Measuring the Mirror: What FinCEN's Human Smuggling Filings Actually Record

On 13 August 2026 the Financial Crimes Enforcement Network (FinCEN) published a Financial Trend Analysis on human smuggling. It examines 67,540 Bank Secrecy Act (BSA) reports filed between 2023 and 2025, carrying more than $4.9 billion in suspicious activity, and its headline finding is a collapse. Filings peaked at 29,266 in 2024 before falling to 11,018 in 2025, a decline of 62 per cent, which the report sets beside US Customs and Border Protection figures showing crossings at the southwest border in May 2025 down 93 per cent year on year. Two numbers falling together invite an obvious inference. Yet a filing count is not a crime rate. Read structurally, this dataset reflects the instrument that produced it as faithfully as the activity it reports.

The report exists because Congress required it. Section 6206 of the Anti-Money Laundering Act of 2020 obliges FinCEN to publish threat pattern and trend information derived from BSA filings, and human smuggling has been a government-wide anti-money laundering priority since June 2021. The dataset was assembled on much narrower terms. FinCEN selected filings that carried the key term issued with its Alert on Human Smuggling along the Southwest Border of 13 January 2023, or that checked human smuggling as a suspicious activity type. The population is therefore co-extensive with a red flag set written for one border at one moment. That is a defensible way to build a dataset and a poor way to measure a global business.

One filer supplies the picture

Concentration in the data is extreme: money services businesses (MSBs) filed 97 per cent of the reports, some 65,238 of them, and accounted for nine of the ten largest filers. Three of those firms produced 86 per cent of all MSB filings. One produced 27,879 reports by itself, 43 per cent of the MSB total and more than two fifths of the entire population. The sums were modest, averaging $7,961 a filing across MSB submissions with a median of $3,560. This is not a national threat picture. It is largely one firm's detection model, published as one.

That matters because the trend line is sensitive to decisions no supervisor announces. A threshold recalibrated after a false positive review, a corridor exited on commercial grounds, a rule retired when a vendor model is replaced: any of these moves the national series while nothing changes on the ground. FinCEN cautions in the report itself that suspicious activity reporting reflects only what institutions have identified and reported, and should not be treated as a complete representation of any activity. The caveat is boilerplate. Its force here is not.

Unverified is not suspicious

The typologies cited by MSB filers deserve closer reading than they have received. The most common, in 38,683 filings or 59 per cent of those lodged by MSBs, was no verifiable familial connection between originator and beneficiary. Money flows that did not follow typical patterns came second, at 26,614; funds sent to locations along known migration routes appeared in 9,018 reports.

The first of those indicators is not a finding. A remittance firm has no obligation to establish that a sender and a receiver are related, and no mechanism capable of proving it. The most frequently cited basis for reporting is therefore the absence of information the filer was never positioned to hold. Scale sharpens the problem: the Bureau of Economic Analysis puts personal remittances from immigrants in the United States to residents abroad at more than $72 billion in 2024, a figure FinCEN itself cited in an alert last November. An indicator that fires on unverified relationships within that population is noisy by construction.

None of this is an argument for filing less. It is an argument about what the aggregate can support. Nearly seven per cent of MSB filings recorded a customer admitting the money was intended to help smuggle a relative or friend, which is direct evidence of intent. In more than nine cases out of ten, the report rests on pattern and inference. Intelligence of that kind is worth having. It is not worth counting.

Banks see money, not movement

The value sits where the volume does not, and the proportions repay precision. Depository institutions filed 2,075 reports, 3 per cent of the dataset, and carried 61 per cent of the value at roughly $3 billion, with a median filing of $82,965 against the MSB median of $3,560. American Banker, analysing FinCEN's published filing statistics, put that contribution in proportion: banks and credit unions submitted 6.19 million suspicious activity reports of all types across the same three years, making these 2,075 about one in every 3,000. Their subjects were domestic almost without exception, with 2,029 naming someone in the United States against 41 for Mexico.

The typologies they described are ordinary retail banking, which is why they travel. One filer reported more than $3 million in charter flight payments suspected of financing movement from the United Arab Emirates to Nicaragua. Another traced more than $195,000 in cash deposits into card purchases from a supplier of thermal binoculars and night vision attachments. Neither pattern needs a border to work.

The division of sight is the real finding. MSBs see the corridor and cannot see aggregation; banks see aggregation and cannot see the corridor. Banks are not blind to smuggling finance. They are recording it as something else, one filing in three thousand at a time.

The label has already moved

The decline did not happen in a quiet period. In March 2025 FinCEN issued a geographic targeting order requiring MSBs in 30 ZIP codes across California and Texas to report cash transactions between $200 and $10,000. In September 2025 it raised that floor to $1,000 and extended the order into Arizona. In March 2026 it renewed the order again and pushed it inland, adding Maricopa and Pima counties in Arizona and three in New Mexico, none of them on the border. Each version has been reissued within days of the last expiring; the current terms run to 2 September 2026. Surveillance of the corridor intensified and widened throughout the period in which smuggling-tagged filings fell by nearly two thirds; it moved into a different report.

