What we found
The Federal Trade Commission announced on 19 August 2026 that it is seeking public comment on an enforcement policy statement regarding personalized pricing, which it describes as the use of personal data to set prices according to what a company believes an individual consumer is willing to spend. FTC Chairman Andrew Ferguson is quoted saying that consumers expect a listed price to be the same price everyone else sees, and that businesses which fail to tell consumers how their personal data is being used to set a price may be in violation of the FTC Act and other laws the agency enforces. According to the announcement, the statement notes that retailers who represent or imply that a price is static when it in fact varies by individual are at risk of misleading customers, and that undisclosed collection or use of personal data for personalized pricing could violate the FTC Act's ban on unfair or deceptive practices. The announcement says informed consumers might take measures to avoid higher personalized prices, such as using a virtual private network or a private browsing session, or avoiding retailers engaged in personalized pricing altogether. The Commission vote authorizing the Federal Register notice was 2-0, and the public will have 30 days to submit comments electronically once the statement is published in the Federal Register. Before you check out on anything that costs real money, reload the same listing in a private browsing window on a device you are not signed in on, and compare the two prices side by side. Our harvested material does not name any company, does not say personalized pricing is currently happening at any particular retailer, and contains no consumer complaints or dollar losses. A draft policy statement out for comment is not a finding that any business broke the law, and we do not know what the final statement will say or when it will take effect. LIVEFRAUD ALERTS is independent and not affiliated with the Federal Trade Commission or any company; this alert was written only from the single official announcement listed below.
- Single official source — an FTC press release quoted directly, with the agency's own wording on what may violate the FTC Act.
- No corroborating press, community or detector rows; nothing independent to test the announcement against.
- The subject is a proposed policy out for public comment, not a documented incident, so there is no victim account, loss figure or named business to anchor it.
- Reviewed by 3 models from independent houses.
THE RULE
Low immediate money risk, high everyday exposure: nobody is contacting you, but the price you are shown at checkout may be built from data about you.
What we don’t know
- Which retailers, if any, are setting prices from individual personal data.
- How many consumers have paid a higher price this way, and how much more.
- What data points feed a personalized price, and whether shoppers can see or challenge them.
- What the final policy statement will say after the comment period, and whether enforcement follows.
- Whether private browsing or a VPN actually changes the price shown — the announcement mentions these as measures consumers might take, not as tested outcomes.
The bench — who voted
3 INDEPENDENT AI MODELS REVIEWED THIS. ALL 3: MODERATE RISK.
The card names a count. Here are the seats behind it, with what each one said.
GOOGLEgemini-flash-latest · MODERATE
The finding includes specific factual assertions not found anywhere in the provided evidence row, including the specific date (19 August 2026), the name and quote of FTC Chairman Andrew Ferguson, and the announcement's explicit discussion of VPNs and private browsing sessions.
GROQopenai/gpt-oss-120b · MODERATE
The finding adds a direct quote from FTC Chairman Andrew Ferguson that is not present in the provided evidence row, stretching beyond the source text.
OPENROUTERopenrouter/free · MODERATE
The finding frames this as a 'warning that businesses...may be breaking the law,' which slightly sharpens the tone compared to a draft seeking comment. The evidence says the draft 'could violate' the FTC Act — 'may be breaking' and 'could violate' are close in meaning but not identical. However, the finding is careful throughout to note this is a draft, not a final statement, and contains no enforcement finding.
Reviewed by 3 independent models: 1 found it carried by the evidence, 2 did not.
▼ Protocol & challenge record
Objection (high): MATERIAL OMISSION FROM THE QUOTE. The source quote contains a limiting clause the finding drops entirely: 'The FTC does not have the legal authority to ban personalized pricing in all circumstances.' The finding reproduces the second half of Ferguson's sentence ('businesses that fail to tell consumers... may be in violation') while silently deleting the half that tells readers personalized pricing is largely lawful. That is selective quotation of a two-clause sentence, and it shifts the reader's takeaway from 'undisclosed use may be deceptive' to 'this practice is probably illegal.' The claim line inherits the distortion. This is the single most fixable and most serious problem in the draft.
Resolved: UNRESOLVED — requires an edit. Restore the limiting clause: add to the Ferguson sentence that he also said the FTC 'does not have the legal authority to ban personalized pricing in all circumstances.' Nothing short of restoring it fixes the quote.
Objection (high): ADVICE UNSUPPORTED BY, AND CONTRADICTED BY, THE DRAFT'S OWN UNKNOWNS. The advice sentence carries row_ids [] and instructs readers to reload listings in a private window 'on a device you are not signed in on' and compare prices. The source mentions VPN/private browsing only as measures 'informed consumers might take' — it makes no claim they work. The draft's own unknown #5 concedes exactly this. Publishing an imperative test procedure whose efficacy the draft admits is unknown is an internal contradiction. The 'device you are not signed in on' detail appears nowhere in the source; it is invented specificity.
