What we found
The Federal Trade Commission has asked for public comment on a draft enforcement policy statement on personalized pricing, which it describes as the use of personal data to set prices according to the amount a company believes an individual consumer is willing to spend. The FTC chairman said businesses that fail to tell consumers how their personal data is being used to set a price may be in violation of the FTC Act, while stating the agency does not have the legal authority to ban personalized pricing in all circumstances. The draft statement says 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 the undisclosed collection or use of personal data for personalized pricing could violate the FTC Act's ban on unfair or deceptive practices. The same statement notes that 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. Separately, the FTC opened comment on a proposed policy statement saying AI companies that distort their systems' outputs to achieve undisclosed objectives could be deceiving consumers under Section 5 of the FTC Act. Both documents are drafts open for public comment, and our harvested material names no company engaged in personalized pricing, reports no enforcement action, and gives no complaint count or consumer loss figure. Neither row describes anyone being contacted, impersonated or asked for money, so we cannot say from this material that a scam is being run on shoppers. Before you buy anything online, load the same listing a second time in a private browsing window or on a device you are not signed in on, and treat any gap between the two prices as a reason to slow down and shop elsewhere. LIVEFRAUD ALERTS is independent and not affiliated with the Federal Trade Commission; this alert rests only on the two published FTC notices cited above.
- Both harvested rows are proposed or draft policy statements open for comment, not final rules or enforcement outcomes.
- No harvested row documents an actual scam, a victim, a loss amount or a named business engaged in the conduct described.
- Only two sources, both from the same agency, with no press, community or detector corroboration.
- Reviewed by 4 models, 3 from independent houses.
THE RULE
Low immediate fraud risk: the harvested material describes a proposed FTC policy on pricing and data disclosure, not an active scheme contacting consumers.
What we don’t know
- Which retailers, if any, currently set prices from individual personal data.
- How many consumers have paid a higher price because of personalized pricing, and by how much.
- Whether either draft statement will be adopted after the comment periods close.
- Whether private browsing or a VPN actually changes the price shown by any particular retailer.
- Whether any enforcement action has followed either proposal.
The bench — who voted
4 INDEPENDENT AI MODELS REVIEWED THIS. 2 OF 4 — TWO MODELS DISSENTED. READ WHY.
The card names a count. Here are the seats behind it, with what each one said.
ANTHROPICclaude-opus-5 · LOW
The finding attributes specific content to sources that the two evidence rows do not contain: a quoted position from the FTC chairman (including that the agency lacks authority to ban personalized pricing outright), the draft's language about retailers representing a price as static, and the draft's discussion of VPNs and private browsing. Row [1] establishes only that a draft statement warns undisclosed use of personal data to set individual prices could violate the FTC Act. On the material supplied, these added specifics are unverified.
DISSENT, PRESERVEDThe headline claim as literally worded is carried: it is hedged ('may be set'), and row [1] does support that the FTC has said undisclosed use of personal data to set prices can violate the FTC Act. My 'overstated' verdict rests on the finding's added detail, not on the claim sentence.
DISSENT, PRESERVEDI want it on record that the finding's own disclaimers are unusually candid — it states plainly that no company is named, no enforcement exists, no losses are quantified, and no one is being contacted or asked for money. That candour is why I grade the reader risk low rather than moderate, and it should not be edited out to make the item read more urgently.
DISSENT, PRESERVEDPublishing a notice-and-comment item under a fraud-alert masthead carries its own cost: it trains readers to treat regulatory proposals as active threats, which dulls response to alerts that describe real losses.
GOOGLEgemini-flash-latest · MODERATE
The finding includes specific details not present in the provided evidence rows, such as quotes/positions attributed to the FTC chairman, references to static price representations, and specific draft statement mentions of consumers using VPNs or private browsing.
GROQopenai/gpt-oss-120b · MODERATE
The finding includes statements about the FTC chairman's comments, specific risks of misleading customers, and consumer mitigation tactics (VPN, private browsing) that are not present in the supplied evidence rows.