Then the labels multiplied. FinCEN issued an alert on bulk cash smuggling by Mexico-based criminal organisations in March 2025, an alert on oil smuggling along the southwest border that May, an advisory on Chinese money laundering networks in August, and on 28 November 2025 an alert on cross-border funds transfers involving people without lawful status in the country. Each carries its own reporting key term, and FinCEN's published index of those terms links the bulk cash alert back to the 2023 human smuggling alert. The trend analysis accompanying the Chinese money laundering advisory found 1,675 filings in its own dataset flagging possible human trafficking or human smuggling, which makes the point: the signal lives in datasets that are not the smuggling dataset.

The pace has not slowed. In the three months before this report appeared, FinCEN issued a notice on trafficking around the 2026 World Cup, a joint advisory on populations without work authorisation and their employers, and a supplemental alert on fuel smuggling along the southern border. Each carries a distinct key term. The June advisory reaches much of the same remittance behaviour the smuggling indicators were written to catch, filed under a heading of financial system integrity.

Designation compounds the effect. Executive Order 14157 of 20 January 2025 directed the designation of cartels as foreign terrorist organisations, and the State Department designated eight groups the following month. FinCEN's director, Andrea Gacki, framed the analysis around smuggling as a revenue stream for larger criminal organisations, several of which now hold that designation.

Where a filer reaches the same conclusion, the natural suspicious activity type becomes terrorist financing rather than human smuggling, and the filing leaves this population altogether.

The collection has legal holes as well. FinCEN records that, in accordance with court orders, MSBs under the jurisdiction of the US District Court for the Southern District of California, and certain Texas-based firms, have not been required to report under the border orders. Part of the densest corridor in the analysis sits outside the reporting.

None of this explains the 2025 fall, which was already visible by the second quarter and which a November alert cannot have caused. It explains something more awkward. The series running forward will not be comparable with the three years behind it, and the break will be invisible to anyone reading the line alone.

Smuggling has moved north

The report's own evidence shows the geography moving. Filers identified more than $32 million in suspicious transactions across MSBs located in Canada, with activity flagged in Minnesota and North Dakota. In May 2026 an Indian national based in New York pleaded guilty to directing smuggling operations across the Canadian border, on a route that began in India and passed through the United Kingdom before reaching North America.

The dataset's geography mirrors the alert that built it: Houston is the highest-value money order sending city, Ciudad Juarez the highest-value receiving city. A template built for that corridor will keep finding it.

The filer becomes the subject

The Bank Secrecy Act reaches few British institutions directly. Designation reaches them all. The Global Irregular Migration and Trafficking in Persons Sanctions Regulations 2025, made under the Sanctions and Anti-Money Laundering Act 2018, created a dedicated designation power covering people smuggling, trafficking in persons and those who provide material support to either. The statutory guidance carries a detail that repays attention: a bespoke licensing ground allows the Office of Financial Sanctions Implementation to permit repayments from designated registered money services businesses to the customers who originally sent the funds. That provision exists because the regime anticipates freezing a payment firm rather than a payer.

Read against the FinCEN data, the inversion is stark. The American dataset treats money services businesses as the eyes of the system, since 97 per cent of what is publicly known about smuggling finance comes from them. The British regime treats the same category of firm as a candidate for designation. A firm can be the principal source of intelligence and the object of an asset freeze at once, and neither framework reconciles those positions.

Three tests follow. First, whether a firm can say how much of its own smuggling intelligence rests on a single rule or a single system, because concentration inside a firm produces the fragility the FinCEN dataset displays nationally. Second, whether a falling internal filing count has ever been examined as anything other than efficiency, since a control output is not a control result. Third, whether narratives distinguish an unverified relationship from a disproved one, because a supervisor testing the quality of a firm's intelligence will start precisely there.

A reflection is not a measurement

Crossings fell in 2025. That is measured elsewhere, by an agency counting people rather than paperwork, and the report's own comparison sets an annual filing total beside a single month. The two series do not move together in any case: crossings fell by 93 per cent on that measure while filings fell by 62 per cent, and the quarterly filing count turned upward again in the final quarter of the year. What the filing curve records is the reach of one red flag set, applied overwhelmingly by three firms, along one border, before the labels multiplied and the route moved north.

Every reporting regime eventually begins to measure itself. The discipline lies in noticing the moment it does, because a falling number is the easiest thing in financial crime compliance to mistake for progress.

Do you know why your firm's reporting volumes fell last year, or only that they did?

If the answer rests on inference rather than evidence, we can help you establish which it is before a supervisor asks the question; contact us.

At OpusDatum, we test whether a firm's financial crime indicators still describe the activity they were written to catch, and whether the reasoning behind a filing decision would withstand examination. We are interested in what a control demonstrates rather than what it was designed to do; that includes what a falling reporting volume proves, and what it merely implies.