Invest1 distinct publisher3 min readPublished
Treasury put $12.7bn behind 33,904 suspicious activity reports about crypto investment fraud, and the more consequential half of that release is the dated indicator list compliance teams are now asked to screen against.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
Divide the total by the window and the release turns into an operating rate: 33,904 reports across 845 days is about 40 filings a day, and $12.7bn across the same span is roughly $15m a day of transactions that somebody inside a covered institution already thought worth writing up [2][3][1][2][3]. Per report, about $375,000 [4]. That is a mean, though, and FinCEN published no median, so the figure is equally consistent with a long tail of five-figure filings sitting beneath a small number of very large ones.
The awkward property of a $12.7bn assembled out of suspicious activity reports is that the institutions being asked to file more are also the measuring instrument [3][4]. FinCEN says as much, calling the number activity visible in the reporting system rather than a full accounting of stolen dollars [5]. If the alert lands, the next total goes up, and no outside reader will be able to separate more fraud from more diligence.
What a compliance team actually receives is the indicator list: large sums sent to newly created investment platforms after an online introduction, accounts used as pass-throughs for many unrelated people, mule-like activity and rapid conversion of funds into crypto [6]. Nothing in the alert as described attaches a penalty or an examination standard to those indicators, and the mechanisms it does name are voluntary, being information sharing under Section 314(b) of the USA PATRIOT Act with its legal safe harbor, and the Rapid Response Program for pushing intelligence to foreign financial intelligence units [13][14][5]. FinCEN's own framing is that timely, detailed filings help connect victims, wallets, accounts and facilitators and support freezing or recovering funds [15], which is an argument about usefulness rather than obligation. If those indicators never surface in a later enforcement matter, the September 3 release is a typology paper with a dollar figure attached [1].
Because reported victims spanned every age group and lived in all 50 states as well as several US territories [9], none of this can be fenced into a high-risk segment; it has to live in general retail monitoring, where the population is everybody.
The supply side is what makes timing unforgiving. FinCEN describes compounds in Southeast Asia operating at industrial scale [7], leaning on networks that handle recruitment, account creation, customer contact and the movement of proceeds [8], with "guarantee marketplaces" selling account creation, phishing assistance and professional laundering to whoever needs it [10]. Specialized launderers then open accounts, form shell companies and route value through mules [11] before stablecoin transfers reach exchanges outside the United States [12]. Account-creation defense, put another way, is a commodity the other side is already pricing against, and the last hop leaves the jurisdiction that could have frozen the money.
Ranked by verification strength, evidence, and original report placement.
The Financial Crimes Enforcement Network, the Treasury bureau that collects and analyzes reports filed under the Bank Secrecy Act, released the finding on September 3, 2026.
FinCEN reviewed 33,904 BSA reports submitted between September 8, 2023, and December 31, 2025.
Those filings described about $12.7 billion in transactions that investigators associated with suspected digital asset investment scams.
FinCEN paired the analysis with an alert telling banks, money transmitters, casinos and other covered institutions to watch for the same tactics and to file suspicious activity reports when they appear.
Officials presented the number as a measure of activity visible in the reporting system, not as a full accounting of every dollar stolen.
The alert's indicators include customers who suddenly send large sums to newly created investment platforms after meeting someone online, accounts used as pass-throughs for many unrelated people, and patterns consistent with mule activity or rapid conversion of funds into crypto.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · September 4, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
invest
Stablecoin KYC and the Line at the Point of Issuance1 distinct publisher
invest
FinCEN moves to close US correspondent accounts for Banque Misr's five UAE branches2 distinct publishers
invest
Regulators clear banks to contact the customers behind their SAR filings1 distinct publisher
leadership
FinCEN shrinks beneficial-ownership reporting from 32 million filers to about 20,0001 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One issuer, one retelling
The filing count, the date range and the $12.7bn all come from a single Treasury release, restated by Crowdfund Insider without a second read. The figures are specific and the provenance is unambiguous, which is worth something; what is missing is anyone outside FinCEN able to say whether 33,904 filings describe 33,904 distinct situations. FinCEN's own note that the total reflects what reached the reporting system bounds the number more honestly than anything in the coverage does.
Filing channel busy, indicator list untested
Institutions clearly use the reporting pipe: about 40 scam-linked filings a day for more than two years is a working habit, not a pilot. The screening ask published on September 3rd has no comparable record behind it. Nothing in this reporting shows a bank, exchange or money transmitter tuning detection to the new indicators, and with 314(b) sharing and the Rapid Response Program both optional, there is no deadline that would force the question.
Scale framing outruns the caveat
FinCEN calls $12.7bn a measure of activity visible in its reporting system; the write-up elevates it to industrial fraud and one of the most damaging threats facing American households, in the same piece that carries the caveat. Divided out, the population is about $15m a day nationwide and an average filing near $375,000 of suspected activity. The typology material — compounds, guarantee marketplaces, the stablecoin exit — is described concretely and is not oversold; the aggregate is.
An agency arguing for its own inputs
FinCEN's release makes the case for FinCEN's data stream: filings are not paperwork for its own sake, and better filings help freeze funds. That is a plausible claim and also a self-interested one, published by the body that receives the filings and asks for more of them. The two remedies named are ones institutions opt into rather than obligations the bureau must defend. On the publishing side, Crowdfund Insider serves crypto and fintech readers who want the red-flag list, which is why the operational half gets the room and the measurement caveat gets a sentence.
Specific, unduplicated
Dates, counts and mechanisms are precise enough that the factual spine holds, and the derived per-day figures follow from them arithmetically. Breadth is the weak side: a single publisher, an issuer that is also the sole data holder, and no way from this reporting to know how much of the $12.7bn overlaps between filings or was ever recovered.