On October 1, the U.S. Treasury's Office of Foreign Assets Control designated the A7 payment network โ a cross-border settlement system that FinCEN estimates moved roughly $17 billion over an 18-month window for sanctioned Russian and Iranian interests. The headline number is doing a lot of work in the coverage. The mechanics deserve more attention.
OFAC did not sanction the token at the center of A7. It could not. A7A5, the network's internal settlement asset, sits on an isolated ledger: no DeFi integrations, no exchange listings, no composability with anything outside its own perimeter. A designation against the token itself would have been a press release, not an enforcement action.
What OFAC hit instead was the door. You don't freeze an asset you cannot reach โ you freeze the counterparty that has to touch it. Here, that counterparty is any desk willing to convert A7A5 into Tether's USDT. And USDT is the point at which a $17 billion structure becomes visible, seizable, and structurally fragile.
That is the actual story. Not the dollar figure. The methodology โ and what it implies for every stablecoin holder who assumes their balance sheet is politically neutral.
Context: a closed loop with exactly one open valve
A7 is not a protocol in any sense a crypto reader would recognize. It is a payment network, and the architecture, reconstructed from the FinCEN proposal and OFAC's designation, runs in a straight line.
Ruble deposits are placed with Promsvyazbank โ PSB โ the Russian state bank that has sat under blocking sanctions since 2018 and functions as a core node in the financing architecture of sanctioned Russian state activity. Those deposits back the issuance of A7A5, an internal accounting and settlement token. A7A5 then has to become something the outside world accepts. It cannot be spent on-chain, cannot be lent, cannot be swapped permissionlessly, cannot be held as collateral. So intermediaries convert it to USDT. USDT moves to an exchange or an OTC desk. The desk pays out fiat.
Four steps. Three of them sit outside U.S. jurisdiction. One of them does not.
It is worth being precise about what A7A5 is not, because the word "stablecoin" keeps getting attached to it. A stablecoin in any conventional sense has a redemption mechanism, a market-making layer defending a peg, and some public accounting of supply. A7A5 has none of that. Redemption is a bilateral arrangement with a sanctioned bank. Supply is undisclosed by design โ publishing it would hand enforcement a target list. There is no arbitrage loop keeping the price honest, because there is no market to arbitrage against. What exists is a credit claim on PSB, denominated in rubles, wrapped in a token that only works if one specific conversion path stays open.
The network's funding sources, per the designation, include Iran's central bank and the Islamic Revolutionary Guard Corps, alongside Russian entities. That composition shapes the system's behavior. State-directed evasion networks do not optimize for yield. They optimize for continuity, and they will pay almost any spread to keep a channel breathing. Which is precisely why the intermediaries are the pressure point: they are the only participants in the chain who have something sanctions can actually take.
FinCEN opened a 30-day public comment period on a proposed rule that would prohibit regulated financial institutions from sending or receiving funds connected to A7 sub-agents โ including transactions with custodial crypto addresses. Institutions receiving funds from a designated address are expected to freeze them where sanctions rules require it, or otherwise reject the transfer and block the receiving party's access. Sub-agent identities will be distributed through FinCEN's secure FI-Portal, and institutions are expected to run risk-based procedures to detect them. OFAC's designation took effect immediately. The 50% rule applies as usual: any entity majority-owned by a designated party is automatically captured, which extends exposure well beyond the named sub-agents.
The U.K. is running parallel. British authorities are investigating a separate pipeline valued at roughly $86 billion and have doubled sanctions penalties. Two jurisdictions, two investigations, one direction of travel.
The plumbing is the vulnerability
I spent two weeks in 2022 auditing the TerraUSD peg mechanics, and the lesson that stuck was not about algorithmic stablecoins specifically. It was that every synthetic settlement system has exactly one valve, and the valve is always the least glamorous part of the design.
A7A5's valve is USDT. The entire $17 billion apparatus rests on a single chokepoint its operators do not control and cannot replace quickly. USDT is the deepest, most widely accepted dollar instrument in crypto. That is exactly what makes it useful to A7 โ and exactly what makes it a liability. There is no substitute with comparable depth. A migration to a smaller stablecoin would cut acceptance and widen slippage on every leg of the trade, which is fatal for a business running on thin, high-frequency flows. A migration to Bitcoin is worse. Post-ETF Bitcoin is a surveilled instrument held largely inside regulated custody wrappers; the peer-to-peer cash story ended the day the custody banks arrived. The asset most useful for evasion is the one Wall Street does not control โ which, today, is a centralized dollar token with a freeze function and a compliance department.
Read the structure and the asymmetry becomes obvious. A7A5 is unfreezable by design because it lives in isolation. That same isolation makes it worthless the moment it cannot exit. The network built a fortress and then put the only gate on someone else's property.
The intermediary layer is where the economics live, and it is worth spelling out. A desk converting A7A5 to USDT is not providing a public service; it is charging a compliance risk premium, and that premium is the real product being sold. The spread widens as enforcement pressure rises, which means the sanctions campaign functions as a tax on the channel rather than a shutdown of it. Desks with U.S. dollar clearing relationships, U.S. counterparties, or U.S. personnel will exit first. Desks without those exposures have no reason to move at all.
