages.

Our read: If the comment file does not force a rewrite, small merchants who run person-level quote tools will spend the next 12 months adding three sentences to checkout rather than ripping out pricing software. If Maryland, Connecticut, and New Jersey become the template, grocery and delivery vendors will have to disable person-level prices in those states while other catalogs still personalize — a split a 10-person shop cannot maintain by hand. Over 12–36 months, state UDAP offices that already look to Section 5 are the more likely letter-writers for a local HVAC or clinic than a federal complaint. A federal ban remains speculation: S.3387 is introduced, and the FTC has reserved whether fully disclosed personalized pricing can still be unfair. Do not staff a rebuild on a ban that is not in force. Do staff a disclosure field and a vendor inventory.

FAQs

Does Personalized Pricing Enforcement Policy ban AI pricing?

No. The FTC states it lacks authority to prohibit personalized pricing in all circumstances and is proposing disclosure duties plus aggressive Section 5 enforcement against deceptive or unfair uses.

What three facts must appear next to a personalized price?

That the price is personalized, the basis for that personalization, and the types of personal data used, disclosed clearly and conspicuously where the buyer reasonably expected a non-personalized price.

Does this apply to a shop that only changes prices with demand?

Ordinary supply, demand, geography, and tax differences are distinguished from personalization. The policy is aimed at prices that move because of the buyer's personal data.

When do comments close?

Holland & Knight reports a 30-day window that starts at Federal Register publication, not at the August 19 announcement. Watch docket FTC-2026-1057.

What if we sell groceries in New Jersey, Maryland, or Connecticut?

Those states already restrict or ban person-level grocery and delivery prices on the dates above. A federal disclosure is not a substitute for a state prohibition.

Do we need a new consent checkbox?

The statement says using personal data for personalized pricing without consent for that specific purpose may violate Section 5. A generic "improve our services" privacy clause is a weak match for a price that just moved. Loyalty programs and bona fide group discounts are preserved in New Jersey's statute; they are not a substitute for naming a willingness-to-pay model.

Glossary

• Personalized pricing: Using personal data to set a price to the amount a firm believes a specific buyer will spend.

• Surveillance pricing: Regulator phrase for person-level prices driven by tracking data rather than a posted group rule.

• Dynamic pricing: Price changes driven by inventory, time, or demand, not by who the buyer is.

• Section 5: The FTC Act provision, 15 U.S.C. § 45, that makes unfair or deceptive acts or practices unlawful.

• Clear and conspicuous: A disclosure that is hard to miss; Part 464 requires interactive disclosures to be unavoidable.

• 6(b) order: A compulsory special report under 15 U.S.C. § 46(b), including the July 2024 surveillance-pricing orders to eight intermediaries.

• UDAP: State unfair-or-deceptive-acts statutes; some tell courts to follow FTC Section 5 readings.

• Purpose limitation: Under the CCPA, using collected personal information for a new incompatible purpose without fresh notice.

Inventory which quote, cart, and booking paths read person-level fields, then put the three facts on those paths. For a workflow that already routes documents and CRM fields, open the agentic-workflow builder and add the disclosure branch on the personalized-price step.