Skip to main content
AVOID.NET
← AI-Powered Deepfake Crypto Scam Infrastructure (2026 Industrialization Wave)reviewed 2026-10-28 · 14 claims checked

Fact-check findings

What an automated fact-checker found when it re-read AI-Powered Deepfake Crypto Scam Infrastructure (2026 Industrialization Wave) against the sources the page cites. Only the most recent review is shown.

Read this first

These findings are produced by an automated reviewer, and its results vary between runs: the same page, checked three times on the same day, came back with 15%, 20% and 34% of its claims disputed, mostly because each run extracted a different number of claims. Treat what follows as leads, not rulings.

“Disputed” means the reviewer could not reconcile the claim with the evidence it cited. It does not mean the claim is false. “Unverifiable” means no reachable source settled it either way.

Nothing here changes the page on its own. A proposed correction is applied only after a human moderator approves it; until then the page reads as it did when reviewed.

unverifiable 3partially supported 5confirmed 60 corrections pending · 0 applied

unverifiable

3 claims

No source the reviewer could reach confirms or contradicts the claim.

  1. #12[unverifiable][awaiting moderator]in section: Celebrity Deepfake Endorsement Scams
    “The submitter's characterization of the campaign as using AI-generated correspondence is not confirmed by the primary government source and should be treated as unverified.”
    reviewerPage refers to a 'submitter' characterization of AI-generated correspondence, but there are no accepted submissionsaccepted_submissions is empty, so the reference to a submitter cannot be checked. The hedging is accurate (neither IRS nor Help Net mentions AI), but the IRS campaign has no confirmed AI link and weakly fits this page's topic. The same applies to the caveats section.
  2. #13[unverifiable][awaiting moderator]in section: Law Enforcement Response
    “As of September 2026, 47 of 50 U.S. states have enacted deepfake-specific legislation, with only Alaska, Missouri, and Ohio having not done so, per the MultiState legislative tracker.”
    reviewerEight of the numbers in the regulatory section: 47 of 50 states have deepfake legislation (only Alaska, Missouri, Ohio lack), iProov 0.1% human accuracyThe cited MultiState article is dated 2026-02-12 and so cannot support an 'as of September 2026' state count. Not fetched or verified. The iProov 0.1% figure is cited to Stingrai (a secondary blog) and also not verified.
    citedwww.multistate.us/insider/2026/2/12/how-…ross-the-country
  3. #14[unverifiable][awaiting moderator]in section: Law Enforcement Response
    “The EU AI Act's Article 50 labeling obligations, requiring machine-readable labels on AI-generated synthetic content, became binding on August 2, 2026.”
    reviewerEU AI Act Article 50 labeling obligations became binding August 2, 2026Not checked in this review. August 2, 2026 was the original application date, but any digital-omnibus delay for Article 50 obligations should be checked against EU sources. A US state tracker is a weak source for an EU date.
    citedwww.multistate.us/insider/2026/2/12/how-…ross-the-country

partially supported

5 claims

The cited evidence supports part of the claim but not all of it.

  1. #3[partially supported][awaiting moderator]in section: Overview and Threat Classification
    “classified scam activity as the only crime category to have reached 'Mature' status, its highest tier”
    reviewerTRM index classified scam activity as the only 'Mature' categoryScams mature and others lower is supported. 'Highest tier' not verified. The overall 54/100 is labelled 'emerging', which the page does not mention.
    citedwww.theblock.co/news/web3/2026-08-21-ai-…crime-trm-412297
  2. #6[partially supported][awaiting moderator]in the timeline
    “2024-10-17”
    reviewerHong Kong police arrested 27 people in a scam stealing ~$46M using real-time deepfake calls, in October 2024, dated 2024-10-17The 2024-10-17 date is the article publication date, not the arrest date. The correction to Oct 14 rests only on a search snippet whose year is unstated, so it is low confidence and optional.
    citedwww.pcworld.com/article/2492762/live-ai-…-convincing.html
  3. #8[partially supported][awaiting moderator]in section: Telegram-Based Illicit Tooling Marketplace
    “Transactions in these channels are predominantly settled in USDT.”
    reviewerMIT Technology Review identified 22 Telegram channels in Chinese, Vietnamese and English; transactions predominantly settled in USDTThe channel count, languages, date and Talsec numbers are confirmed. The USDT settlement claim is only loosely supported (Tether, in a laundering context).
    citedwww.technologyreview.com/2026/04/15/1135…g-bank-telegram/
  4. #9[partially supported][awaiting moderator]in section: Celebrity Deepfake Endorsement Scams
    “Letters contained Treasury and IRS branding, fabricated notice numbers, tax year ranges spanning 2017 through 2026, and QR codes linking to a fraudulent website that mimicked official government portals.”
    reviewerLetters contained fabricated notice numbers and tax year ranges spanning 2017 through 2026The 2017-2026 range was not found in either cited source (Coinbase and Journal of Accountancy not checked, so the range may come from one of them). The correction removes only the unsupported specifics.
    citedwww.irs.gov/compliance/criminal-investig…currency-holderswww.helpnetsecurity.com/2026/08/04/fake-…nce-portal-scam/
  5. #10[partially supported][awaiting moderator]in section: Celebrity Deepfake Endorsement Scams
    “The IRS Criminal Investigation Chief stated: 'The IRS did not send it. The IRS does not operate a Digital Asset Compliance Portal. This is a scam.'”
    reviewerIRS CI Chief quote: 'The IRS did not send it. The IRS does not operate a Digital Asset Compliance Portal. This is a scam.'The exact attribution of the full quote to the Chief was not verified; it appears to be an IRS statement rather than a verbatim Chief quote. Low severity.
    citedwww.irs.gov/compliance/criminal-investig…currency-holders

