What we found
The FBI's Internet Crime Complaint Center published a public service announcement on 15 October 2026 warning that scammers could exploit the 2026 US midterm election to carry out financial fraud schemes targeting victims across the United States. According to the advisory, scammers use the names, images, logos and slogans of candidates to solicit campaign contributions, sell merchandise that is never delivered, or steal personally identifiable information for further fraud. The advisory states that scammers also use artificial intelligence tools to generate highly realistic voices, images and written messages impersonating candidates, campaigns, government entities and trusted contacts, which it says makes the schemes harder to detect. Four named patterns are described: a campaign investment pool that promises a return on contribution dollars after the candidate wins and rewards recruiting others; a fake political action committee the scammers have no affiliation with; an online purchase scheme where political merchandise is never shipped; and a voter registration scheme in which a text message or email claims you are not registered and pushes you to a fraudulent state registration page. The advisory states plainly that campaign donations are not an investment and will not increase in value nor be returned. If a message says you are not registered to vote, do not touch the link — open a new browser window yourself and check your registration at vote.gov instead. The advisory does not report how many people have been targeted, how much money has been lost, or which states, candidates or committees have been impersonated. Our harvested material contains a single official advisory and no complaint data, victim accounts or detector results, so we cannot say how widespread these schemes currently are. LIVEFRAUD ALERTS is an independent consumer awareness desk and is not affiliated with the FBI, the Federal Election Commission, or any campaign, party or committee.
- The sole source is a dated public service announcement from the FBI's Internet Crime Complaint Center, published 2026-10-15, and every claim sentence tracks its wording.
- The advisory describes the scheme mechanics in detail and is an update to a prior 2024 PSA on the same playbook.
- Confidence covers the existence and shape of the warning only — no victim counts, loss figures or named impersonation targets are given, so scale is unmeasured.
- Reviewed by 2 models from independent houses.
THE RULE
Money sent as a 'campaign investment' or to a fake PAC is gone, and a fake registration page hands over the identity details used for later fraud.
What we don’t know
- How many people have been contacted or have lost money in the 2026 cycle.
- The dollar amounts involved and the payment methods the scammers request.
- Which states, candidates, campaigns or committees have been impersonated.
- Whether the AI-generated voices and messages described have been observed in these specific schemes or are cited as a general capability.
The bench — who voted
2 INDEPENDENT AI MODELS REVIEWED THIS. 1 OF 2 — ONE MODEL DISSENTED. READ WHY.
The card names a count. Here are the seats behind it, with what each one said.
GROQopenai/gpt-oss-120b · HIGH
The evidence row only confirms that the FBI issued a PSA warning about potential election‑related fraud. It does not enumerate the four specific scam patterns, the use of AI‑generated voices/images, or the detailed advice about vote.gov. Those details are added beyond what the source text states, so the finding overreaches the evidence.
OPENROUTERopenrouter/free · MODERATE
Raised no objection; read the evidence as holds.
DISSENT, PRESERVEDThe PSA is preventive and offers no evidence of actual incidents, so the risk to an ordinary reader may be higher than moderate.
Reviewed by 2 independent models: 1 found it carried by the evidence, 1 did not.
▼ Protocol & challenge record
Objection (high): Tense overclaim in the claim line. The claim asserts scammers 'are using' the 2026 midterm to run these schemes. The source body says only that scammers 'could exploit the 2026 US Midterm Election' — conditional and forward-looking. The only present-tense wording is the PSA's own headline, which the FBI contradicts in its first sentence. The finding sentence gets it right ('could exploit'), so the headline claim is out of step with the very finding it summarizes. Readers who see only the claim line will believe a live, observed 2026 campaign is underway. Fix: 'The FBI warns that scammers could exploit the 2026 U.S. midterm election...' or attribute the present tense explicitly to the PSA title.
Resolved: Desk can resolve by rewriting the claim line to 'The FBI warns that scammers could exploit the 2026 U.S. midterm election to run fake campaign-donation, fake-merchandise and fake voter-registration schemes.' This matches the first sentence of the source exactly and costs nothing in reader value.
Objection (high): Dropped 'As previously reported.' The entire paragraph describing candidate-name misuse, non-delivered merchandise, PII theft and AI-generated voices/images is prefaced in the source by 'As previously reported' — i.e. it is carried forward from the 29 October 2024 PSA, not a fresh 2026 observation. Finding sentences 2 and 3 present this material as current 2026 characterisation with no such marker. This is a date-currency problem, not a nitpick: the AI capability claim in particular is a 2024-era statement being read as 2026 evidence. The unknowns list half-acknowledges this for AI only; it should be on the face of the sentences. Fix: 'Repeating language from its 2024 advisory, the FBI says scammers use...'
Resolved: Insert the source's own qualifier into finding sentences 2 and 3 ('As previously reported' / 'Repeating its 2024 advisory'), and promote the existing AI unknown so it sits adjacent to the AI sentence rather than at the foot of the item.
Objection (medium): 'Fake political action committee' mischaracterises the ScamPAC pattern. The source describes scammers duping victims into believing they are donating to a legitimate PAC or candidate committee 'but the scammers have no affiliation with the PAC and keep the funds' — i.e. impersonation of a real or real-seeming entity by unaffiliated actors. The finding's 'a fake political action committee the scammers have no affiliation with' is internally incoherent (if the PAC is fake, affiliation is meaningless) and points the reader at the wrong defensive check. The risk_line repeats 'fake PAC'. The source also warns specifically that 'some scam PACs and SuperPACs may appear registered' — a nuance this framing erases.