OPENROUTERopenrouter/free · LOW
The finding correctly notes that the FTC documents are drafts/proposals, which matches the evidence.
Reviewed by 4 independent models: 1 found it carried by the evidence, 3 did not.
▼ Protocol & challenge record
Objection (high): Date currency failure inside a limitation sentence. The draft asserts "Both documents are drafts open for public comment." That is verifiably false for row 303f1a36: the source states comments closed July 31, 2026, and the release is dated 2026-07-01. The pricing statement (pub 2026-08-19) allows 30 days from Federal Register publication, so its window is also likely closed or nearly so. Presenting closed comment periods as open both misinforms readers who might want to comment and inflates the item's currency. The status of each docket as of publication date must be checked and stated separately, not merged.
Not resolved — preserved on the record.
Objection (high): The advice sentence is not supported by either source and invites false positives. The FTC release only observes that "informed consumers might take measures" like VPN or private browsing — a description of consumer behavior, not an FTC-validated detection method. The draft's own unknowns concede "Whether private browsing or a VPN actually changes the price shown by any particular retailer" is unknown. Yet the advice tells readers to "treat any gap between the two prices as a reason to slow down and shop elsewhere." Price gaps between a logged-in session and a private window have many innocuous causes: loyalty/member discounts, applied coupons, A/B testing, geolocation and currency, shipping/tax estimation, cart-abandonment promos, time-of-day dynamic pricing, and cached pages. The source itself says consumers should expect prices to change with supply and demand. As written the advice manufactures a diagnostic that the evidence does not support.
Resolved: Partially acknowledged in unknowns ("Whether private browsing or a VPN actually changes the price shown by any particular retailer"), but unknowns do not travel with the advice sentence; a reader acts on the advice, not the metadata. Unresolved.
Objection (high): Source-to-claim stretch on the AI row. The source says AI companies that distort outputs "to achieve undisclosed ideological objectives" could be deceptive; the finding drops "ideological" and writes "undisclosed objectives," materially broadening the legal theory. The finding also omits the entire operative context of that release — a December Trump executive order directing the FTC to issue it, and an express preemption attack on Colorado's AI Act. Stripping that context converts a politically directed preemption document into a generic consumer-protection item, which misrepresents what the source is.
Not resolved — preserved on the record.
Objection (medium): Relevance/corroboration inflation. The AI-accuracy row has no connection to personalized pricing beyond sharing an agency, a chairman and Section 5. Including it as a second SOURCED evidence row creates the appearance of two-source support for a one-source claim. The confidence_reasons even note "only two sources, both from the same agency" without noting that only one of them is on topic. Effective sourcing here is n=1.
Not resolved — preserved on the record.
Objection (medium): The headline claim asserts a fact about the world that neither source establishes. "Some online prices may be set from your own personal data" is presented as descriptive of current retail practice, but the FTC document names no retailer, cites no study, no complaint count and no enforcement action. The FTC is warning about a practice and putting firms "considering" it on notice — which is compatible with the practice being rare or not yet documented at scale. The hedge "may" carries a lot of weight that the audience callout ("ATTENTION: ONLINE SHOPPERS") and the buy-time advice immediately undo.
Resolved: Partially pre-empted: the finding's limitation sentence already states the material "names no company engaged in personalized pricing, reports no enforcement action, and gives no complaint count or consumer loss figure," and confidence is set to low. The residual problem is the gap between that limitation and the assertive headline/callout/advice, which the limitation does not reach.
Objection (medium): "the FTC says failing to disclose that can break the law" overstates the source's modality and provenance. The document is an unadopted draft; the operative language is "may be in violation" (Chairman Ferguson, a quote in a press release) and "could violate." "Can break the law" reads as a settled agency legal position. Additionally the authorizing vote was 2-0, i.e. a two-member Commission — a thin quorum whose draft statement carries less institutional weight than the phrase "the FTC says" implies, and which a successor Commission can abandon.
Resolved: Partially mitigated by the finding's accurate rendering of "may be in violation" and by the explicit note that both documents are drafts. The unresolved part is the claim line itself and the absence of any note on the 2-0 vote / Commission composition.