confirmed

6 claims

The cited evidence supports the claim as written.

  1. #1[confirmed][no action needed]in section: Scale of Losses: Chainalysis 2026 Crypto Crime Report
    “documented that cryptocurrency scams siphoned at least $14 billion on-chain in 2025, with projections reaching approximately $17 billion as investigators continued uncovering illicit wallets”
    reviewerChainalysis reported at least $14B in 2025 scam losses, projected ~$17B, with 1,400% growth in impersonation scamsConfirmed via secondary reporting of the Chainalysis report; the primary page was not fetched directly.
    citedwww.chainalysis.com/reports/crypto-crime-2026/
  2. #2[confirmed][no action needed]in section: Scale of Losses: Chainalysis 2026 Crypto Crime Report
    “Chainalysis separately found that scam operations with on-chain links to AI service vendors generated 4.5 times more revenue per operation than those without AI tooling, averaging approximately $3.2 million per operation versus $719,000 for non-AI schemes.”
    reviewerScams with on-chain links to AI vendors earned 4.5x more per operation ($3.2M vs $719K)Matches secondary reporting.
    citedwww.chainalysis.com/blog/crypto-scams-2026/
  3. #4[confirmed][no action needed]in section: Overview and Threat Classification
    “The index additionally reported that deepfake-scam losses in 2026 to date already exceeded all of 2025 by 263% as of the report's publication.”
    reviewerDeepfake-scam losses in 2026 to date exceeded all of 2025 by 263%Supported. The timeline entry calls it 'year-over-year', which is a slight mislabel, see next finding.
    citedwww.theblock.co/news/web3/2026-08-21-ai-…crime-trm-412297
  4. #5[confirmed][no action needed]in the timeline
    “reporting a 40% year-over-year increase in AI use across crypto crime, an adoption score of 54/100, and that 2026 deepfake scam losses already exceed all of 2025 by 263%.”
    reviewerThe TRM report shows a 40% year-over-year increase in AI use alongside a 54/100 scoreSupported by The Block. The TRM primary page was not fetched.
    citedwww.trmlabs.com/reports-and-whitepapers/…e-adoption-index
  5. #7[confirmed][no action needed]in section: Active Campaign: Fake IRS Digital Asset Compliance Portal (2026)
    “announced on April 23, 2026 the restraint of over $701 million in cryptocurrency, the seizure of 503 fraudulent investment websites, and one Telegram channel with more than 6,000 followers”
    reviewerScam Center Strike Force announced on April 23, 2026 restraint of over $701M, seizure of 503 websites and a Telegram channel with 6,000+ followers; two nationals charged with wire fraud conspiracy over Shunda compoundFully supported, including the page's note that no AI is alleged. The DOJ says the site was seized 'more than 503', a trivial difference.
    citedwww.justice.gov/opa/pr/scam-center-strik…enters-targeting
  6. #11[confirmed][no action needed]in section: Celebrity Deepfake Endorsement Scams
    “Coinbase investigators and threat intelligence firm DarkTower traced the fraudulent domain to a Hong Kong registrar, with the site hosted on Romanian infrastructure previously linked to banking and delivery-service phishing.”
    reviewerCoinbase and DarkTower traced the domain to a Hong Kong registrar with Romanian hostingCore attribution confirmed.
    citedwww.irs.gov/compliance/criminal-investig…currency-holders
How this fits together. The reviewer reads the published page and its cited sources and records one finding per claim. A human moderator decides whether each proposed correction is applied; those decisions, and the score changes they cause, appear in the audit log. Earlier review runs are not shown here; only the latest reflects the page as it stands.