FinCEN Finishes a Case in 156 Days, or It Does Not Finish

FinCEN did three things in two days. On 5 October it proposed banning American financial institutions from moving money to companies controlled by the A7 Network, a Russia-linked sanctions evasion service. On 6 October it abandoned two crypto rulemakings, one of which had been sitting open since December 2020.

Set against the record, the pattern is sharper than any of the three announcements. HaiPay paired every FinCEN special-measure proposal since 2015 with whatever happened to it. The ones that became prohibitions took between 124 and 164 days. The ones that were abandoned took at least 357.

The record

Eleven proposals, each naming a bank, a jurisdiction, a group of businesses or a class of transactions.

Four were carried through. North Korea took 159 days from proposal to final rule, Bank of Dandong 124, Al-Huda Bank 154, Huione Group 164. The median is 156 days, and the whole range fits inside six weeks.

Three were withdrawn. Banca Privada d'Andorra after 357 days, convertible virtual currency mixing after 1,079, and ABLV Bank after 2,415 days, which is six years and seven months.

Four are open today. The proposal on ten Mexican gambling establishments has been pending for 326 days, MBaer Merchant Bank for 221, Banque Misr UAE for 38, and the A7 Network sub-agents for four.

Two of those four are already older than every case FinCEN has ever finished.

What that means if you are the one named

A special-measure proposal is not an order. Nothing is prohibited while it is pending, and a bank reading the Federal Register learns only that a named counterparty has been found to be a primary money laundering concern.

In practice the finding does the work. The clearest case is ABLV. FinCEN published its finding on 16 February 2018; one week later the European Central Bank declared the bank failing or likely to fail, citing the run on deposits and the loss of dollar funding that followed the notice, and its licence was withdrawn that July. The proposed prohibition was never imposed. It was withdrawn six and a half years later, on the ground that ABLV was no longer of primary money laundering concern. Correspondent banks act on the finding, not on the final rule, which is why the gap between the two matters.

This is also the part of the pipeline nobody is required to report on. There is no statutory clock between a FinCEN finding and a final rule, and the agency publishes no queue.

The week itself

The new proposal is not made under section 311 of the USA PATRIOT Act, the authority behind most of the list. It uses section 9714(a) of the Combating Russian Money Laundering Act, the Russia-specific power added in 2021.

The target is defined by control rather than by name: any company operating outside the United States that is controlled by the A7 Network. OFAC sanctioned the network itself on 1 October under the transnational criminal organisation authority. The proposed measure would prohibit covered financial institutions from transmitting funds to or from the sub-agents.

The comment period is 30 days, closing 4 November. That is now the house length. The four proposals from 2015 to 2018 each ran 60 days, and the convertible virtual currency mixing proposal ran 91. Every proposal since January 2024, six of them, has run 30.

Going the other way, the two withdrawals were the only crypto-wide items in the pipeline. The 2023 finding on convertible virtual currency mixing is gone, and so is the December 2020 proposal that would have made banks and money services businesses report, record and verify identities for transactions involving unhosted wallets. On that one FinCEN says it will take no further action.

The shape of the shift

Read together, the week moves the tool from categories to names. A rule about mixing services or unhosted wallets applies to a technique. A prohibition on the sub-agents of one network applies to a counterparty list, and it reaches whoever is on it.

For a payments business the practical difference is where the compliance burden lands. A category rule is a build: new reporting, new record retention, new identity checks across a product. A named-target prohibition is a screening problem: the same sanctions and watchlist plumbing, pointed at one more entity, with the hard part being the sub-agents that are not individually named.

What this is not

HaiPay counted proposals that name a target and propose a measure. Orders issued without a prior proposal are excluded, which removes Bitzlato, PM2BTC and the three Mexican institutions sanctioned in June 2025, where FinCEN went straight to an order. Including them would shorten nothing, because they had no proposal stage to measure.

The pairing of each proposal with its outcome is HaiPay's, made by matching targets across FinCEN's own Federal Register documents. Two of the eleven are judgement calls: the FBME sequence of 2015 and 2016 ran through a reopened comment period and two final rules, and is not counted as a single proposal-to-rule pair, and the 2016 withdrawals covering Liberty Reserve and JSC CredexBank concern findings issued before 2015, so their proposals fall outside the window.

Days to outcome are not a measure of difficulty, and a withdrawal is not a mistake. The ABLV withdrawal records that the bank had lost its licence and entered supervised liquidation.

The claim that the A7A5 rouble-backed stablecoin passed $100 billion in transactions appears in the proposal as a citation to a blockchain analytics firm, not as a FinCEN measurement, and HaiPay has not verified it.

HaiPay did not contact FinCEN, OFAC or any named institution.

What to watch

Whether the Mexican gambling proposal, at 326 days, gets a final rule or a withdrawal. It is the oldest live case and the one that will tell you which group the current pipeline belongs to.

Whether the A7 prohibition lands inside the 124 to 164 day band. If it does, it would be finalised in late February or early March.

And whether anything replaces the two crypto rules. The withdrawals leave the technique-level gap open, and FinCEN said nothing about filling it.