Objection (medium): Watch icons contradict the finding. "card" and "person" signal payment fraud and impersonation, while the finding explicitly states "Neither row describes anyone being contacted, impersonated or asked for money" and the risk_line says low immediate fraud risk. The iconography re-inserts the scam frame the text just disclaimed.
Not resolved — preserved on the record.
Objection (medium): Generation-pipeline corruption in the option lists. "Send this to any altogether you know" and "Forward this to the altogethers in your life" are nonsense derived from the word "altogether" in the source paragraph. "ATTENTION: SCAMMEDS" is not a word. "ATTENTION: VETERANS" has zero basis in either row — inventing a targeted vulnerable audience with no evidentiary hook is the exact pattern that produces harmful mis-targeting. These being "options" rather than the selected values is not a defense; a pipeline emitting them can select them.
Not resolved — preserved on the record.
Objection (low): Verification hygiene: the finding says "The FTC chairman said" without naming Andrew N. Ferguson, weakening reader verifiability. Separately, the AI row's harvested HTML contains an internal inconsistency — the comment link points to docket FTC-2026-0859 while the title attribute reads FTC-2026-0727. Minor, but it means the docket number should not be relied on from this harvest, and it is a small flag on harvest fidelity.
Not resolved — preserved on the record.
Objection (low): Genre mismatch. Everything substantive here is a proposed policy statement with no victim, no loss, no actor and no contact vector. Running it under a fraud-alert masthead with an "ATTENTION" callout and a share directive applies scam-alert urgency to a rulemaking notice. The disclaimer and risk_line partially mitigate, but format is a claim too.
Resolved: Mitigated by risk_line and the independence disclaimer, which together tell the reader this is a policy proposal and not an active scheme.
Preserved dissent
ON THE RECORDI would drop the AI-accuracy row entirely. It shares an agency and a statute with the pricing item and nothing else, and its inclusion is doing corroboration work it cannot do. If it stays, it must carry the executive-order and Colorado-preemption context and the word 'ideological' must be restored; as currently paraphrased it misdescribes the document.
ON THE RECORDI would cut the advice sentence. Telling readers to compare a private-window price against a logged-in price and treat any difference as a warning sign is a test with a high false-positive rate and no support in the record — the draft's own unknowns say we do not know whether it works. If some action step is required by format, the only defensible one is: comparison-shop across retailers and be aware that a listed price may not be universal.
ON THE RECORDI do not think this should run as a fraud alert. There is no actor, no victim, no contact, no loss and no money changing hands — the risk_line concedes as much. A proposed policy statement dressed in ATTENTION-shopper urgency trains readers to treat regulatory news as an active threat, which degrades the value of the masthead when a real alert issues.
ON THE RECORDThe 'altogether' directives, 'SCAMMEDS', and the unsourced 'VETERANS' callout indicate the option-generation stage is producing artifacts from raw source text and inventing audiences. That is a pipeline defect, not a copy nit, and I would not sign off on any output from this run until it is fixed.
The sources
Official sourceFTC Seeks Comment on Enforcement Policy Statement Regarding Personalized Pricing2026-08-19
The FTC is seeking comment on a draft enforcement policy statement warning that undisclosed use of personal data to set individual prices could violate the FTC Act.
Authority: official. Retrieved 2026-08-20.
Limitation: A draft open for a 30-day comment period; it names no company, cites no enforcement case and quantifies no consumer harm.
Open the original source →Official sourceFTC Seeks Public Comment on Policy Statement Addressing AI Accuracy2026-07-01
The FTC is seeking comment on a proposed statement that AI companies distorting their systems' outputs contrary to consumer expectations could be deceiving consumers under Section 5.
Authority: official. Retrieved 2026-08-20.
Limitation: Also a proposal rather than a finding; it describes possible conduct and legal theory, not any identified incident or affected consumer.
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.
- ✗ groq returned "overstated"; openrouter raised 1 objection(s); anthropic returned "overstated"; google 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 2026-08-20.
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