Resolved: UNRESOLVED — requires an edit. Either cut the advice sentence, or rewrite it as attribution rather than instruction: 'The FTC's announcement suggests informed consumers might use a VPN or private browsing; it does not say these change the price shown.' Delete the invented 'device you are not signed in on' clause.
Objection (medium): ALTERNATIVE EXPLANATIONS UNADDRESSED FOR THE ADVISED TEST. If a reader follows the advice and sees two different prices, the draft gives them no way to interpret it. Price deltas between sessions are routinely produced by A/B testing, time-based or inventory-based dynamic pricing, geolocation/IP-based regional pricing, currency and tax inference, cart-abandonment coupons, cached promotions, and logged-in loyalty discounts — none of which is personalized pricing in the sense the FTC defines (pricing to estimated willingness to pay). The advice as written manufactures false positives and will generate reader accusations against retailers that the evidence cannot support.
Resolved: UNRESOLVED. If the advice survives at all, it must carry the caveat that a price difference between sessions has many ordinary causes and is not by itself evidence of personalized pricing.
Objection (medium): RISK LINE CONTRADICTS THE LIMITATION LINE. risk_line asserts 'high everyday exposure' and 'the price you are shown at checkout may be built from data about you.' The limitation two sentences later states the harvested material 'does not say personalized pricing is currently happening at any particular retailer.' A prevalence characterisation ('high everyday exposure') is a factual claim about the world and has no row backing it. The hedge 'may be' does not rescue 'high.'
Resolved: UNRESOLVED — requires an edit. Cut 'high everyday exposure' or replace with a sourced formulation, e.g. 'no immediate money risk; the FTC says it is putting businesses on notice, and names none.'
Objection (high): CALLOUT AND DIRECTIVE OPTION SETS ARE BROKEN. 'ATTENTION: VETERANS' and 'ATTENTION: SCAMMEDS' are unsourced audience targeting of exactly the kind targeting_dropped correctly rejected for 'shopping' — the rule was applied to the dropped items but not to the surviving option list. Worse, 'SCAMMEDS' is not a word. Three of four directive options are corrupt: 'Send this to any altogether you know' and 'Forward this to the altogethers in your life' are token-salvage from the adverb 'altogether' in the source sentence about 'avoiding retailers engaged in personalized pricing altogether.' If any of these is selected the alert ships gibberish under an official-looking banner. This is a generation defect, not a judgement call.
Resolved: UNRESOLVED — requires regeneration. Strike 'ATTENTION: VETERANS' and 'ATTENTION: SCAMMEDS' and all three 'altogether' directives. Survivors: 'ATTENTION: EVERYONE', 'ATTENTION: ANYONE WHO SEES A LISTED PRICE', 'Forward this to whoever in your house does the online buying', 'Tell a friend who assumes everyone sees the same price.'
Objection (medium): DATE CURRENCY / PERISHABILITY. The source is dated 2026-08-19 and the operative fact is a 30-day comment window that starts on Federal Register publication — a date the draft does not have and does not flag as missing. Depending on when this alert is read, the window may be open, unopened, or long closed, and the statement may already be final. The draft tells readers they have 30 days without telling them from when, and offers no way to check. A one-line pointer to federalregister.gov would fix it.
Resolved: PARTIALLY RESOLVED by the limitation 'we do not know what the final statement will say or when it will take effect.' Close the rest by stating the announcement date inline as the anchor and noting the 30 days runs from Federal Register publication, date not yet set as of the announcement.
Objection (medium): SINGLE-SOURCE, SINGLE-DOMAIN DEPENDENCY WITH NO INDEPENDENT ANCHOR. Every claim sentence maps to one row. The confidence_reasons acknowledge this but 'moderate' still feels generous for a body of text that cannot be falsified against anything. Note also the URL is bare 'ftc.gov' rather than 'www.ftc.gov'; the row is tagged official on trust, not on verification. For a policy action of this kind the Federal Register notice itself is a second, checkable primary document and its absence from the harvest is a real gap, not a stylistic one.
Resolved: PARTIALLY RESOLVED — confidence_reasons already name single-sourcing explicitly and the disclaimer says the alert was written from one announcement. Residual: I would still hold confidence at moderate and not above, and I would flag the missing Federal Register notice as a named gap.
