Thousands of low-value transfers, reportedly sent from wallets linked to sanctioned HTX, reportedly landed in Kraken customer accounts, per CoinDesk. The customers reportedly did nothing. They reportedly asked for nothing. And Kraken briefly locked their accounts anyway.
That is the scene to hold in your head while reading the Blockchain Association's new nine-point framework, because the framework's central claim is that compliance controls "can, and should sit where institutions and customers actually exercise control."
In the Kraken case, reportedly nobody exercised anything, and sanctioned funds reportedly simply arrived, per CoinDesk.
The framework's central claim is that compliance controls can, and should, sit where institutions and customers actually exercise control.
Start with the argument on its own terms, because it is a good one. The paper — reportedly from Rebecca Rettig, Omid Malekan, and Michael Mosier — sets out to answer a narrow question: can U.S. financial institutions build and transact on permissionless networks while complying with the Bank Secrecy Act and sanctions law?
Their answer is yes, and the legal reasoning holds up. The BSA and U.S. sanctions laws require "reasonable, risk-based compliance programs," and the paper is right that these laws "understand the reality that institutions cannot eliminate every possible risk."
From that premise comes the framework's key move: pulling compliance off the validator set. For blockchains, the paper argues, these laws "do not require a permissioned network, known validator set, or closed group of participants."
Instead, controls belong "at points of genuine intermediation on the application layer" — KYC checkpoints, wallet and counterparty screening, transaction monitoring, Travel Rule compliance, sanctions requirements. The paper names the specific chokepoints: on- and off-ramps, wallet connection, and the moments "when a network interoperates with the institution itself."
This is the correct reading of how risk-based law actually works. It does not demand that a bank vet every validator or recreate a closed system; it demands that the bank manage risk at the points it controls. The paper's slogan — "permissionless does not mean non-compliant" — is defensible as a statement about legal architecture.
It is also a coherent doctrine, not a one-off. The same week, in its rebuttal to the National Sheriffs' Association over the CLARITY Act, the Association argued the bill "regulates based on what an intermediary actually does and what control it holds over customer funds and transactions."
That is the identical logic: obligations attach to control, not to proximity or to protocol architecture. The intermediation framework and the CLARITY defense are two applications of one control-based theory of who owes what. Read together, they tell you this is a considered lobbying position the industry intends to hold.
So the law is sound. The problem is operational, and it showed up the same week the framework did.
Consider what OFAC actually did. Treasury/OFAC issued its first-ever sectoral determination covering Iran's digital assets sector, under a newly launched program called Operation Economic Outcast.
Sectoral is the word that matters. This is not a designation of a named exchange or a specific mixer; it reaches an entire category of activity. The determination also targets a cyber-espionage group inside Iran's Ministry of Intelligence and Security, naming multiple cryptocurrency addresses tied to several members.
Now line that up against the framework's checkpoints. On/off-ramps screen customers moving between fiat and crypto. Wallet connection screens the counterparty a customer chooses to connect to. Transaction monitoring watches what the customer does.
Every one of those controls governs a moment the customer initiates. A sectoral determination does not politely confine itself to those moments. It sweeps across addresses and flows, and much of what it touches will never pass through an application-layer chokepoint at all — it lives at the settlement layer, on the chain itself.
The Kraken episode makes the gap concrete rather than theoretical. Wallets linked to sanctioned HTX reportedly sent thousands of low-value transfers in what looked like a dust attack, per CoinDesk.
Dust attacks are inbound by definition. The sender chooses the recipient; the recipient has no say. There is no on-ramp to screen, no counterparty the customer connected to, no transaction the customer authorized — because the customer did not transact.
The framework governs the moment a customer chooses to act. A dust attack is precisely the case where the customer chooses nothing.
So Kraken did what an application-layer institution can reportedly do when sanctioned value shows up uninvited: it froze accounts.
This is the seam the paper leaves open. Its nine points are built around intermediation — the transactional handshake between institution, customer, and network. But sanctions enforcement, reportedly, is increasingly ambient and inbound. It arrives as sectoral scope and as unsolicited on-chain contact, neither of which waits for a customer to exercise control.
None of this refutes the framework's legal claim. It refines the boundary of it. "Permissionless does not mean non-compliant" is true for the risks that flow through intermediation points. It says nothing about risks that arrive at addresses the institution never chose to interact with.
And here is the exposure. When sanctioned funds land in a wallet through no act of the holder, what does a reasonable, risk-based program require? Freeze everything, as Kraken did, and you punish the innocent and hand attackers a cheap denial-of-service weapon — dust a competitor's customers and watch their accounts lock. Do nothing, and you are holding sanctioned value.
The framework does not answer that, because it was not written to. It answers whether institutions can participate in open networks at all, and it answers well. But the live enforcement events point at a second question the paper does not reach: what is the standard for unsolicited, inbound, on-chain contact?
Until a regulator blesses a specific screening standard for that case — a safe harbor for received dust, a rule for handling addresses caught by sectoral scope rather than named designation — institutions are left improvising with the blunt instrument of the freeze.
That is the real state of play. The Blockchain Association has made the strongest available legal case that open networks and financial-integrity law are compatible, and it has made it twice now with the same control-based logic.
But a legal argument is not a safe harbor. The framework tells an institution where to put its checkpoints; it cannot tell the institution what to do when sanctions enforcement routes around every checkpoint and lands directly at the settlement layer.
Iran's sectoral determination and Kraken's frozen customers arrived in the same week as the paper, and together they mark the exact edge of its coverage. The framework is right about the moment of intermediation. Enforcement has reportedly already moved past that moment.
Until the standard for inbound, ambient sanctions contact exists in writing, "permissionless does not mean non-compliant" remains what it is: a sound argument, and not yet a defense you can bank on.