On-chain forensics firms will find the conversion leg quickly. It is the only step in the chain that touches public infrastructure, and therefore the only step that Chainalysis, Elliptic, and TRM Labs can cluster, label, and hand to enforcement. My working assumption, based on how these attribution pipelines behaved during the 2020 Compound flash-loan window and every sanctions case since, is that the relevant address clusters were assembled before the designation dropped, not after. Designations at this scale are the output of an investigation, not its opening move. The public is reading the conclusion.
Tether's incentive structure is not ambiguous. The company has a long record of freezing addresses on law-enforcement request, and it will almost certainly do so here, at scale, because cooperation buys regulatory goodwill at a moment when stablecoin issuers are fighting for legislative legitimacy in Washington and Brussels. Expect the freeze to be fast, public, and framed as a compliance success. It is also the clearest demonstration yet that the largest dollar instrument in crypto ships with a censor switch.
Now the data hygiene, which the coverage has mostly skipped. FinCEN's figure is roughly $17 billion across 18 months. Russian self-reported figures put network throughput near $91.5 billion and claim it handles about 13% of relevant trade flows. Those numbers cannot both be right, and the self-reported set includes a date range extending into the future โ a tell that it was assembled for marketing rather than audit. Independent verification of the network's 2,000-plus daily transactions does not exist. When a sanctions-evasion operation publishes its own volume statistics, the correct posture is skepticism with a wide error bar, not credulity in either direction.
What the numbers do establish is the shape of the thing. A7A5 is not a Ponzi. It has no yield promise and no new-money-pays-old-interest structure. It is a credit-dependent shadow currency: its stability equals PSB's ability to honor ruble deposits, multiplied by the continued availability of a USDT conversion path. That is a weaker foundation than any algorithmic stablecoin I have audited, because the failure mode is not a bank run โ it is a bank and a chokepoint failing at the same time, with no governance process capable of responding to either.
The angle nobody is pricing
The designation is not the story. The story is that this strategy works as legislative ammunition far more effectively than it works as enforcement.
Every case like this โ Tornado Cash, Garantex, and now A7 โ becomes a data point in the argument that stablecoins require an AML framework with freeze capability embedded in the standard. That argument is already winning in Brussels and advancing in Washington. A7 is not the target of the campaign; it is the exhibit. The enforcement action is short-term and bounded. The regulatory precedent is long-term and permanent.
Liquidity doesn't respect sanctions regimes โ it respects gradients. Strike A7 and the flow does not stop. It migrates toward whatever channel carries the lowest compliance exposure, which today means unregulated OTC desks in the UAE, Turkey, and Kazakhstan, and tomorrow means whatever replaces them. The real bottleneck in this system was never inside Russia. It was always the off-ramp jurisdictions, and that is where the secondary-sanctions question gets decided. A desk with no U.S. dollar clearing relationship and no U.S. nexus has no reason to comply. That is jurisdiction shopping, and it is why I expect this campaign to produce headlines for years and friction reduction rather than elimination.
Strategic pivots aren't announced โ they are repriced by whoever has to absorb the compliance cost. Watch the exchanges with U.S. exposure: they will upgrade counterparty screening, reprice Russian-adjacent business, and quietly exit segments they cannot defend. The collateral damage lands on ordinary users whose addresses happen to share a cluster with something flagged. De-risking is not a side effect of this policy. It is the policy, expressed through risk-averse intermediaries that will always prefer over-blocking to ambiguity.
Two consequences follow that the market has not fully digested. The most certain winners are the compliance and forensics vendors. Matching FI-Portal lists against live address clusters in real time is not a spreadsheet exercise; it is a product, and demand for it just became non-discretionary across every institution with a U.S. touchpoint. Separately, the screening logic will not stay confined to exchanges. DeFi front-ends, bridges, and RPC providers with any institutional ambition will be pulled into the same apparatus, and the governance theater that already produces arbitrary interest-rate curves on Aave and Compound will be asked to enforce a sanctions regime it has no mechanism to adjudicate. That is a governance problem wearing a compliance costume.
For anyone holding stablecoins through this bear market, the practical question is not whether A7 survives. It is whether the instrument in your wallet has a censor switch, who controls it, and what happens to your balance when that switch is flipped for a reason that has nothing to do with you. Today's answer is uncomfortable, and it is now documented in a federal filing.
What to watch
Three signals, ordered by information value. Tether's blacklist activity: a large cluster of A7-linked freezes confirms the chokepoint thesis and reopens the stablecoin-censorship debate in a way price action will not capture. The final FinCEN rule text once the comment period closes: a softened rule tells you the industry's lobbying apparatus still has leverage, a hardened one tells you the framework is being locked in for a decade. And the emergence of replacement pipes โ new ruble- or rial-linked settlement tokens appearing alongside fresh OTC intermediaries.
That third signal is the one that matters. Sanctions are a tax on evasion, not a prohibition of it. The only question left is who pays it.