Objection (low): THE 2-0 VOTE IS REPORTED WITHOUT CONTEXT. The FTC is statutorily a five-member body. A 2-0 authorization means only two commissioners participated. Reported bare, readers will read it as unanimous full-Commission backing. Either give the composition context or drop the vote count as uninformative to a consumer audience.
Resolved: Accept either remedy: add 'only two of the Commission's seats were voting' or drop the vote count. Desk's choice; low stakes.
Objection (low): WATCH ICONS DO NOT MATCH THE VECTOR. 'link' and 'person' signal phishing links and a human contacting the reader. The draft's own risk_line says 'nobody is contacting you.' Only 'card' is arguably apt, and even that overstates a transaction-side risk that is nowhere documented.
Resolved: UNRESOLVED. Drop 'link' and 'person'. Retain 'card' only if the Desk is comfortable defending a transaction icon on an evidence base with no transactions in it.
Objection (medium): CATEGORY STRETCH: A CONSULTATION IS NOT A FRAUD EVENT. A brand called LIVEFRAUD ALERTS publishing a notice-and-comment proceeding with no perpetrator, no victim, no loss and no named business dilutes the channel's signal. The limitation lines are honest, but honesty about having nothing does not by itself justify issuing an alert. The strongest defence of publication is the actionable advice — and per OBJ-02 that advice is the least supported element in the draft.
Resolved: NOT RESOLVABLE BY EDIT — a publish/hold decision for the Desk. My position is recorded in dissent.
Objection (low): OMITTED FRAMING, NOTED BOTH WAYS. The source frames this as an action 'under President Donald Trump's leadership' by 'the Trump-Vance FTC.' Stripping it is defensible neutrality, but it should be a recorded editorial decision rather than a silent one, because a reader who follows the link will see framing the alert did not prepare them for.
Resolved: RESOLVED as a recorded editorial choice. I do not object to the omission; I object only to it being silent. Logging it here discharges that.
Preserved dissent
ON THE RECORDI would not publish this as an alert in its current form. The one genuinely actionable line in it — reload the page in a private window and compare — is the one line with no source behind it, and the draft's own unknowns section says we do not know whether it works. Strip that line and what remains is a government consultation notice with no company, no victim, no loss and no deadline we can state. That is a newsletter item, not a fraud alert.
ON THE RECORDThe Ferguson quote as reproduced is not a fair rendering of the source. The FTC Chairman said in one breath that the agency cannot ban personalized pricing in all circumstances and that non-disclosure may violate the Act. We kept the second half and dropped the first. Whatever the reason for that, the effect is to make a hedged official statement read as a harder warning than it is, and it is the kind of edit that, if a reader clicked through, would cost us more credibility than the whole item is worth.
ON THE RECORD'High everyday exposure' in the risk line is a prevalence claim we invented. Our own limitation paragraph says the material does not establish that personalized pricing is happening anywhere. We cannot hold both positions in the same document and call the result moderate confidence.
ON THE RECORDThe callout and directive option lists contain generated nonsense — 'SCAMMEDS', 'the altogethers in your life' — and unsourced targeting of veterans that the same pipeline correctly refused for shoppers. Whatever produced those strings is not working, and I want it on the record that the failure was visible in this draft before anything shipped.
The sources
Official sourceFTC Seeks Comment on Enforcement Policy Statement Regarding Personalized Pricing2026-08-19
The FTC is seeking comment on an enforcement policy statement saying undisclosed use of personal data to set individual prices could violate the FTC Act, and that retailers implying a price is static when it varies by individual risk misleading customers.
Authority: official. Retrieved 2026-08-31.
Limitation: A draft statement open for a 30-day comment period, approved 2-0; it names no company, cites no enforcement action, and reports no consumer harm or losses.
Open the original source →Other checks
Published under standing founder pass (A9) — every claim source-mapped by the machine.
▼ What the machine checked
- ✓ Not a community submission.
- ✗ Draws on an FTC enforcement release, which names a defendant: "FTC Seeks Comment on Enforcement Policy Statement Regarding Personalized Pricing".
- ✓ All 5 material sentence(s) map to FTC.
- ✗ openrouter raised 1 objection(s); google returned "overstated"; groq returned "overstated" — published on the receipt, not blocking (A9 amendment).
- ✓ No audience band is set.
No human affirmed these. They were verified by the classifier described in Amendment A9, on 2026-09-01.
Something wrong here? Tell us and we'll correct it — corrections are published, not quietly edited.
Phishy? Send it → sharelivefraud.com/squire-it
Not affiliated with any government agency, credit bureau, bank, platform, or law-enforcement agency. Informational only — not legal or financial advice.
Naming a source is not an endorsement, and being named here is not an accusation against any company.
Powered by SquireIt™
Verify this receipt at squireit.com