Resolved: Rewrite as: 'a political action committee scheme in which scammers solicit donations while posing as a legitimate PAC, Super PAC or candidate committee they have no affiliation with, and keep the funds.' Amend risk_line to drop 'fake PAC' in favour of 'a PAC the scammers do not represent.'
Objection (medium): Unsupported targeting in the option set. 'ATTENTION: VETERANS' appears in callout_options and 'Send this to any candidate you know' / 'Forward this to the candidates in your life' appear in directive_options. Nothing in the harvested row mentions veterans, and the PSA addresses donors, purchasers and voters as victims — not candidates as the audience. targeting_dropped is empty, which leaves no record that fabricated targeting was considered and rejected. The selected callout and directive are fine; the audit trail is not.
Resolved: Record 'ATTENTION: VETERANS' and both candidate-directed share directives in targeting_dropped with the reason 'no basis in harvested material', so the rejection is visible on the record.
Objection (medium): The single verification mechanism the FBI actually gives for the donation schemes is omitted. The PSA tells readers to check PAC/party/candidate committee registration and financial filings on the FEC website, and flags high fundraising/consulting/compliance expense ratios as a scam signal. The finding carries the voter-registration advice (vote.gov) but gives donors nothing actionable, even though two of the four named patterns are donation schemes and the risk_line leads with donated money. This is an advice-coverage gap that makes the item less useful than its source.
Resolved: Add one advice-role sentence covering the FEC filings check and the high-overhead red flag, drawn directly from the source's Tips section, with the row_id attached.
Objection (medium): Confidence 'high' is attached to a claim line that the source does not support at that strength. The confidence_reasons carefully scope confidence to 'the existence and shape of the warning only', but the claim as written asserts ongoing conduct. Either the claim is hedged (OBJ-1) or the confidence should drop. Also note the evidence base is a single official row with no corroboration, no complaint data and no second source — defensible for a 'the FBI said X' item, indefensible for a 'this is happening' item.
Resolved: If OBJ-1 is accepted, 'high' is sustainable as scoped. If the claim line keeps present tense, confidence must drop to medium.
Objection (low): 'Four named patterns are described' risks implying exhaustiveness. The source says election-related fraud 'come[s] in many varieties, including the schemes detailed below.' Suggest 'Four patterns are described, which the advisory says are examples rather than a complete list.'
Resolved: Add 'including' / 'examples rather than a complete list' to the four-patterns sentence.
Objection (low): Channel coverage mismatch. watch_icons are text-message, email, link. The PSA's first protective tip names 'unsolicited calls, texts, emails, or surveys' — voice calls are explicitly in scope and are the channel most exposed to the AI-voice capability the finding highlights. A phone icon is missing.
Resolved: Add a phone/voice-call icon to watch_icons.
Objection (low): Shelf-life. The item hangs on a 15 October 2026 PSA about a November 2026 election. The voter-registration hook and the donation-pool hook both decay sharply after election day. Nothing in the record states the check date relative to the election, so a reader encountering this later cannot tell whether the warning window is still open.
Resolved: State the advisory date prominently (already done in sentence 1) and add a limitation noting the warning is tied to the November 2026 election cycle.
Preserved dissent
ON THE RECORDI do not accept the claim line as written. The FBI says scammers 'could exploit' the 2026 midterm; the desk's claim says they 'are using' it. That is a prediction converted into a report of fact. The PSA's own headline uses the present tense, but the FBI contradicts its headline in its very first sentence, and a consumer desk should follow the body, not the SEO line. Publishing 'are using' at confidence 'high' tells readers an observed wave is underway when the only harvested evidence is a preventive notice with zero complaint volumes, zero loss figures and zero named incidents.
ON THE RECORDThe removal of 'As previously reported' is the more serious of the two problems and the one I expect to be waved through as a formality. It is not a formality. That phrase is the FBI marking its own material as recycled from October 2024. Strip it and a two-year-old threat description is republished as a 2026 finding. The AI paragraph is the sharpest example: it is the most alarming content in the item, it is the content most likely to be quoted onward, and it is the content furthest from any 2026 observation.
ON THE RECORD'Fake PAC' is wrong and I want the objection preserved even if the desk judges it cosmetic. The FBI's point is that the entity can look real, can appear registered, and that the fraud lies in the solicitor's lack of affiliation. Telling readers to watch for a 'fake PAC' trains them to look for the wrong thing and sets them up to be reassured by a legitimate-looking registration — precisely the failure mode the PSA calls out.
ON THE RECORDDropping the FEC filings check while keeping the donation warnings is a defensible editing choice but a poor one. The item tells donors twice that their money is at risk and never tells them the one verification step the source supplies.
ON THE RECORDI note for the record that 'ATTENTION: VETERANS' was generated as a candidate audience with no support anywhere in the harvested row. The selected callout is sound, so no harm reached the page, but an empty targeting_dropped field conceals that an unsupported demographic hook was on the table.
The sources
Official sourceScammers Exploit 2026 U.S. Midterm Election to Perpetrate Multiple Fraud Schemes2026-10-15
The FBI's Internet Crime Complaint Center published a public service announcement on 15 October 2026 warning that scammers could exploit the 2026 US midterm election to carry out financial fraud schemes targeting victims across the United States.
Authority: official. Retrieved 2026-10-16.
Limitation: The PSA is forward-looking and preventive; it gives no complaint volumes, loss totals, dates of specific incidents, or identified offenders.
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.
- ✓ No entity is named.
- ✓ All 5 material sentence(s) map to FBI/IC3.
- ✗ groq returned "overstated"; openrouter recorded dissent — 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 2026-10-16.
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