What we found
The IRS publishes a page in its newsroom titled "Dirty Dozen tax scams". Our harvested copy of that page is a title and link only — it contains no description of how any of the listed scams works, no contact method, no dollar figures and no publication date. Because nothing in our harvested material names who was targeted or how, this alert does not describe any specific approach, message or payment demand. Treat any unexpected contact about your taxes as unconfirmed until you have closed it down and reached the tax authority yourself through the phone number or website you already use. LIVEFRAUD ALERTS is an independent consumer fraud awareness publication and is not affiliated with, endorsed by or speaking for the IRS; the agency name appears here only because it is the source of the page cited above.
- Only one harvested row, and its body text is empty — title and URL only.
- No publication date on the row, so we cannot say the listing is current.
- No harvested detail on contact method, target group, losses or impersonated brands, so no scam mechanics can be stated.
- No second source to corroborate or extend the single official row.
- Reviewed by 2 models, 1 from independent houses.
THE RULE
Unknown exposure: the single official row names a tax scam listing but gives us no mechanics, so readers cannot be told what approach to expect.
What we don’t know
- How contact is made in any of the listed scams (call, text, email, letter, in person).
- What the scammers ask for and how any money or data moves.
- Which brands or agencies are impersonated.
- Who is most affected — no age, occupation, income or location is described.
- When the cited page was published or last updated.
- Number of people affected and any loss amounts.
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.
ANTHROPICclaude-opus-5 · MODERATE
The closing advice frames the topic as 'unexpected contact about your taxes', which imports an assumption the evidence does not carry. The harvested row is a bare title; nothing in it establishes that the listed items involve inbound contact at all. The IRS 'Dirty Dozen' historically mixes impersonation contact scams with promoter-driven abusive tax schemes and bad preparer conduct, so a reader could be steered toward the wrong defensive posture. The finding should say plainly that it does not know whether contact is involved rather than defaulting to contact-based guidance.
DISSENT, PRESERVEDI record that the finding is honestly scoped: it claims only the page's existence and the emptiness of the harvested copy, and it does not manufacture scam mechanics, victim profiles or loss figures. On the evidence question it holds.
DISSENT, PRESERVEDMy moderate grade reflects the underlying category — tax-authority-linked fraud can carry large losses — not this write-up, which gives a reader nothing they could use to recognise an approach. If the panel prefers a lower grade on the ground that no recognisable pattern is described, that is a defensible reading and I would not object strongly.
DISSENT, PRESERVEDI would prefer this be published as a source pointer directing readers to the IRS page itself, not as a fraud alert.
GROQopenai/gpt-oss-120b · LOW
Raised no objection; read the evidence as holds.
Reviewed by 2 independent models; all found the finding carried by the evidence.
▼ Protocol & challenge record
Objection (high): No news hook and no date: the only row is an undated, body-empty title. The IRS 'Dirty Dozen' is an annually reissued evergreen list, so without a publication or last-updated date the harvested row cannot establish that anything happened recently, or which year's list this is. Dressing a dateless index entry as an 'alert' with 'ATTENTION: EVERYONE' and a share directive manufactures urgency from zero substantive content. The honest output here is no alert, not a low-confidence alert.
Not resolved — preserved on the record.
Objection (high): Internal contradiction between watch_icons and the finding. The finding states twice that no contact method is known and unknowns lists 'How contact is made (call, text, email, letter, in person)' as open — yet watch_icons asserts phone, email and link. Icons are read by readers as mechanics. This asserts exactly the fact the draft says it cannot assert.
Resolved: Clear watch_icons to an empty list; no channel may be signalled while contact method is listed as unknown.
Objection (medium): Alternative explanation not considered: an empty body with only a title and URL is equally consistent with a failed or blocked harvest (redirect, 403, paywall/JS render failure, dead path) as with a real live page. The draft treats the row as SOURCED proof that 'The IRS publishes a page...' in the present tense. Nothing in the row confirms the URL resolved at harvest time. The evidence limitation should say the row supports only that this title/URL pair was captured, not that the page currently exists as described.
Resolved: Rewrite the evidence limitation to: row captures a title and URL only, with no fetch confirmation or date, so it supports the existence of the citation and not the current state of the page.
Objection (medium): Source-to-claim stretch on 'in its newsroom' and on the quoted title. 'Newsroom' is inferred from a URL path segment, not from any harvested field. The harvested title is 'Dirty dozen tax scams'; the draft presents it inside quotation marks as 'Dirty Dozen tax scams', silently changing capitalisation of a purportedly verbatim title. Also note the harvested URL is a bare 'irs.gov/newsroom/dirty-dozen' — the IRS's actual canonical page for this list is not established by the row, so citing it as 'the page cited above' overstates what we hold.
Resolved: Quote the title exactly as harvested ('Dirty dozen tax scams') or drop quotation marks; attribute 'newsroom' to the URL path explicitly ('at a newsroom path on irs.gov') rather than to the page itself.
Objection (medium): Subject-matter mismatch in the advice. The IRS Dirty Dozen list has historically mixed consumer-facing impersonation scams with abusive tax-avoidance schemes and promoter arrangements that taxpayers enter voluntarily. With no body text harvested, the draft has no basis to pivot to 'treat any unexpected contact about your taxes as unconfirmed' — that presumes an inbound-contact scam pattern the evidence does not support. The advice is generic hygiene being passed off as responsive to this source.
Resolved: If published at all, replace the inbound-contact advice with a neutral pointer: consult the IRS's own list via a route the reader already uses, with no assertion about how any listed scheme reaches people.
Objection (low): 'ATTENTION: EVERYONE' is unsupported and unnecessarily broad given targeting_dropped is empty and no affected group is described; 'ATTENTION: TAXPAYERS' is at least tied to the page's subject. The share_directive's reference to 'household tax paperwork' also implies a household-level threat vector nowhere in evidence.
Resolved: Select 'ATTENTION: TAXPAYERS' from callout_options and remove the 'household tax paperwork' framing from the share directive.
Objection (low): The claim line says the row 'carries no scam details, victim descriptions or publication date' — accurate — but the finding then spends five sentences on disclaimers and boilerplate. Net informational content for a reader is zero beyond 'the IRS has a scam page'. Even if published, this fails a utility floor.
Not resolved — preserved on the record.
Preserved dissent
ON THE RECORDI would not publish this. The entire evidentiary basis is one undated link with no body text, and the draft's own confidence_reasons concede there are no mechanics, no date, no targets, no losses and no corroboration. What remains is a true but trivial statement that the IRS has a scam page, wrapped in generic advice and a disclaimer, carried under 'ATTENTION: EVERYONE' with an instruction to forward it. That combination borrows the IRS's authority to give an empty item the shape of a warning, which is the specific harm the disclaimer sentence is trying to pre-empt and does not cure. My recommendation is to hold pending a successful re-harvest that yields a date and at least one scam mechanic.
ON THE RECORDSeparately, and regardless of whether the Desk publishes: the phone/email/link watch icons must come off. Asserting three contact channels in iconography while the text twice says the contact method is unknown is not a nuance a reader will parse — it is a false specific, and it is the most concretely misleading element in the draft.
The sources
Official sourceDirty dozen tax scams
The IRS publishes a page in its newsroom titled "Dirty Dozen tax scams".
Authority: official. Retrieved 2026-08-17.
Limitation: Row holds only a title and URL with an empty publication field; it supports the existence and subject of the page and nothing about how any scam operates, who it hits, or when it was posted.
Open the original source →Other checks
Approved by ihubglobalhq on 2026-08-17, after review of the alert and its sources.
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